From danielk at umd.edu Mon Nov 12 08:58:23 2007 From: danielk at umd.edu (Daniel Kessler) Date: Mon, 12 Nov 2007 09:58:23 -0500 Subject: [thelist] frameset styles Message-ID: I have a basic frameset and I'd like to add our stylesheet to it. The left frame is a bunch of links that populate the right frame. The pages that go in the right frame have no styles set. When I add the stylesheet to the page containing the frameset, the style doesn't change in the pages. If I add a "style='font- family:blah;'" to the frame declaration the font isn't set to the page that goes in that frame. It appears that I have to add the style to all the pages that will be called. Is there a way to add this style to the frameset? Am I missing something? thanks. -- Daniel Kessler University of Maryland College Park School of Public Health 3302E HHP Building College Park, MD 20742-2611 Phone: 301-405-2545 http://sph.umd.edu From nan at nanharbison.com Mon Nov 12 14:55:03 2007 From: nan at nanharbison.com (Nan Harbison) Date: Mon, 12 Nov 2007 15:55:03 -0500 Subject: [thelist] How do I delete uploaded files in PHP? Message-ID: <20071112205507.6C84DAB8254@smtp4.34sp.com> Hi all, I am working on a website where clients can upload a resume, and can upload a new resume in its place. This is working just fine. However, I am having the file names be randomly generated, so the old file is not overwritten. I need to be able to delete the old file, but I am having trouble a tutorial on how to do this, all the links I get in google are for little applications to manage file uploads and downloads. -I tried using 'unlink' but it didn't work -I tried this but it isn't working. I didn't get any error messages, it just doesn't delete the old resume. $file = 'resumes/'.$_POST['oldresume']; $ftp_server = 'my ftp address'; $ftpuser = 'myusername; $ftppass = my password; // set up basic connection $conn_id = ftp_connect($ftp_server); // login with username and password $login_result = ftp_login($conn_id, $ftpuser, $ftppass); // try to delete $file if (ftp_delete($conn_id, $file)) { // the ftp function echo '$file deleted successful'; } else { echo 'could not delete $file'; } // close the connection ftp_close($conn_id); Can someone please explain to me how I should do this, or point me to a resource that show me? Thanks! Nan From anthony at baratta.com Mon Nov 12 15:09:21 2007 From: anthony at baratta.com (Anthony Baratta) Date: Mon, 12 Nov 2007 13:09:21 -0800 Subject: [thelist] How do I delete uploaded files in PHP? Message-ID: How are your resumes being posted to the site?? Are you using the built-in PHP File Upload functions? http://us.php.net/manual/en/features.file-upload.php I'm curious as to why you are using FTP at all. Are you receiving them via the web service and the transporting them via FTP to a specific location? Are the files local to the web server? If you login manually with your FTP credentials can you see the files you expect? Can you delete them manually? If the files are local to the web sever, make sure that the web server has the proper permissions on the files. You should be using this PHP function on files that are local: http://us.php.net/manual/en/function.unlink.php -----Original message----- From: "Nan Harbison" nan at nanharbison.com Date: Mon, 12 Nov 2007 12:55:03 -0800 To: thelist at lists.evolt.org Subject: [thelist] How do I delete uploaded files in PHP? > Hi all, > > I am working on a website where clients can upload a resume, and can upload > a new resume in its place. This is working just fine. However, I am having > the file names be randomly generated, so the old file is not overwritten. I > need to be able to delete the old file, but I am having trouble a tutorial > on how to do this, all the links I get in google are for little applications > to manage file uploads and downloads. > > -I tried using 'unlink' but it didn't work > > > -I tried this but it isn't working. I didn't get any error messages, it just > doesn't delete the old resume. > > $file = 'resumes/'.$_POST['oldresume']; > $ftp_server = 'my ftp address'; > $ftpuser = 'myusername; > $ftppass = my password; > > // set up basic connection > $conn_id = ftp_connect($ftp_server); > > // login with username and password > $login_result = ftp_login($conn_id, $ftpuser, $ftppass); > > // try to delete $file > if (ftp_delete($conn_id, $file)) { // the ftp function > echo '$file deleted successful'; > } else { > echo 'could not delete $file'; > } > > // close the connection > ftp_close($conn_id); > > Can someone please explain to me how I should do this, or point me to a > resource that show me? > > > Thanks! > > Nan From nan at nanharbison.com Mon Nov 12 15:20:09 2007 From: nan at nanharbison.com (Nan Harbison) Date: Mon, 12 Nov 2007 16:20:09 -0500 Subject: [thelist] How do I delete uploaded files in PHP? In-Reply-To: Message-ID: <20071112212012.C6602AB8200@smtp4.34sp.com> Anthony, I tried this ftp method as a stupid wild stab at something that might work, I found it on a website explaining how to delete files. Resumes are uploaded using PHP file upload functions, mainly using this function, written by Marcos Thiago : function moveFileToDestination(){ if(is_uploaded_file($this->tmp_name) && $this->dst_dir){ if($this->checkUploadConditions()){ $dst_file_name = ($this->randon_name) ? $this->generateFileName() : $this->fixFileName($this->name); $full_destination_path = $this->dst_dir."/".$dst_file_name; if(file_exists($full_destination_path) && !$this->replace) { $this->error_type = 7; } else { if(@move_uploaded_file($this->tmp_name,$full_destination_path)){ @chmod ($this->dst_dir."/".$dst_file_name,$this->file_perm); $this->error_type = 0; } else { $this->error_type = 5; } } } } else { $this->error_type = 6; } if($this->error_type != 0){ $this->fail_files_track[$this->succeed_track_index]["file_name"] = $this->name; $this->fail_files_track[$this->succeed_track_index]["new_file_name"] = $dst_file_name; $this->fail_files_track[$this->succeed_track_index]["destination_directory"] = $this->dst_dir; $this->fail_files_track[$this->succeed_track_index]["field_name"] = $this->fld_name; $this->fail_files_track[$this->succeed_track_index]["file_size"] = $this->size; $this->fail_files_track[$this->succeed_track_index]["file_type"] = $this->type; $this->fail_files_track[$this->succeed_track_index]["error_type"] = $this->error_type; $this->fail_files_track[$this->succeed_track_index++]["msg"] = $this->msg[$this->error_type]; return FALSE; } else { $this->succeed_files_track[$this->fail_track_index]["file_name"] = $this->name; $this->succeed_files_track[$this->fail_track_index]["new_file_name"] = $dst_file_name; $this->succeed_files_track[$this->fail_track_index]["destination_directory"] = $this->dst_dir; $this->succeed_files_track[$this->fail_track_index]["field_name"] = $this->fld_name; $this->succeed_files_track[$this->fail_track_index]["file_size"] = $this->size; $this->succeed_files_track[$this->fail_track_index]["file_type"] = $this->type; $this->succeed_files_track[$this->fail_track_index]["error_type"] = $this->error_type; $this->succeed_files_track[$this->fail_track_index++]["msg"] = $this->msg[$this->error_type]; return TRUE; } } ++++++++++++++++++++++++++ The unlink function didn't work for me. I will reread the php.net page you sent me. Thanks Nan -----Original Message----- From: thelist-bounces at lists.evolt.org [mailto:thelist-bounces at lists.evolt.org] On Behalf Of Anthony Baratta Sent: Monday, November 12, 2007 4:09 PM To: thelist at lists.evolt.org Subject: Re: [thelist] How do I delete uploaded files in PHP? How are your resumes being posted to the site?? Are you using the built-in PHP File Upload functions? http://us.php.net/manual/en/features.file-upload.php I'm curious as to why you are using FTP at all. Are you receiving them via the web service and the transporting them via FTP to a specific location? Are the files local to the web server? If you login manually with your FTP credentials can you see the files you expect? Can you delete them manually? If the files are local to the web sever, make sure that the web server has the proper permissions on the files. You should be using this PHP function on files that are local: http://us.php.net/manual/en/function.unlink.php -----Original message----- From: "Nan Harbison" nan at nanharbison.com Date: Mon, 12 Nov 2007 12:55:03 -0800 To: thelist at lists.evolt.org Subject: [thelist] How do I delete uploaded files in PHP? > Hi all, > > I am working on a website where clients can upload a resume, and can > upload a new resume in its place. This is working just fine. However, > I am having the file names be randomly generated, so the old file is > not overwritten. I need to be able to delete the old file, but I am > having trouble a tutorial on how to do this, all the links I get in > google are for little applications to manage file uploads and downloads. > > -I tried using 'unlink' but it didn't work > > > -I tried this but it isn't working. I didn't get any error messages, > it just doesn't delete the old resume. > > $file = 'resumes/'.$_POST['oldresume']; $ftp_server = 'my ftp > address'; $ftpuser = 'myusername; $ftppass = my password; > > // set up basic connection > $conn_id = ftp_connect($ftp_server); > > // login with username and password > $login_result = ftp_login($conn_id, $ftpuser, $ftppass); > > // try to delete $file > if (ftp_delete($conn_id, $file)) { // the ftp function > echo '$file deleted successful'; } else { > echo 'could not delete $file'; } > > // close the connection > ftp_close($conn_id); > > Can someone please explain to me how I should do this, or point me to > a resource that show me? > > > Thanks! > > Nan -- * * Please support the community that supports you. * * http://evolt.org/help_support_evolt/ For unsubscribe and other options, including the Tip Harvester and archives of thelist go to: http://lists.evolt.org Workers of the Web, evolt ! From hughesj at firemtn.com Mon Nov 12 15:20:09 2007 From: hughesj at firemtn.com (Jon Hughes) Date: Mon, 12 Nov 2007 13:20:09 -0800 Subject: [thelist] The most semantic way to code a keypad? Message-ID: I am coding a keypad, exactly like a phone. I am unsure what the most semantic way for doing this would be, and am hoping I can get some feedback from you all to help guide my decision. The three options I have are: 1) Table. 2) Div's, floating. 3) Unordered list, floating. The keypad looks like this: -1-2-3- -4-5-6- -7-8-9- -*-0-#- (without the hyphens) What do you think the best way to code this would be? Thanks, - Jon From hughesj at firemtn.com Mon Nov 12 15:27:36 2007 From: hughesj at firemtn.com (Jon Hughes) Date: Mon, 12 Nov 2007 13:27:36 -0800 Subject: [thelist] Semantic Question.... In-Reply-To: <003701c82410$6ee90220$4cbb0660$@net> References: <003701c82410$6ee90220$4cbb0660$@net> Message-ID: > -----Original Message----- > From: thelist-bounces at lists.evolt.org [mailto:thelist- > bounces at lists.evolt.org] On Behalf Of Jon Hughes > Sent: Saturday, November 10, 2007 7:11 PM > To: thelist at lists.evolt.org > Subject: [thelist] Semantic Question.... > > I am creating a keypad in HTML. The full: > > .... Sorry for the apparent double-post list.. I thought this email was "lost in the mail" so to speak. From joel at streamliine.com Mon Nov 12 15:28:37 2007 From: joel at streamliine.com (Joel D Canfield) Date: Mon, 12 Nov 2007 13:28:37 -0800 Subject: [thelist] Semantic Question.... References: <003701c82410$6ee90220$4cbb0660$@net> Message-ID: <72E9FAA171D63B48AAC707C72900E6B4D89F3A@ireland.spinhead.com> > 1 2 3 > > 4 5 6 > > 7 8 9 > > * 0 # semantically, it's an ordered list, yes? or, perhaps, unordered, if we consider the numerical glyphs to be simply glyphs instead of numbers. semantically, it depends somewhat on what the keypad references. phone numbers? they're just glyphs. a calculator? they're numbers. joel From anthony at baratta.com Mon Nov 12 15:32:28 2007 From: anthony at baratta.com (Anthony Baratta) Date: Mon, 12 Nov 2007 13:32:28 -0800 Subject: [thelist] How do I delete uploaded files in PHP? Message-ID: Two quick questions: What are the permissions being set or defaulted to when the file is moved to it's final destination by the code example below? e.g. what is the value of: $this->file_perm And if you "print file_exists($full_destination_path_and_name_to_remove);" What do you get? -----Original message----- From: "Nan Harbison" nan at nanharbison.com Date: Mon, 12 Nov 2007 13:20:09 -0800 To: "'Anthony Baratta'" anthony at baratta.com, thelist at lists.evolt.org Subject: RE: [thelist] How do I delete uploaded files in PHP? > Anthony, > > I tried this ftp method as a stupid wild stab at something that might work, > I found it on a website explaining how to delete files. Resumes are uploaded > using PHP file upload functions, mainly using this function, written by > Marcos Thiago : From anthony at baratta.com Mon Nov 12 15:34:24 2007 From: anthony at baratta.com (Anthony Baratta) Date: Mon, 12 Nov 2007 13:34:24 -0800 Subject: [thelist] How do I delete uploaded files in PHP? Message-ID: P.S. Another option would be to "name" the new file as the olde file and force a delete via your upload code. From meganwh at wi.rr.com Mon Nov 12 16:20:48 2007 From: meganwh at wi.rr.com (Megan Holbrook) Date: Mon, 12 Nov 2007 16:20:48 -0600 Subject: [thelist] [JOB] ASP.NET programmer job Message-ID: <4738D1C0.5020500@wi.rr.com> Hi all - We received a call from www.memorytogo.com who is looking for an ASP.NET programmer. I don't have more information than that, but please contact Azad Faliv (azad at componex.com) if you are interested, or can recommend someone. Thanks, M. -- Megan Holbrook - megan at kapow.com Partner - Business Development kapow, inc. (www.kapow.com) - website design, development & consulting Most recent website launches: PBS Seeing in the Dark (www.pbs.org/seeinginthedark) Skirball Cultural Center (www.skirball.org) Declare Yourself (www.declareyourself.com) kapow, inc. Milwaukee kapow, inc. Los Angeles 2405 E. Wyoming Place 1301 Montana Ave., Suite A Milwaukee, WI 53202 Santa Monica, CA 90403 T: 414.273.2446 * F: 414.278.9056 T: 310.394.5276 * F: 310.394.5278 www.linkedin.com/in/meganholbrook From nan at nanharbison.com Mon Nov 12 18:10:21 2007 From: nan at nanharbison.com (Nan Harbison) Date: Mon, 12 Nov 2007 19:10:21 -0500 Subject: [thelist] How do I delete uploaded files in PHP? In-Reply-To: Message-ID: <20071113001025.BD24EAB8265@smtp4.34sp.com> Anthony, The file->perm is 0444, and the print you suggested to print out had two // 's in it --- resumes//file_to_delete.txt. So that was the problem! I finally got the unlink function to work. Wow, did I waste a lot of time on that. If it didn't work soon, I was going to use your second suggestion, to overwrite the file. Thanks for all your help! Nan -----Original Message----- From: thelist-bounces at lists.evolt.org [mailto:thelist-bounces at lists.evolt.org] On Behalf Of Anthony Baratta Sent: Monday, November 12, 2007 4:32 PM To: thelist at lists.evolt.org Subject: Re: [thelist] How do I delete uploaded files in PHP? Two quick questions: What are the permissions being set or defaulted to when the file is moved to it's final destination by the code example below? e.g. what is the value of: $this->file_perm And if you "print file_exists($full_destination_path_and_name_to_remove);" What do you get? -----Original message----- From: "Nan Harbison" nan at nanharbison.com Date: Mon, 12 Nov 2007 13:20:09 -0800 To: "'Anthony Baratta'" anthony at baratta.com, thelist at lists.evolt.org Subject: RE: [thelist] How do I delete uploaded files in PHP? > Anthony, > > I tried this ftp method as a stupid wild stab at something that might > work, I found it on a website explaining how to delete files. Resumes > are uploaded using PHP file upload functions, mainly using this > function, written by Marcos Thiago : -- * * Please support the community that supports you. * * http://evolt.org/help_support_evolt/ For unsubscribe and other options, including the Tip Harvester and archives of thelist go to: http://lists.evolt.org Workers of the Web, evolt ! From martin at easyweb.co.uk Mon Nov 12 18:57:01 2007 From: martin at easyweb.co.uk (Martin Burns) Date: Tue, 13 Nov 2007 00:57:01 +0000 Subject: [thelist] Top 10 reasons to make your page accessible... In-Reply-To: <30bd6ffd0711110901y160ed7d1x5720813a94d435fd@mail.gmail.com> References: <8720012.111169546702017.JavaMail.root@redroom.customer.ultraspeed.co.uk> <4734D209.6080808@charlestonwebsolutions.com> <30bd6ffd0711110901y160ed7d1x5720813a94d435fd@mail.gmail.com> Message-ID: <93133133-AAB4-4706-8899-EABE137AEBD7@easyweb.co.uk> On 11 Nov 2007, at 17:01, Christian Heilmann wrote: > Why not collect the top 10 reasons why you shouldn't make your site > accessible? That is less evangelizing and more "ah that is a problem, > but here is a solution". > http://www.digital-web.com/articles/ten_reasons_clients_dont_care_about_accessibility/ > > Feeling as if we are in a time bubble... Yes and no... current day job is a new venture for very large consumer company that has an enormous interest in branding, particularly aimed at youth markets. So as you'd imagine, being 'cool' is not unimportant. New venture has a very significant web element to it (ie nearly all new customers are expected to come in over via the web channel, and the call centre app is pretty much an expanded version of the same site). So, given the above, you'd expect that said site would have to be strong-armed into accessibility, right? Sorry, no cigar. The CEO of the client (you'd recognise the name...) is reported by those who talk to him as having a personal concern that the site be accessible. Trouble is, as you've eluded in the digi-web article, it's seen as a technical matter, so cue lots of retro-fitting towards the end of the build phase. The HTML bits of the site were relatively easy (not least because they were built in a fairly accessibility-friendly way to start with). However... much of the operational bit of the site makes heavy use of Adobe Flex (which ain't anything like as bad as old- stylee Flash for accessibility), so the level of pain involved is not insignificant. Cheers Martin -- > Spammers: Send me email -> yumyum at easyweb.co.uk to train my filter > http://dspam.nuclearelephant.com/ From bobm at dottedi.biz Mon Nov 12 19:46:35 2007 From: bobm at dottedi.biz (Bob Meetin) Date: Mon, 12 Nov 2007 18:46:35 -0700 Subject: [thelist] transition out of package site Message-ID: <473901FB.2020807@dottedi.biz> I am preparing to transition a client out of a package site. It came with a ton of generic duplicate content files (50-75). The new site with have between 15-20, but with my help, more focused, individualized. Almost all the files on the old site are templated html as opposed to php on the new. Page rank to the extent that it means anything is abysmal on all pages. Is there anything I need to do to tell google about the pages that are getting their walking papers? For the ones that are staying but renamed to .php should I do anything such as creating links or redirects to the new pages? -- Bob From dejan at kozina.com Mon Nov 12 20:53:47 2007 From: dejan at kozina.com (Dejan Kozina) Date: Tue, 13 Nov 2007 03:53:47 +0100 Subject: [thelist] Microformats question - SOLVED In-Reply-To: <4734B9FF.8090002@kozina.com> References: <4732B09E.4050007@kozina.com> <4eedb92a0711081649h610df43by7063779607fb6f9b@mail.gmail.com> <4734B9FF.8090002@kozina.com> Message-ID: <473911BB.4090900@kozina.com> Thanks to some good advice from the microformats list I got the answer and I'm passing it along just in case. Google Calendar started to behave when I shortened the query down to 779 urlencoded chars. All it took was to move the 'span class="description"' closing tag forward to contain approx. 400 characters in plain text and Google played ball. djn Dejan Kozina wrote: > Thank you. At least now I know it's not something on my PC. I've just > subscribed to the microformats mailing list in hope somebody there will > be willing to take a look at it. > > djn > -- ----------------------------------------- Dejan Kozina Web design studio Dolina 346 (TS) - I-34018 Italy tel./fax: +39 040 228 436 - cell.: +39 348 7355 225 skype: dejankozina http://www.kozina.com/ - e-mail: dejan at kozina.com From codepo8 at gmail.com Tue Nov 13 04:16:58 2007 From: codepo8 at gmail.com (Christian Heilmann) Date: Tue, 13 Nov 2007 10:16:58 +0000 Subject: [thelist] The seven rules of Unobtrusive JavaScript Message-ID: <30bd6ffd0711130216q2dcc11c8k688db9d506dd2a5d@mail.gmail.com> I am giving a workshop in Paris on Saturday on Unobtrusive JavaScript and took the opportunity to analyze my older materials on the subject. Here's what I came up with. http://icant.co.uk/articles/seven-rules-of-unobtrusive-javascript/ After the workshop I'll also have the code to go along with it and will send that around in case you want to run own trainings on that. Regards Chris -- Chris Heilmann Book: http://www.beginningjavascript.com Blog: http://www.wait-till-i.com Writing: http://icant.co.uk/ From david at dorward.me.uk Tue Nov 13 04:12:21 2007 From: david at dorward.me.uk (David Dorward) Date: Tue, 13 Nov 2007 10:12:21 +0000 Subject: [thelist] frameset styles In-Reply-To: References: Message-ID: <82BEA39C-6BFA-4562-A4B0-ABCC9F42AF15@dorward.me.uk> On 12 Nov 2007, at 14:58, Daniel Kessler wrote: > I have a basic frameset and I'd like to add our stylesheet to it. > The left frame is a bunch of links that populate the right frame. Uh oh. http://allmyfaqs.net/faq.pl?Problems_with_using_frames > The pages that go in the right frame have no styles set. > When I add the stylesheet to the page containing the frameset, the > style doesn't change in the pages. Pages loaded into frames are separate documents and don't get styles applied from framesets containing them. -- David Dorward http://dorward.me.uk/ http://blog.dorward.me.uk/ From lee.kowalkowski at googlemail.com Tue Nov 13 05:43:28 2007 From: lee.kowalkowski at googlemail.com (Lee Kowalkowski) Date: Tue, 13 Nov 2007 11:43:28 +0000 Subject: [thelist] Semantic Question.... In-Reply-To: <003701c82410$6ee90220$4cbb0660$@net> References: <003701c82410$6ee90220$4cbb0660$@net> Message-ID: <610592c90711130343h13e54bf2s9faac59ec7bd8ac5@mail.gmail.com> On 11/11/2007, Jon Hughes wrote: > My question is, is it semantically correct to use a table in this instance, > or to float them in block-level elements, or to make an unordered list? Interesting, why would floating affect semantics? This is really tricky, I think you should settle for what makes most sense without CSS, in which case, it's got to be a headerless table, which sounds troubling. The semantics are going to come from the affordance of laying out the design of something which has become an everyday thing. If you're worried about accessibility, I'd suggest you have a text-input as fallback. It's really difficult to provide insight without more background. Good question. -- Lee From austin at dotmail.co.uk Tue Nov 13 05:46:09 2007 From: austin at dotmail.co.uk (Austin Harris) Date: Tue, 13 Nov 2007 11:46:09 +0000 (GMT) Subject: [thelist] Top 10 reasons to make your page accessible... In-Reply-To: <4734D209.6080808@charlestonwebsolutions.com> Message-ID: <16843829.261194954369077.JavaMail.root@redroom.customer.ultraspeed.co.uk> As Christian points out, "we must be in a time bubble"... We must be as I can't remember the last time I posted on here, (apart from a javascript question this week). God only knows how long ago my post containing the rather open "Any other points..?" was from. ----- "Jono" wrote: > Austin Harris wrote: > > Any other points that should be raised? > > > The biggest "disabled" visitor/constituent/user is Google. An > accessible site is a Google-friendly site...and of course Yahoo! and > MSN > count too. > > > -- > > *JONO YOUNG* > Designer | Developer | Illustrator > Charleston Web Solutions > /Bringing Higher Standards to the Lowcountry/ > > -- > > * * Please support the community that supports you. * * > http://evolt.org/help_support_evolt/ > > For unsubscribe and other options, including the Tip Harvester > and archives of thelist go to: http://lists.evolt.org > Workers of the Web, evolt ! From dtohidy at hotmail.com Tue Nov 13 06:06:49 2007 From: dtohidy at hotmail.com (DAVOUD TOHIDY) Date: Tue, 13 Nov 2007 07:06:49 -0500 Subject: [thelist] Layout Stability In-Reply-To: <4736662C.3090108@ij.net> References: <47313D1A.5030705@ij.net> <47328C11.4030206@ij.net> <4734CF1E.9060503@ij.net> <4736662C.3090108@ij.net> Message-ID: On Date: Sat, 10 Nov 2007 21:17:16 -0500 Felix Miata wrote: //////////////// Your comment starts here ///////////// >I had nothing to do with the creation of the site from which >that image wasmade. http://dancesrq.homestead.com/ is> the very first site ever built by afriend's daughter. >She used homestead.com's garbageware to create> it becauseit was "easy" and inexpensively hosted. >She created it shortly before herwindoz computer broke>and was replaced by a Mac. Since homestead.com >garbageware is not supported or offered for the Mac, >shehas not yet done any updating based upon any >knowledge she gained fromexamining > http://mrmazda.no-ip.com/auth/Sites/ksc/ >which on the other handwas created by >me _based_upon_her_ homestead.com site. >I did it originally asan exercise designed to show her the>difference between valid standards-basedHTML and CSS >and the tag soup that homestead.com litters the internet with. ////////////// Your comment ends here //////////////// Here it is the URL for the above comment as below: http://lists.evolt.org/archive/Week-of-Mon-20071105/192558.html My thread's subject is "Layout Stability" not "Homestead.com (was: Layout Stability)"? . I did bring all of your previous comment within myreply and changed the sunject back to original subject to make sure that if somebody browses my thread by subject will not miss my comment in here and your previous comment above!! How you can let yourself to change subject of somebody else's thread??? This is sure enough an unprofessional and unethical behaviour. A forth grade student will act more professional than you . You definitely need to take some courses onhow to act ethical and professional within a technicaland professional community. >I did it originally asan exercise designed to show her... Did you do it to show her or to show me? And have you doneit so many years ago or just NOW? Here it is the proof in your previous comment about it: ////////// Your previous comment starts here //////////// >OK, you want revisit? Here it is: >http://mrmazda.no-ip.com/SS/Davoto/davoto >It has needed doing for some time, >so I took quite a bit of time to do it now>to use both here and as future reference to show poorly-adaptive versus>web-friendly resolution-agnostic design. //////// Your previous comment ends here ///////////// and here it is the url for above comment: http://lists.evolt.org/archive/Week-of-Mon-20071105/192550.html and I will leave it to the List to determine if your comment in regards to recoding that site for that lady is true or a lie. >fromexamining http://mrmazda.no-ip.com/auth/Sites/ksc/ Well Thanks a lot for the link and I hope everybody has accessto the above URL. Even though I just received another reply from you, , this time, with the correct subject!!! with different contentbut I am not going to really go deeper in your layout and find defects for FUTURE REFFERENCE!! . I will not bring myself that down professionally. BTW I will reply to your next reply seperately. Sure enough if i really want to, I will find enough defect in your own layout. Further more you need to recode your own layout athttp://mrmazda.no-ip.com/menu.html which is a table based layout,and make it accessible and usable before talking about theaccessibilty and usability of somebody else's layout. I did not bother looking for other defects in your own site at http://mrmazda.no-ip.com . And I do not intend to make a demo here to show it to everybody. Appraently you are just migrating from table based layout to css based layout. I will have a quick comment on Dancesrq layout and againI will not bother going deeper . Definitely, there is no such a perfect site today on theinternet to be considered as an ultimate solution and all we do is to find a point in the middle where we can satisfy the mostcommon browsers and resolutions. That is what you have donethere too. I don't have access to higher resolutions today becauseI am out of office , I will try it later myself. However it would benice if you could provide screen captures of your layout at higher resolutions,specificly those that you were asking me to design for them like 1920x1200 and 1680x1050 ,as well as screen captures on different browsers and platforms please. I have your current layout in my computer FOR FUTURE referrence, so please just show me the screen captures of that layout without any modification! Browsercam has a trial version and it is free. It would be greatif i can see some of it in browsercam. It does not provide screen capture at resolutions of higher than 1280 though. Apparently your layout will be like a very narrow column whena specific css rules applied. Otherwise it will not be a stable layout. Even though your layout fits nicely on screen at the resolution of 600 * 400 however this resolution is deprecated :) and I will not design for this resolution ever, unless my client asks me to. But I will check my layout to see how it look like. Having a scroll bar at this resolution is not important to me andI consider this resolution as a last priority. Browse your layout and tap on the ctrl+ at different resolutionsand see the effect of it. Surely there is no layout stability after some incerements. But that is O.K to me. It is you showing it as a defect in my layout. I can see there is no layout stability at some resolutions based on your different css rules or if there is then there is a scroll bar for example when browsing your layout with the dancesrqdpref.cssrule at 800 it does not have layout stability at all and when browsing it with dancesrqdno100.css it does have scrollbar at this resolution. Remember that your demo shows my portfolio has a scrollbarat 600*400 and you have mentioed it in your demo to show it as a defect. But your layout has a scroll bar at 800 and or isnot a stable layout. I would rather to have a stable layout without a scroll bar at 800 *600 and have a stable layout with scroll barat 600 *400, rather than having a scroll bar or unstable layout at 800 and an unstable layout without a scroll bar at 600. Surely I can provide a Demo for FUTURE refferrence! However, I do not intend to demonstrate any defect in your layout!! I believe technical capability is necessary to be a professional but it is not enough. To be considered as a professional, one needs to have professional ethics before even having the necessarytechnical knowledge. One before even SHOWING that s/he is trying to solve a problemfor somebody else, needs to solve his/her own problems in real life and in his/her own web site and then providea solution for somebody else. _________________________________________________________________ Are you ready for Windows Live Messenger Beta 8.5 ? Get the latest for free today! http://entertainment.sympatico.msn.ca/WindowsLiveMessenger From martin at easyweb.co.uk Tue Nov 13 06:10:20 2007 From: martin at easyweb.co.uk (Martin Burns) Date: Tue, 13 Nov 2007 12:10:20 +0000 Subject: [thelist] Semantic Question.... In-Reply-To: <003701c82410$6ee90220$4cbb0660$@net> References: <003701c82410$6ee90220$4cbb0660$@net> Message-ID: <054A9967-26B4-43E7-9418-6D0553B75B1F@easyweb.co.uk> On 11 Nov 2007, at 03:10, Jon Hughes wrote: > I am creating a keypad in HTML. The full: > > 1 2 3 > 4 5 6 > 7 8 9 > * 0 # See, I have a bigger question - why? Looks like a case of stretching the metaphor too far to me. Would you create a full onscreen keyboard for every text input? This input layout is great for physical fingers, but a royal PITA with a mouse. So why not just allow the user to use their real world keypad and have a simple input box? I'm sure there's a Jakob on this somewhere... Cheers Martin -- > Spammers: Send me email -> yumyum at easyweb.co.uk to train my filter > http://dspam.nuclearelephant.com/ From kasimir.k.lists at gmail.com Tue Nov 13 07:11:17 2007 From: kasimir.k.lists at gmail.com (kasimir-k) Date: Tue, 13 Nov 2007 13:11:17 +0000 Subject: [thelist] Semantic Question.... In-Reply-To: <003701c82410$6ee90220$4cbb0660$@net> References: <003701c82410$6ee90220$4cbb0660$@net> Message-ID: <4739A275.3050707@gmail.com> Jon Hughes scribeva in 11/11/2007 3:10: > I am creating a keypad in HTML. The full: > > > > 1 2 3 > > 4 5 6 > > 7 8 9 > > * 0 # > > > > My question is, is it semantically correct Which means..? With a paragraph of text it's easy: you mark it up using the element for paragraph. What is the text in this case then? Well, a keypad, but we don't have . Is it tabular data? Do the columns and rows have meanings? I didn't think so either. Is it a list then? - 1 - 2 - 3 - 4 - 5 - 6 - 7 - 8 - 9 - * - 0 - # Nope, doesn't make sense as a list. > to use a table in this instance, > or to float them in block-level elements, or to make an unordered list? And as Lee pointed out, floting really has nothing to do with semantics here. But what you keypad looks to me is preformatted text... If you need links remember that you can have inline elements inside 
 (except 
img, object, sup, sub, big, small).

.k

From danielk at umd.edu  Tue Nov 13 07:11:24 2007
From: danielk at umd.edu (Daniel Kessler)
Date: Tue, 13 Nov 2007 08:11:24 -0500
Subject: [thelist] frameset styles
Message-ID: 

>> I have a basic frameset and I'd like to add our stylesheet to it.
>> The left frame is a bunch of links that populate the right frame.
>>
>
> Uh oh.
>
> http://allmyfaqs.net/faq.pl?Problems_with_using_frames

oh dang.  Well I suspected as much.



thank you for replying.


-- 

Daniel Kessler

University of Maryland College Park
School of Public Health
3302E HHP Building
College Park, MD  20742-2611
Phone: 301-405-2545
http://sph.umd.edu




From dtohidy at hotmail.com  Tue Nov 13 09:02:09 2007
From: dtohidy at hotmail.com (DAVOUD TOHIDY)
Date: Tue, 13 Nov 2007 10:02:09 -0500
Subject: [thelist] Layout Stability
In-Reply-To: <473654B7.6070701@turmel.org>
References: 
	<47313D1A.5030705@ij.net>	
	<47328C11.4030206@ij.net>	
	<4734CF1E.9060503@ij.net> 
	<473654B7.6070701@turmel.org>
Message-ID: 


 
Phil Turmel wrote:> You are responding with hostility where Felix's 
>tone has been, at worst,> exasperated. 
 
Dear Phil,
 
I have talked to other people in other lists in regards to
"layout stability". You might want to check them out as
well. I have even accepted to learn from some of them. 
 
I do not know what hostility means, until I see there is 
somebody who is following a certain purpose to downsize 
or downgrade somebody or somebody else's work purposely.
 
>That is, he has pointed out a limitation of your> definition of stability, and you keep arguing how well you 
>are meeting that definition. 
 
There is no doubt that there is a limit for "layout stability" and
please read the thread yourself again because I have mentioned 
it in my previous posts. And even I have mentioned a real time
example to help him grasp the definition of layout stability that
I have provided.
 
I am not arguing, I am just telling the facts. It is obvious
that after some increments my layout becomes unstable.
 
However, That is what I have designed for and that is what
serves the purpose. 
 
>You have NOT answered his points. 
 
Well he has not answered mine. I have asked for feedback on
layout stability NOT feedback on my portfolio. And If I have
not then how come you have [snip screed /] :) over there?
 
He has not even given any answer in regards to accepting
or rejecting the idea of coding for layout stability. 
 
Further more he is pushing to downgrade my layout.
 
As I have mentioned before in my previous reply, I do not
suggest that my layout is an ultimate solution.
 
>You clearly weren't expecting criticism when you asked
> for comments in your original post,
Well as I said the subject of my thread is "layout stability".
If I was asking for a feedback on my portfolio everybody
was welcome to say anything about it.
 
But the fact is that I was only expecting about layout 
stability.
 
Surely I am open to any critique as long as I see it is not
in the way of purposely down grading my layout. Again
please read the other lists and you will notice it yourself.
 
Otherwise I wouldn't even responding to him.
 
> and you are therefore taking offense at what seems to 
>me was polite constructive criticism.
 
I am not taking offense at all. He is definitely trying to
downgrade my portfolio purposely.
 
Let me give you example of what he is doing. He changed
the subject of my thread last time to prevent those 
who are or will follow this thread by subject from seeing that
specific comment which he has made.
 
He then gives two urls as below in his last post:
 
> http://mrmazda.no-ip.com/SS/Davoto/sc-davoto3 (screenshot)> http://mrmazda.no-ip.com/SS/Davoto/sc-davoto-i3 (URL from which screenshot made)
 
Please feel free to browse them. The second one "sc-davoto-i3",
will take you to a screen capture. click on it to enlarge and
get the URL from address bar of the screen capture. 
 
Here it is the URL :
 
http://mrmazda.no-ip.com/SS/Davoto/sc-davoto3.html
 
Feel free to browse it and see what kind of person who is and
what personality he has. Don't forget to see how he has 
manipulated my portfolio. 
 
And don't forget to see the
 " The quick brown fox jumps over the lazy dog" sentence
which has been repeated many times just next to my picture. And 
please don't forget to see that there is no juming column
in my portfolio in the screen captures.
 
Juming column will exist only after couple of extra increments 
which I have not designed for it anyways.
 
Just in case if you cannot have access to that url, I am putting
screen captures of it in my server.
 
Here they are the urls:
 
http://cssfreelancer.awardspace.com/captures/fel1.html
 
http://cssfreelancer.awardspace.com/captures/fel2.html
 
 
Now, Won't you be ashamed of calling somebody like him as
a native of yours?
 
There are some "*** " that need to take their medicine before
they join to a professional community.
 
 
> Please set your ego aside and re-read the thread.
 
I will re read,please do so yourself as well. 
 
I will not reply to him anymore. Strike three OUT.
 
Please feel free to contact me off list at any time.
 
[snip: screed /] :)
 
 
Best
davoud
http://cssfreelancer.awardspace.com 
 
_________________________________________________________________
R U Ready for Windows Live Messenger Beta 8.5? Try it today!
http://entertainment.sympatico.msn.ca/WindowsLiveMessenger

From kasimir.k.lists at gmail.com  Tue Nov 13 09:47:47 2007
From: kasimir.k.lists at gmail.com (kasimir-k)
Date: Tue, 13 Nov 2007 15:47:47 +0000
Subject: [thelist] Layout Stability
In-Reply-To: 
References: 	<47313D1A.5030705@ij.net>		<47328C11.4030206@ij.net>		<4734CF1E.9060503@ij.net>
		<4736662C.3090108@ij.net>
	
Message-ID: <4739C723.8050607@gmail.com>

DAVOUD TOHIDY scribeva in 13/11/2007 12:06:
> How you can let yourself to change subject of somebody else's
> thread??? This is sure enough an unprofessional and unethical
> behaviour.

Davoud, please have a look here: 


Point 4 clearly states: "If an ongoing discussion drifts from the 
initial subject, please change the subject line to reflect the new subject."

So Felix has done the right, ethical and professional thing, and read 
the list guidelines before posting here.

> A forth grade student will act more professional than you .

Oh, and remember to read the first point too, the one that reads 
"Respect each other. Diversity in Web development is not only tolerated 
but encouraged. Argument and debate are fine. Character digs and 
personal attacks are not."

> You definitely need to take some courses onhow to act ethical and
> professional within a technicaland professional community.
...
> ... I will not bring myself that down professionally.
...
> I believe technical capability is necessary to be a professional but
> it  is not enough.
...
> To be considered as a professional, one needs to have professional
> ethics before even having the necessarytechnical knowledge.

Davoud, even though the listinfo (link above) tells that "The list is 
intended for professional Web workers", I think you might now be getting 
a little carried away with your cries for professionalism... If your 
intention is to convince the readers about the level of *your* 
professionalism, there is a chance that the intended and actual effects 
don't fully meet.

.k



From dtohidy at hotmail.com  Tue Nov 13 10:10:24 2007
From: dtohidy at hotmail.com (DAVOUD TOHIDY)
Date: Tue, 13 Nov 2007 11:10:24 -0500
Subject: [thelist] Layout Stability
In-Reply-To: <4739C723.8050607@gmail.com>
References: 
	<47313D1A.5030705@ij.net>	
	<47328C11.4030206@ij.net>	
	<4734CF1E.9060503@ij.net>	
	<4736662C.3090108@ij.net> 
	<4739C723.8050607@gmail.com>
Message-ID: 


Date: Tue, 13 Nov 2007 15:47:47 +0000 kasimir.k Wrote:
 
> So Felix has done the right, ethical and professional thing..
 
I don't find it ethical. You can only change subject of your own 
thread not somebody else's.
 
> Oh, and remember to read the first point too,..
 
read the entire thread before replying.
 
 
> Davoud, even though the listinfo ...
 
I do not cry, I do laugh at those who cannot act professionally.
I do not intend to convince anybody. Everybody has the right to
beleive in what they believe in.
 
 
 
 
_________________________________________________________________
Send a smile, make someone laugh, have some fun! Start now!
http://www.freemessengeremoticons.ca/?icid=EMENCA122

From mrmazda at ij.net  Tue Nov 13 10:44:14 2007
From: mrmazda at ij.net (Felix Miata)
Date: Tue, 13 Nov 2007 11:44:14 -0500
Subject: [thelist] Homestead.com (was: Layout Stability)
In-Reply-To: 
References: 	<47313D1A.5030705@ij.net>		<47328C11.4030206@ij.net>		<4734CF1E.9060503@ij.net>
		<4736662C.3090108@ij.net>
	
Message-ID: <4739D45E.6050703@ij.net>

On 2007/11/13 07:06 (GMT-0500) DAVOUD TOHIDY apparently typed:

> My thread's subject is "Layout Stability" not

> "Homestead.com (was: Layout Stability)"? .

That was the subject of a thread you originated, but the material you
interjected about http://dancesrq.homestead.com/ in
http://lists.evolt.org/archive/Week-of-Mon-20071105/192554.html was about
neither my site nor your own nor the subject of your original post.

My changing the subject line was merely adherence to good thread etiquette,
which recommends to change the subject line to match the content whenever
there is a significant divergence in subject from the original thread subject.

I have done the same thing here, because none of what I write here is
directly about "Layout Stability", just as what I wrote in
http://lists.evolt.org/archive/Week-of-Mon-20071105/192558.html had nothing
directly to do with "Layout Stability".
-- 
"   A patriot without religion . . . is as great a
paradox, as an honest man without the fear of God."
	                             John Adams

 Team OS/2 ** Reg. Linux User #211409

Felix Miata  ***  http://mrmazda.no-ip.com/

From hughesj at firemtn.com  Tue Nov 13 11:12:25 2007
From: hughesj at firemtn.com (Jon Hughes)
Date: Tue, 13 Nov 2007 09:12:25 -0800
Subject: [thelist] "Rotate" div + contents in non-IE browser
Message-ID: 

Howdy list.

I am making an application for a phone, and need it to be landscape
view, not portrait.

I already have it done in HTML/CSS, but (as a lack of foresight on my
behalf) I didn't think about how it would show up on the handheld
device.

So I am, looking for something that will turn this:


 - - -
 - - -
 - - -
 - - -
 - - -

Into this:

| | | | |
| | | | |
| | | | |

Essentially, just rotating the entire content 90 degrees
counter-clockwise.

I would do this with the graphics, but there is dynamic text and an
input box...


Any ideas?

Thanks,

 - Jon

From morrison.ben at gmail.com  Tue Nov 13 11:22:04 2007
From: morrison.ben at gmail.com (ben morrison)
Date: Tue, 13 Nov 2007 17:22:04 +0000
Subject: [thelist] "Rotate" div + contents in non-IE browser
In-Reply-To: 
References: 
Message-ID: <6073aef90711130922y637eff45obe90e0c2239d5c6d@mail.gmail.com>

On Nov 13, 2007 5:12 PM, Jon Hughes  wrote:
> I am making an application for a phone, and need it to be landscape
> view, not portrait.
>
> I already have it done in HTML/CSS, but (as a lack of foresight on my
> behalf) I didn't think about how it would show up on the handheld
> device.
>
...
> Essentially, just rotating the entire content 90 degrees
> counter-clockwise.
>

You'll have to forgive my ignorance when it comes to mobile
application support, but could I ask a few questions?

non-IE browser - which browser or rather Mobile platform is your
application for?

>From my limited knowledge, CSS and in fact HTML support is very poor
when it comes to mobile phones - are you developing for only one
particular mobile device - such as an iPhone....

Can you explain why you have to rotate the contents - I would have
thought that layout is generally done by widths - which can
accommodate different sizes.

When you say Application - is this something that is installed on the
device or a website?

ben
-- 
Ben Morrison
http://www.benjaminmorrison.com

From mrmazda at ij.net  Tue Nov 13 11:35:34 2007
From: mrmazda at ij.net (Felix Miata)
Date: Tue, 13 Nov 2007 12:35:34 -0500
Subject: [thelist] Layout Stability
In-Reply-To: 
References: 	<47313D1A.5030705@ij.net>		<47328C11.4030206@ij.net>		<4734CF1E.9060503@ij.net>
		<4736662C.3090108@ij.net>
	
Message-ID: <4739E066.2060808@ij.net>

On 2007/11/13 07:06 (GMT-0500) DAVOUD TOHIDY apparently typed:

> On Date: Sat, 10 Nov 2007 21:17:16 -0500 Felix Miata wrote:

>> http://mrmazda.no-ip.com/auth/Sites/ksc/

> Browsercam has a trial version and it is free. It would be great if i can
> see some of it in browsercam. It does not provide screen capture at
> resolutions of higher than 1280 though.

Unfortunately, Browsercam has the limitations you found, and others.
Browsercam falls short of providing a full range of window size and screen
resolutions. To test beyond these limitations requires the sophisticated web
design professional to use more comprehensive means, such as, but not limited to:

1-more than one computer
2-more than one display
3-more than one operating system
4-more than one window size
5-more than one actual working DPI
6-more than just a few browser text sizes or zoom levels

At my disposal I have more than 20 working computers, most of which have 3 or
more operating systems installed. I also have one LCD display, and many
different CRT displays, all the latter of which are capable of a wide range
of resolutions, enabling me to switch among them to thoroughly test both the
pages I create, and those created by others, such as
http://cssfreelancer.awardspace.com/ created by you.

> I can see there is no layout stability at some resolutions based on your
> different css rules or if there is then there is a scroll bar for example
> when browsing your layout with the dancesrqdpref.css rule at 800 it does
> not have layout stability at all and when browsing it with
> dancesrqdno100.css it does have scrollbar at this resolution.

http://mrmazda.no-ip.com/auth/Sites/ksc/ was created as a styling exercise,
and includes both those two stylesheets for the precise purpose of comparing
behavior between two basic styling methods, one of which constrains the
content to viewport with, and the other of which does not. Naturally the one
that does not presents the opportunity for a horizontal scrollbar to appear
when viewing it, which is the very nature of the constraint to produce.

> Remember that your demo shows my portfolio has a scrollbar at 600*400 and
> you have mentioed it in your demo to show it as a defect.

Yours has a scrollbar that is roughly 25% of the window width, while mine at
the same time has none at all.

> But your layout
> has a scroll bar at 800  and or is not a stable layout.

I consider the most stable layout to be one that does the best job of
adapting to the font size, DPI, and window size without degrading the user
experience any more than necessary, preferably with no degrading at all.
http://mrmazda.no-ip.com/auth/Sites/ksc/ does a good job of that regardless
which stylesheet is selected, and regardless of the above mentioned variables
in the visitor's viewing environment.

> I would rather to have a stable layout without a scroll bar at 800 *600
> and have a stable layout with scroll bar at 600 *400, rather than having a
> scroll bar or unstable layout at 800 and an unstable layout without a
> scroll bar at 600.

A scrollbar is not always a bad thing, particularly when the amount of scroll
is small and no content is missing from either side of the viewport. Arguably
it is not a problem even when substantial, as when an entire vertical nav
list/menu is missing after scrolling sizeways to bring the entirety of
reasonable width paragraphs into view, allowing the content to be easily read.

In the cases where IE7 and Opera are used to view pages, the default zoom
method is classified as "page". The other kind, used by IE6 (technically not
really zoom), Safari and Gecko browsers, is classified as "text only". The
former means scrollbars are commonplace because of the design of the
browser's zoom feature - they cannot be easily avoided through design of the
browser. OTOH, the latter browsers permit much reduced likelihood that
horizontal scrollbars will become necessary.

I tested http://mrmazda.no-ip.com/auth/Sites/ksc/ with several browsers. Both
IE7 and Opera produced horizontal scrollbars, but only when the resulting
text size grew to an unreasonably large proportion of the window width.
Firefox remained free of horizontal scrollbars even when text size was
increased well beyond a reasonable proportion of window width.

The consequence of these tests show that
http://mrmazda.no-ip.com/auth/Sites/ksc/ gives the user a better experience
than http://cssfreelancer.awardspace.com/, regardless how or whether you
consider either as fitting your definition of "layout stability". Good user
experience is more important than meeting your definition of "layout stability".
-- 
"   A patriot without religion . . . is as great a
paradox, as an honest man without the fear of God."
	                             John Adams

 Team OS/2 ** Reg. Linux User #211409

Felix Miata  ***  http://mrmazda.no-ip.com/

From hughesj at firemtn.com  Tue Nov 13 11:48:07 2007
From: hughesj at firemtn.com (Jon Hughes)
Date: Tue, 13 Nov 2007 09:48:07 -0800
Subject: [thelist] "Rotate" div + contents in non-IE browser
In-Reply-To: <6073aef90711130922y637eff45obe90e0c2239d5c6d@mail.gmail.com>
References: 
	<6073aef90711130922y637eff45obe90e0c2239d5c6d@mail.gmail.com>
Message-ID: 

> 
> non-IE browser - which browser or rather Mobile platform is your
> application for?
> 
> >From my limited knowledge, CSS and in fact HTML support is very poor
> when it comes to mobile phones - are you developing for only one
> particular mobile device - such as an iPhone....
> 
> Can you explain why you have to rotate the contents - I would have
> thought that layout is generally done by widths - which can
> accommodate different sizes.
> 
> When you say Application - is this something that is installed on the
> device or a website?
> 
> ben


Ben,

The browser is essentially firefox (possibly Opera...) - I mentioned
non-IE because I don't want to use proprietary IE "filters" to rotate
the content.

The browser support to pretty much the same as Firefox, it supports
javascript and CSS2.

I am a contractor on this job, and was simply told that it had to be
rotated.  In my own eyes, I agree with you, but that's not what the
client wants :)

And to answer your final question, it is going to be a "website"
fundamentally, but the handheld will be set on this one website, and you
will be unable to navigate away. (the devices sole purpose is to be on
this website)

I hope that helps,

 - Jon

From kendsnyder at gmail.com  Tue Nov 13 12:42:39 2007
From: kendsnyder at gmail.com (Ken Snyder)
Date: Tue, 13 Nov 2007 11:42:39 -0700
Subject: [thelist] "Rotate" div + contents in non-IE browser
In-Reply-To: 
References: 	<6073aef90711130922y637eff45obe90e0c2239d5c6d@mail.gmail.com>
	
Message-ID: <4739F01F.9070300@gmail.com>

Jon Hughes wrote:
> Ben,
>
> The browser is essentially firefox (possibly Opera...) - I mentioned
> non-IE because I don't want to use proprietary IE "filters" to rotate
> the content.
>
> ...
> And to answer your final question, it is going to be a "website"
> fundamentally, but the handheld will be set on this one website, and you
> will be unable to navigate away. (the devices sole purpose is to be on
> this website)...
>   
It seems like you would want the device to run Firefox in landscape mode 
rather than have Firefox rotate the content.

However, Google did produce this idea of embedding HTML in a sideways 
SVG image with Firefox 3's "foreignObject":
http://starkravingfinkle.org/blog/2007/07/firefox-3-svg-foreignobject/

- Ken Snyder

From mrmazda at ij.net  Tue Nov 13 12:44:40 2007
From: mrmazda at ij.net (Felix Miata)
Date: Tue, 13 Nov 2007 13:44:40 -0500
Subject: [thelist] Fair Use under US Copyright Law and Custom (was:
 Layout Stability)
In-Reply-To: 
References: 	<47313D1A.5030705@ij.net>		<47328C11.4030206@ij.net>		<4734CF1E.9060503@ij.net>
		<473654B7.6070701@turmel.org>
	
Message-ID: <4739F098.9060707@ij.net>

On 2007/11/13 10:02 (GMT-0500) DAVOUD TOHIDY apparently typed:

> Phil Turmel wrote:

>> You are responding with hostility where Felix's 

> I am not taking offense at all. He is definitely trying to
> downgrade my portfolio purposely.

> He then gives two urls as below in his last post:

>> http://mrmazda.no-ip.com/SS/Davoto/sc-davoto3 (screenshot)
>> http://mrmazda.no-ip.com/SS/Davoto/sc-davoto-i3 (URL from which screenshot
made)

> Please feel free to browse them. The second one "sc-davoto-i3",
> will take you to a screen capture. click on it to enlarge and
> get the URL from address bar of the screen capture.

> Here it is the URL :

> http://mrmazda.no-ip.com/SS/Davoto/sc-davoto3.html

That's the same page as http://mrmazda.no-ip.com/SS/Davoto/sc-davoto3 which
is something every astute web professional should have noticed if he had gone
to each, and thus would not have needed to mention as a digression from the
subject matter of a mailing list thread.

> Don't forget to see how he has manipulated my portfolio. 

> And don't forget to see the
>  " The quick brown fox jumps over the lazy dog" sentence
> which has been repeated many times just next to my picture.

I did not "manipulate your portfolio". I performed the function of an
educator and critic, making edits designed to demonstrate flaws in the design
and content of the original in order to demonstrate presentation and content
errors, and opportunities to improve the original. This falls under the
purview of copyright law exceptions as an acceptable and fair use of your web
page, something we do here on a routine basis, usually using a subject line
that includes "site check" or its equivalent. The relevant law is quoted below:

"Copyright Act of 1976, 17 U.S.C. ? 107:

    "Notwithstanding the provisions of sections 106 and 106A, the fair use of
     a copyrighted work ... for purposes such as criticism, comment ...
     teaching ... is not an infringement of copyright."

If you take some time to examine http://mrmazda.no-ip.com/auth/Sites/davoto/
, which is the bottom portion of
http://mrmazda.no-ip.com/SS/Davoto/sc-davoto3 set underneath your own
"portfolio" page for the purpose of making a screenshot set in an identical
environment, you should be able to see that that is exactly what it is and
what I did. I hope you can improve your CSS coding, XHTML Freelancing,
Freelance web designing, and cross platform website Freelance Front End
Coding skills from it, so that someday you might become a competent
professional web designer and developer.
-- 
"   A patriot without religion . . . is as great a
paradox, as an honest man without the fear of God."
	                             John Adams

 Team OS/2 ** Reg. Linux User #211409

Felix Miata  ***  http://mrmazda.no-ip.com/

From martin at easyweb.co.uk  Tue Nov 13 14:49:48 2007
From: martin at easyweb.co.uk (Martin Burns)
Date: Tue, 13 Nov 2007 20:49:48 +0000
Subject: [thelist] frameset styles
In-Reply-To: <82BEA39C-6BFA-4562-A4B0-ABCC9F42AF15@dorward.me.uk>
References: 
	<82BEA39C-6BFA-4562-A4B0-ABCC9F42AF15@dorward.me.uk>
Message-ID: <4871A58B-4825-45E4-B072-0827B7E7D799@easyweb.co.uk>


On 13 Nov 2007, at 10:12, David Dorward wrote:

> On 12 Nov 2007, at 14:58, Daniel Kessler wrote:
>
>> I have a basic frameset and I'd like to add our stylesheet to it.
>> The left frame is a bunch of links that populate the right frame.
>
> Uh oh.


As per the other thread, it feels like I'm in a time capsule...  
whyever would one use frames these days?[1]

Cheers
Martin

[1] Actually, there is one reasonable use case - to critique another  
site in context, as per ye olde McLibel site.

--
 > Spammers: Send me email -> yumyum at easyweb.co.uk to train my filter
 > http://dspam.nuclearelephant.com/






From julian.rickards at gmail.com  Wed Nov 14 08:18:55 2007
From: julian.rickards at gmail.com (Julian Rickards)
Date: Wed, 14 Nov 2007 09:18:55 -0500
Subject: [thelist] :before and :after appearing together
Message-ID: <7613e9690711140618r6d90b752y1c3296d7f1818a88@mail.gmail.com>

Hi:

I can't show you the site because it is internal. However, here is the basic
structure:



  
...



-- 
Blog: http://pen-and-ink.ca
E-mail: julian dot rickards at gmail dot com

From webdev at mountain.ch  Wed Nov 14 08:26:14 2007
From: webdev at mountain.ch (Mark Howells-Mead)
Date: Wed, 14 Nov 2007 15:26:14 +0100
Subject: [thelist] Ruby on Rails: why?
Message-ID: <20E8922C-F68B-4022-B3E6-90E6528FE235@mountain.ch>

Hi all

Just to be a bit controversial and perhaps a little OT.

Is Ruby on Rails just a trendy, client-server version of programs  
like GoLive and Frontpage? I see an awful lot of job adverts these  
days demanding experience with RoR, but from the introductory videos  
on the official website, it doesn't seem to be much more than a  
GoLive-type environment, albeit with integral Server-Side features.  
Am I missing something? It seems like a tool to enable developers to  
produce websites with dynamic content using less than optimal code  
and standardized functions, instead of had coding, which is in the  
most cases far superior. I am infinitely quicker coding by hand and  
RoR seems to add steps to the development process, instead of taking  
steps away....

Just $0.02, no offence intended to users and RoR developers!

Cheers
Mark Howells-Mead
- www.permanenttourist.ch
- www.flickr.com/photos/mhowells/



From julian.rickards at gmail.com  Wed Nov 14 08:32:45 2007
From: julian.rickards at gmail.com (Julian Rickards)
Date: Wed, 14 Nov 2007 09:32:45 -0500
Subject: [thelist] :before and :after appearing together
Message-ID: <7613e9690711140632jac5c1eeq11321e6bba391b02@mail.gmail.com>

Hi:

Sorry, I pressed Enter too soon.

Here is a sample of the code:



  
...

  
...

  
...


 

I have floated #primarynav, #primarycontent and #secondarycontent to the
left so in order to have the background of #content appear behind all of
them, I floated #content to the left as well. This works fine.

However, the designer wants curved corners on the background which means
three backgrounds, a top, a middle and a bottom. I could do it with multiple
layered 
s but I decided to apply the backgrounds using #content:before
(for the top background image), #content for the main background image and
#content:after for the bottom background image.

Here is the code:

#content {background: url(middle-bg.png) repeat-y;}
#content:before {content: ""; display:block; height: 10px; background: url(
top-bg.png) no-repeat;}
#content:after {content: ""; display:block; height: 10px; background: url(
bottom-bg.png) no-repeat;}

The results in FF and Opera are that the backgrounds for all of them appear
but the order of the backgrounds is :before, :after, #content so that it
appears as if the curved top is immediately followed by the curved bottom
and then the middle background.

I assumed that #content:after would appear after #content but this doesn't
seem to be the case.

Any ideas?

----------------
Now playing: Ozric Tentacles -
Erpsongs
posted with FoxyTunes 

-- 
Blog: http://pen-and-ink.ca
E-mail: julian dot rickards at gmail dot com

From beertastic at gmail.com  Wed Nov 14 11:01:28 2007
From: beertastic at gmail.com (Tris)
Date: Wed, 14 Nov 2007 17:01:28 +0000
Subject: [thelist] SQL - hardcore... at least it looks that way to me,...
Message-ID: <8b20a7490711140901i6f664a24i34afc42c6310cb@mail.gmail.com>

A mate came to me with a simple SQL query...
what she's asked, I can do in PHP'Mysql with my eyes closed, but she's
given me code I don't recognse, and therefore, how to answer her..
Does the following message mean anything to anyone?

She basically need so to adda  studnet attendance, based on a name and
a date. and the SQl ias supposed to go find the student ID and
update...


-======================


SQL> describe student;
Name Null? Type
----------------------------------------- -------- --------------
STUDENT_ID NOT NULL NUMBER
FIRST_NAME VARCHAR2(15)
LAST_NAME VARCHAR2(20)
SEX CHAR(1)
TITLE VARCHAR2(6)

SQL> DESCRIBE Attendance;
Name Null? Type
----------------------------------------- -------- ----------
STUDENT_ID NOT NULL NUMBER
ATT_DATE DATE

Those are my two tables.

I have written one procedure that changes the title when it gets the
name passed:
CREATE OR REPLACE PROCEDURE addtitle(p_fn varchar2, p_ln varchar2)
IS
v_sex varchar2(1);
v_sid number;

BEGIN
SELECT student_id, sex INTO v_sid, v_sex
FROM student s
WHERE first_name = p_fn AND last_name = p_ln;

IF v_sex = 'M' THEN
UPDATE student SET title = 'Mr' WHERE student_id = v_sid;
ELSIF v_sex = 'F' THEN
UPDATE student SET title = 'Ms' WHERE student_id = v_sid;
END IF;
COMMIT;
END;
/

EXEC addtitle ('Tom', 'Smith')
EXEC addtitle ('Sweta', 'Patel')

And now I was asked to: Write a procedure to enter records in teh
Attendance table when the student's name and date of attendance are
passed to the procedure. And i have this, which doesn't work (not
surprised!)

CREATE OR REPLACE PROCEDURE addAttendance(p_fn varchar2, p_ln
varchar2, p_ad date)
IS
v_att_date date;
v_sid number;

BEGIN
SELECT student_id, att_date INTO v_sid, v_att_date
FROM Attendance a, Student s
WHERE first_name = p_fn AND last_name = p_ln;

IF v_att_date IS NULL THEN
UPDATE attendance SET Att_date = att_date WHERE student_id = v_sid;
ELSE
UPDATE attendance SET Att_date = Att_date WHERE student_id = v_sid;
END IF;
COMMIT;
END;
/


Thanks for looking at this. I just tried to change the first one round
but need to somehow join them to get at the student id and I can't do
that at the best of times without procedures



LINE/COL ERROR
-------- ----------------------------------------------------
7/1 PL/SQL: SQL Statement ignored
7/8 PL/SQL: ORA-00918: column ambiguously defined



=======================

-- 
Give a man a fish and he'll feed himself for a day.
Give a man a religion and he'll starve to death praying for a fish.
Anon

`We are what we pretend to be, so we must be careful what we pretend to be.`
Kurt Vonnegut

`When a person can no longer laugh at himself, it is time for others
to laugh at him.`
Thomas Szasz

From Ron.Luther at hp.com  Wed Nov 14 11:35:15 2007
From: Ron.Luther at hp.com (Luther, Ron)
Date: Wed, 14 Nov 2007 17:35:15 +0000
Subject: [thelist] (no subject)
Message-ID: 

Tris asked about some code:


Hi Tris,


At a quick glance it looks like this error:

>>7/8 PL/SQL: ORA-00918: column ambiguously defined

Is being caused in the select statement, here:

>>SELECT student_id, att_date INTO v_sid, v_att_date FROM Attendance a, Student s

Because Oracle isn't sure if you want a.student_id or s.student_id


I didn't look much beyond that, so correcting that may lead to more interesting errors!  ;-)


HTH,
RonL.

From Ron.Luther at hp.com  Wed Nov 14 11:38:40 2007
From: Ron.Luther at hp.com (Luther, Ron)
Date: Wed, 14 Nov 2007 17:38:40 +0000
Subject: [thelist] SQL - hardcore... at least it looks that way to me,
	...
In-Reply-To: <8b20a7490711140901i6f664a24i34afc42c6310cb@mail.gmail.com>
References: <8b20a7490711140901i6f664a24i34afc42c6310cb@mail.gmail.com>
Message-ID: 

Tris asked about some code:


Hi Tris,


At a quick glance it looks like this error:

>>7/8 PL/SQL: ORA-00918: column ambiguously defined

Is being caused in the select statement, here:

>>SELECT student_id, att_date INTO v_sid, v_att_date FROM Attendance a,
>>Student s

Because Oracle isn't sure if you want a.student_id or s.student_id


I didn't look much beyond that, so correcting that may lead to more interesting errors!  ;-)


HTH,
RonL.

(Sorry - Not sure why my first reply went out with no subject line - hamstahs in the servah again I reckon.)


From danielk at umd.edu  Wed Nov 14 12:10:51 2007
From: danielk at umd.edu (Daniel Kessler)
Date: Wed, 14 Nov 2007 13:10:51 -0500
Subject: [thelist] frameset styles
Message-ID: <7A868D55-AFCF-43BE-9838-B18DD3A88AB5@umd.edu>

> On 13 Nov 2007, at 10:12, David Dorward wrote:
>
>
> As per the other thread, it feels like I'm in a time capsule...
> whyever would one use frames these days?[1]


Well because when that part of the site was coded years ago, frames  
were the thing to use.  As it's an internal set of components that  
only need to be readable to be used, there's no reason to spend  
precious time rewriting it to something "correct" when it otherwise  
works fine and there's real work to do.

-- 

Daniel Kessler

University of Maryland College Park
School of Public Health
3302E HHP Building
College Park, MD  20742-2611
Phone: 301-405-2545
http://sph.umd.edu




From Paul.Bennett at wcc.govt.nz  Wed Nov 14 13:00:58 2007
From: Paul.Bennett at wcc.govt.nz (Paul Bennett)
Date: Thu, 15 Nov 2007 08:00:58 +1300
Subject: [thelist] Ruby on Rails: why?
In-Reply-To: <20E8922C-F68B-4022-B3E6-90E6528FE235@mountain.ch>
References: <20E8922C-F68B-4022-B3E6-90E6528FE235@mountain.ch>
Message-ID: 

Wow, bound to get a few replies here! 

This is a long reply, so bear with me here.

First of all, Ruby on Rails isn't a GUI product like Frontpage or GoLive, so I think you've been misled by the videos there.

RoR is a *framework* for producing dynamic web apps - basically if you send any amount of time building browser-based apps you'll soon get tired of solving the same problems and writing the same code over and over (and over) again.

RoR and other MVC frameworks (Zend, CakePHP, codeIgniter to name only some popular PHP alternatives) offer the separation (or near enough) of  data, business logic and presentation code, to allow you to get stuck into solving business problems first.

They take away much of the repetitive grunt work from building web apps.

Poor code? Bear in mind that most popular MVC frameworks are also *very* popular open-source projects with many intelligent developers not only extending but also securing and optimising the code, as well as identifying and fixing bugs. Also some pretty high profile, high traffic and high-load applications run on MVC frameworks (i.e.: all of 37Signals stuff) so they've passed the production environment test.

Personally, I've used cakePHP on several projects and am completely enamoured with it - it not only makes my development easier and faster, I get to focus more on problem solving and delivering something quickly and effectively than writing yet another db interface or presentation class.

Definitely, definitely take a look at using MVC frameworks if you spend any time building data-driven applications. The investment in time getting familiar with the MVC way of 'doing things' will be paid off many times over.

My 3c :)

Paul
http://thingsilearn.wordpress.com

-----Original Message-----
From: thelist-bounces at lists.evolt.org [mailto:thelist-bounces at lists.evolt.org] On Behalf Of Mark Howells-Mead
Sent: Thursday, November 15, 2007 3:26 AM
To: Evolt List
Subject: [thelist] Ruby on Rails: why?

Hi all

Just to be a bit controversial and perhaps a little OT.

Is Ruby on Rails just a trendy, client-server version of programs  
like GoLive and Frontpage? I see an awful lot of job adverts these  
days demanding experience with RoR, but from the introductory videos  
on the official website, it doesn't seem to be much more than a  
GoLive-type environment, albeit with integral Server-Side features.  
Am I missing something? It seems like a tool to enable developers to  
produce websites with dynamic content using less than optimal code  
and standardized functions, instead of had coding, which is in the  
most cases far superior. I am infinitely quicker coding by hand and  
RoR seems to add steps to the development process, instead of taking  
steps away....

Just $0.02, no offence intended to users and RoR developers!

Cheers
Mark Howells-Mead
- www.permanenttourist.ch
- www.flickr.com/photos/mhowells/


-- 

* * Please support the community that supports you.  * *
http://evolt.org/help_support_evolt/

For unsubscribe and other options, including the Tip Harvester 
and archives of thelist go to: http://lists.evolt.org 
Workers of the Web, evolt ! 

From joel at streamliine.com  Wed Nov 14 13:09:38 2007
From: joel at streamliine.com (Joel D Canfield)
Date: Wed, 14 Nov 2007 11:09:38 -0800
Subject: [thelist] Ruby on Rails: why?
References: <20E8922C-F68B-4022-B3E6-90E6528FE235@mountain.ch>
	
Message-ID: <72E9FAA171D63B48AAC707C72900E6B4D89F6A@ireland.spinhead.com>

> Definitely, definitely take a look at using MVC frameworks if 
> you spend any time building data-driven applications. The 
> investment in time getting familiar with the MVC way of 
> 'doing things' will be paid off many times over.
> 
> My 3c :)

Wow. Paul, you get my 'helpful post of the day' award.

*If* I decide to continue with any kind of coding instead of tossing it
over for business analysis/consultancy/public speaking, I'll spend the
time to love an MVC framework. Your post whacks so much of what I'm
bugged about when I'm coding.

joel

From fredthejonester at gmail.com  Wed Nov 14 12:17:31 2007
From: fredthejonester at gmail.com (Fred Jones)
Date: Wed, 14 Nov 2007 21:17:31 +0300
Subject: [thelist] Ruby on Rails: why?
In-Reply-To: <72E9FAA171D63B48AAC707C72900E6B4D89F6A@ireland.spinhead.com>
References: <20E8922C-F68B-4022-B3E6-90E6528FE235@mountain.ch>	
	<72E9FAA171D63B48AAC707C72900E6B4D89F6A@ireland.spinhead.com>
Message-ID: <473B3BBB.5010405@gmail.com>

Joel D Canfield wrote:
>> Definitely, definitely take a look at using MVC frameworks if 
>> you spend any time building data-driven applications. The 
>> investment in time getting familiar with the MVC way of 
>> 'doing things' will be paid off many times over.
>>
>> My 3c :)
> 
> Wow. Paul, you get my 'helpful post of the day' award.
> 
> *If* I decide to continue with any kind of coding instead of tossing it
> over for business analysis/consultancy/public speaking, I'll spend the
> time to love an MVC framework. Your post whacks so much of what I'm
> bugged about when I'm coding.

There is no question about this. If I was to code from scratch a web 
app, I wouldn't even consider anything but CakePHP.

MVC is to programmers what GUI is to laymen.

lol

Fred

From santilal at scorpioneng.co.nz  Wed Nov 14 14:18:55 2007
From: santilal at scorpioneng.co.nz (Santilal Parbhu)
Date: Thu, 15 Nov 2007 09:18:55 +1300
Subject: [thelist] Mysql connection failed
Message-ID: 

Hi

 

This is a very unusual query.  I have been building a PHP & Mysql
application and testing it on a Abria Merlin platform using PHP 4.0.4 and
Mysql 3.23.22-beta with an Apache 1.3 server.  The development has been
going fine and I have been using it successfully for about 18 months.  After
some minor tweaking (mostly HTML tidy-ups and improving security) I have
found that on some occasions I can't connect to the database.  On other
occasions, I can, even though there have been no changes in-between.
Weird!!

 

This is the error I get: (I am printing out the host name, username and
password for debugging purposes).

 

"Warning: MySQL Connection Failed: Access denied for user
'admin'@'localhost' (using password: YES) in mysqlconnect.php on line 12
localhost, admin, access

 

 

The script is:

 

Could not select the database because: '
. mysql_error() . 
');

            }

            

            //mysql_close(); // Close the connection.

 

} else {

            print "$hostName, $dbUserName, $dbPassword";

            die ('
Could not connect to MySQL because: ' .
mysql_error() . '
');

 

}

 

?>

 

The script that holds the parameters is:

 



 

I have even tried accessing the database directly using MysqlMan, but this
failed too.  It is as though the database is not recognizing the username
and password, but I have successfully used these before.

 

Does anyone know a way to get into the database through a back-door so I can
see what is happening?

 

Thanks.  Oh and I an using Windows XP Pro.

 

 

 

Santilal Parbhu

 


From Paul.Bennett at wcc.govt.nz  Wed Nov 14 15:06:15 2007
From: Paul.Bennett at wcc.govt.nz (Paul Bennett)
Date: Thu, 15 Nov 2007 10:06:15 +1300
Subject: [thelist] Mysql connection failed
In-Reply-To: 
References: 
Message-ID: 

Hi Santilal,

You mention 'improving security' - any chance you've tightened up database security? It looks as if your database access is failing due to an incorrect username or password, or the user account doesn't have access to the database.

:)
Paul

From Chris at activeide.com  Wed Nov 14 15:24:02 2007
From: Chris at activeide.com (Chris Anderson)
Date: Wed, 14 Nov 2007 21:24:02 -0000
Subject: [thelist] Mysql connection failed
References: 
	<50202E8619084085AA65DE6D30ADA916@ActiveIDE.local>
Message-ID: <09DEC79B6F995248B402303C86CA4912237AAF@scarlet.ActiveIDE.local>

> You mention 'improving security' - any chance you've tightened up
> database security? It looks as if your database access is failing due
> to an incorrect username or password, or the user account doesn't have
> access to the database.

+1 - but also check the host because of the MySql way of treating
admin at localhost and admin at devmachine as separate accounts.

If it's intermittent, it *may* even be the localhost thing - perhaps
MySql is sometimes seeing the host as localhost and sometimes as
servername, so check you have 'admin'@'localhost' *and*
'admin'@'servername' (or just define 'admin'@'%' to cover all hosts if
you are not worried about third party access)

Chris

From BreezyGraphics at aol.com  Wed Nov 14 16:24:20 2007
From: BreezyGraphics at aol.com (Breezy)
Date: Wed, 14 Nov 2007 17:24:20 -0500
Subject: [thelist] Link problems on html on desktop
Message-ID: <000a01c8270d$1d71bb30$0501a8c0@pc00012345678>

Ok, here's the story.

I was showing the kid how to make an html page in notepad so that he could
make links to swf games that he could right click and save target as so he
could play them without being on the internet. Example worked on my computer
just as it should. Then we moved to his laptop which has IE and Firefox.
Well, Firefox didn't have save target as AND would not open the link. Opened
the page from IE and it didn't even show up as a link. Has anyone ever run
across this before or better yet know a solution? 

Thanks, Breezy.

From evolt_org at striderweb.com  Wed Nov 14 17:16:55 2007
From: evolt_org at striderweb.com (Stephen Rider)
Date: Wed, 14 Nov 2007 17:16:55 -0600
Subject: [thelist] Domain masking and $_SERVER['HTTP_HOST']
Message-ID: <381C9C56-6F9C-4EC5-A788-4133DCA36EAF@striderweb.com>

I'm working on a multiblog system that allows you to run multiple  
blogs of a single install of the blog platform.  (Currently it's only  
for WordPress, but should be modifiable to work with others).

The basic concept is that you install WordPress -- that's your first  
blog.  Then you make symbolic links to that directory, and those are  
you other blogs.  My system calls a different set of configuration  
files depending on what "directory" it thinks it's being called from.

In my attempts to make it _really_ easy to configure, I have code in  
it that auto-detects which directory it's in, or if it's the root  
directory, what domain it is.  It then looks for (or creates) tables  
in the database based on the directory or domain.  Different set of  
tables == different blog.

The problem:

I see a potential security problem here.  What would stop a person  
from pointing their own domain at my site and thus auto-creating  
their own blog?  Without access to my hosting, they can't directly  
point a domain to my directories, but what about domain masking?

I used domain masking to point a different (unhosted) domain to my  
site, in an attempt to test it.  My regular site (and thus "proper"  
blog) showed, up.  I was somewhat expecting it to allow me to install  
an auto-created blog based on the masked domain.

The Question:

Why _didn't_ the domain masking work?  The auto-config works based on  
looking at $_SERVER['HTTP_HOST'].  When using masking does this  
variable see the "real" domain and not the masked one?

Are there other security problems I should be looking for?

Regards,
Stephen Rider

From morrison.ben at gmail.com  Thu Nov 15 03:40:19 2007
From: morrison.ben at gmail.com (ben morrison)
Date: Thu, 15 Nov 2007 09:40:19 +0000
Subject: [thelist] Link problems on html on desktop
In-Reply-To: <000a01c8270d$1d71bb30$0501a8c0@pc00012345678>
References: <000a01c8270d$1d71bb30$0501a8c0@pc00012345678>
Message-ID: <6073aef90711150140h3c0cdfc2xd53c425d13bfd483@mail.gmail.com>

On Nov 14, 2007 10:24 PM, Breezy  wrote:
> Ok, here's the story.
>
> I was showing the kid how to make an html page in notepad so that he could
> make links to swf games that he could right click and save target as so he
> could play them without being on the internet. Example worked on my computer
> just as it should. Then we moved to his laptop which has IE and Firefox.
> Well, Firefox didn't have save target as AND would not open the link. Opened
> the page from IE and it didn't even show up as a link. Has anyone ever run
> across this before or better yet know a solution?

Without seeing the HTML code it is difficult for us to help you . . . .

Could you put the page online?

ben
-- 
Ben Morrison
http://www.benjaminmorrison.com

From webdev at mountain.ch  Thu Nov 15 06:27:21 2007
From: webdev at mountain.ch (Mark Howells-Mead)
Date: Thu, 15 Nov 2007 13:27:21 +0100
Subject: [thelist] Ruby on Rails: why?
In-Reply-To: 
References: <20E8922C-F68B-4022-B3E6-90E6528FE235@mountain.ch>
	
Message-ID: <1FBD8D01-0E5A-4644-8276-5D23BBAF7475@mountain.ch>

Hi Paul, thanks for the detailed reply!

> RoR is a *framework* for producing dynamic web apps - basically if  
> you send any amount of time building browser-based apps you'll soon  
> get tired of solving the same problems and writing the same code  
> over and over (and over) again.

So if one has a central batch of function (method) libraries, which  
can be appended to projects as required, then that amounts to the  
same thing? RoR is simply a framework of libraries, which developers  
can use to create apps and dynamically-driven website?

> RoR and other MVC frameworks (Zend, CakePHP, codeIgniter to name  
> only some popular PHP alternatives) offer the separation (or near  
> enough) of  data, business logic and presentation code, to allow  
> you to get stuck into solving business problems first.
>
> They take away much of the repetitive grunt work from building web  
> apps.

That certainly makes sense, though if I've been using the concept of  
separation between content, layout and functionality for nearly ten  
years, building new globally useful library methods every week, then  
does that amount to the same thing?

> Poor code? Bear in mind that most popular MVC frameworks are also  
> *very* popular open-source projects with many intelligent  
> developers not only extending but also securing and optimising the  
> code, as well as identifying and fixing bugs.

OK, I retract what I said; what I actually meant was poor code in  
terms of code which isn't *quite* right for our application, so that  
it always has to be modified and often has to be improved or extended.

> Also some pretty high profile, high traffic and high-load  
> applications run on MVC frameworks (i.e.: all of 37Signals stuff)  
> so they've passed the production environment test.

Absolutely, but they've only passed the production environment test  
of that application environment. For example, none of the high  
profile shareware/freeware CMS systems, like WordPress for example,  
don't properly support multilingualism. There are of course plug-ins,  
but the majority of them are likely to be way less than they could  
be, because they're based on WordPress itself.

> Definitely, definitely take a look at using MVC frameworks if you  
> spend any time building data-driven applications. The investment in  
> time getting familiar with the MVC way of 'doing things' will be  
> paid off many times over.

Do you have any links you can share? It sounds as though I've been  
doing this for many years, and that (for example) RoR is simply the  
most recent and popular framework.

Mark Howells-Mead
- www.permanenttourist.ch
- www.flickr.com/photos/mhowells/

From fredthejonester at gmail.com  Thu Nov 15 05:42:23 2007
From: fredthejonester at gmail.com (Fred Jones)
Date: Thu, 15 Nov 2007 14:42:23 +0300
Subject: [thelist] Ruby on Rails: why?
In-Reply-To: <1FBD8D01-0E5A-4644-8276-5D23BBAF7475@mountain.ch>
References: <20E8922C-F68B-4022-B3E6-90E6528FE235@mountain.ch>	
	<1FBD8D01-0E5A-4644-8276-5D23BBAF7475@mountain.ch>
Message-ID: <473C309F.5020609@gmail.com>

> So if one has a central batch of function (method) libraries, which  
> can be appended to projects as required, then that amounts to the  
> same thing? RoR is simply a framework of libraries, which developers  
> can use to create apps and dynamically-driven website?

No, RoR (and CakePHP) are MVC, which is something more than just a 
library of functions.

> That certainly makes sense, though if I've been using the concept of  
> separation between content, layout and functionality for nearly ten  
> years, building new globally useful library methods every week, then  
> does that amount to the same thing?

MVC is indeed a method of separating data, logic and presentation. So 
could be it is the same thing. A major advantage of RoR or CakePHP is 
that there is a team of developers contributing to and constantly 
improving the framework.

> OK, I retract what I said; what I actually meant was poor code in  
> terms of code which isn't *quite* right for our application, so that  
> it always has to be modified and often has to be improved or extended.

Without specific examples that's quite a sweeping statement. :)

>> Definitely, definitely take a look at using MVC frameworks if you  
>> spend any time building data-driven applications. The investment in  
>> time getting familiar with the MVC way of 'doing things' will be  
>> paid off many times over.
> 
> Do you have any links you can share? It sounds as though I've been  
> doing this for many years, and that (for example) RoR is simply the  
> most recent and popular framework.

I myself don't know nor use RoR, but you can see these:

http://en.wikipedia.org/wiki/Model-view-controller
http://cakephp.org/

:)

Fred

From webdev at mountain.ch  Thu Nov 15 06:53:47 2007
From: webdev at mountain.ch (Mark Howells-Mead)
Date: Thu, 15 Nov 2007 13:53:47 +0100
Subject: [thelist] Ruby on Rails: why?
In-Reply-To: <473C309F.5020609@gmail.com>
References: <20E8922C-F68B-4022-B3E6-90E6528FE235@mountain.ch>	
	<1FBD8D01-0E5A-4644-8276-5D23BBAF7475@mountain.ch>
	<473C309F.5020609@gmail.com>
Message-ID: <6823EA0E-8CE7-4E77-8317-0E9D16F77B56@mountain.ch>

>> Do you have any links you can share? It sounds as though I've been
>> doing this for many years, and that (for example) RoR is simply the
>> most recent and popular framework.
>
> I myself don't know nor use RoR, but you can see these:
>
> http://en.wikipedia.org/wiki/Model-view-controller
> http://cakephp.org/

Thanks for the links, Fred. Though I know that CakePHP is about PHP,  
it puts me off using a development tool, when I go to the project  
website and within two clicks start finding programming errors or  
oversights! (Namely, Javascript errors and an comments section full  
of spam, which is fairly easily blocked these days. I know this  
doesn't have any direct association with the "product" but it does  
give a poor first impression of the developers who have created it,  
which would make me question the stability of their project!)

Cheers
Mark Howells-Mead
- www.permanenttourist.ch
- www.flickr.com/photos/mhowells/


From beertastic at gmail.com  Thu Nov 15 08:25:24 2007
From: beertastic at gmail.com (Tris)
Date: Thu, 15 Nov 2007 14:25:24 +0000
Subject: [thelist] SQL multiple Joins?
Message-ID: <8b20a7490711150625g7622c1a5kc33be11c529c3b7b@mail.gmail.com>

Is it possible to have multiple joins..

Ont eh query below. it failes.. Buuuut, if I limit it to one join, it
works great..
My database logs transactions and I'm trying to produce a report that
shows each transaction total, within a certain period.. (Oct and sept
so far..)

but it's cfalling over..
Anyhoo, off to google... race you!

============

SELECT
	 SUM(activitiesSep.points) as sepPoints
	,SUM(activitiesOct.points) as octPoints
	,users.ID
	,users.Username
	,users.Password
	,users.Name
	,users.Company
	,users.Address1
	,users.Address2
	,users.Address3
	,users.Address4
	,users.Address5
	,users.Postcode
	,users.Email
	,users.CustomerType
	,users.Active
	,users.LogIns
	,users.GAURef
	,users.JRRef
	,users.GAUMember
	,users.GAUContact
	,users.RegForm
	,users.Created
	,users.Updated

FROM `users`

RIGHT JOIN `activities` as `activitiesSep` on
	activitiesSep.UserID = users.ID
WHERE activitiesSep.`Timestamp` >= '1188604800' AND
activitiesSep.`Timestamp` <= '1191196799'

RIGHT JOIN `activities` as `activitiesOct` on
	activitiesOct.UserID = users.ID
WHERE activitiesOct.`Timestamp` >= '1191196800' AND
activitiesOct.`Timestamp` <= '1193875199'

GROUP BY users.ID

================

-- 
Give a man a fish and he'll feed himself for a day.
Give a man a religion and he'll starve to death praying for a fish.
Anon

`We are what we pretend to be, so we must be careful what we pretend to be.`
Kurt Vonnegut

`When a person can no longer laugh at himself, it is time for others
to laugh at him.`
Thomas Szasz

From fredthejonester at gmail.com  Thu Nov 15 07:33:30 2007
From: fredthejonester at gmail.com (Fred Jones)
Date: Thu, 15 Nov 2007 16:33:30 +0300
Subject: [thelist] Ruby on Rails: why?
In-Reply-To: <6823EA0E-8CE7-4E77-8317-0E9D16F77B56@mountain.ch>
References: <20E8922C-F68B-4022-B3E6-90E6528FE235@mountain.ch>		<1FBD8D01-0E5A-4644-8276-5D23BBAF7475@mountain.ch>	<473C309F.5020609@gmail.com>
	<6823EA0E-8CE7-4E77-8317-0E9D16F77B56@mountain.ch>
Message-ID: <473C4AAA.1000608@gmail.com>

>> http://cakephp.org/
> 
> Thanks for the links, Fred. Though I know that CakePHP is about PHP,  
> it puts me off using a development tool, when I go to the project  
> website and within two clicks start finding programming errors or  
> oversights! (Namely, Javascript errors and an comments section full  
> of spam, which is fairly easily blocked these days. I know this  
> doesn't have any direct association with the "product" but it does  
> give a poor first impression of the developers who have created it,  
> which would make me question the stability of their project!)

What URL has errors and what URL has spam comments?

Thanks.

From jason.handby at corestar.co.uk  Thu Nov 15 08:37:47 2007
From: jason.handby at corestar.co.uk (Jason Handby)
Date: Thu, 15 Nov 2007 14:37:47 -0000
Subject: [thelist] SQL multiple Joins?
References: <8b20a7490711150625g7622c1a5kc33be11c529c3b7b@mail.gmail.com>
Message-ID: <9A50776858A21848A96469CDFCBCDEFF012042CD@exch-be12.exchange.local>

Hi Tris,


> Is it possible to have multiple joins..

The short answer is "yes".

However, the slightly longer answer is, as with anything, that you will
only get the results you expect if you use them right :-)


> Ont eh query below. it failes.. Buuuut, if I limit it to one join, it
> works great..
> My database logs transactions and I'm trying to produce a report that
> shows each transaction total, within a certain period.. (Oct and sept
> so far..)
> 
> but it's cfalling over..


> 
> SELECT
> 	 SUM(activitiesSep.points) as sepPoints
> 	,SUM(activitiesOct.points) as octPoints
> 	,users.ID
> 	,users.Username
> 	,users.Password
> 	,users.Name
> 	,users.Company
> 	,users.Address1
> 	,users.Address2
> 	,users.Address3
> 	,users.Address4
> 	,users.Address5
> 	,users.Postcode
> 	,users.Email
> 	,users.CustomerType
> 	,users.Active
> 	,users.LogIns
> 	,users.GAURef
> 	,users.JRRef
> 	,users.GAUMember
> 	,users.GAUContact
> 	,users.RegForm
> 	,users.Created
> 	,users.Updated
> 
> FROM `users`
> 
> RIGHT JOIN `activities` as `activitiesSep` on
> 	activitiesSep.UserID = users.ID
> WHERE activitiesSep.`Timestamp` >= '1188604800' AND
> activitiesSep.`Timestamp` <= '1191196799'
> 
> RIGHT JOIN `activities` as `activitiesOct` on
> 	activitiesOct.UserID = users.ID
> WHERE activitiesOct.`Timestamp` >= '1191196800' AND
> activitiesOct.`Timestamp` <= '1193875199'
> 
> GROUP BY users.ID


OK, so one pretty nice way to do this would be to use subqueries,
perhaps like this:


SELECT
	 (SELECT SUM(points) FROM activities
		WHERE activities.`Timestamp`  >= '1188604800'
		AND activities.`Timestamp` <= '1191196799'
		AND activities.UserID = users.ID
	) AS sepPoints,
	 (SELECT SUM(points) FROM activities
		WHERE activities.`Timestamp` >= '1191196800'
		AND activities.`Timestamp` <= '1193875199'
		AND activities.UserID = users.ID
	) AS octPoints,
	,users.ID
	,users.Username
	,users.Password
	,users.Name
	,users.Company
	,users.Address1
	,users.Address2
	,users.Address3
	,users.Address4
	,users.Address5
	,users.Postcode
	,users.Email
	,users.CustomerType
	,users.Active
	,users.LogIns
	,users.GAURef
	,users.JRRef
	,users.GAUMember
	,users.GAUContact
	,users.RegForm
	,users.Created
	,users.Updated


I've just tested something like this in MS-SQL and it works fine, so
hopefully it should give you something to go on.




Jason


From webdev at mountain.ch  Thu Nov 15 08:43:31 2007
From: webdev at mountain.ch (Mark Howells-Mead)
Date: Thu, 15 Nov 2007 15:43:31 +0100
Subject: [thelist] Ruby on Rails: why?
In-Reply-To: <473C4AAA.1000608@gmail.com>
References: <20E8922C-F68B-4022-B3E6-90E6528FE235@mountain.ch>		<1FBD8D01-0E5A-4644-8276-5D23BBAF7475@mountain.ch>	<473C309F.5020609@gmail.com>
	<6823EA0E-8CE7-4E77-8317-0E9D16F77B56@mountain.ch>
	<473C4AAA.1000608@gmail.com>
Message-ID: <813E367E-E2C5-4720-BD9A-A06B322FBCD4@mountain.ch>

>>> http://cakephp.org/
>>
> What URL has errors and what URL has spam comments?

http://cakephp.org/maps : no map and a Javascript error
http://cakephp.org/maps/markers: most of the recent entries are spam

Mark Howells-Mead
- www.permanenttourist.ch
- www.flickr.com/photos/mhowells/


From evolt_org at striderweb.com  Thu Nov 15 08:59:35 2007
From: evolt_org at striderweb.com (Stephen Rider)
Date: Thu, 15 Nov 2007 08:59:35 -0600
Subject: [thelist] Ruby on Rails: why?
In-Reply-To: <1FBD8D01-0E5A-4644-8276-5D23BBAF7475@mountain.ch>
References: <20E8922C-F68B-4022-B3E6-90E6528FE235@mountain.ch>
	
	<1FBD8D01-0E5A-4644-8276-5D23BBAF7475@mountain.ch>
Message-ID: 

Your phrasing is confusing -- don't know if you intended the double- 
negative, BUT just for the record...

WordPress is most definitely multi-lingual.


Plugins are frequently multilingual as well, though of course that's  
limited to how many multilingual people  a single developer can get  
to make .mo files for his plugin. :)

Stephen

On Nov 15, 2007, at 6:27 AM, Mark Howells-Mead wrote:

> For example, none of the high
> profile shareware/freeware CMS systems, like WordPress for example,
> don't properly support multilingualism. There are of course plug-ins,
> but the majority of them are likely to be way less than they could
> be, because they're based on WordPress itself.


From webdev at mountain.ch  Thu Nov 15 09:04:26 2007
From: webdev at mountain.ch (Mark Howells-Mead)
Date: Thu, 15 Nov 2007 16:04:26 +0100
Subject: [thelist] Ruby on Rails: why?
In-Reply-To: 
References: <20E8922C-F68B-4022-B3E6-90E6528FE235@mountain.ch>
	
	<1FBD8D01-0E5A-4644-8276-5D23BBAF7475@mountain.ch>
	
Message-ID: 

> Your phrasing is confusing -- don't know if you intended the double-
> negative, BUT just for the record...
>
> WordPress is most definitely multi-lingual.
> 

Sorry, I naturally meant "none of the high profile shareware/freeware  
CMS systems, like WordPress for example, properly support  
multilingualism".

That's to say, I can't use the default installation of WordPress to  
power a site where the visitor can switch between (say) English,  
French, German and Italian content at will. That requires an input  
environment which is capable of handling content in several languages  
concurrently, which WordPress - and, as far as I am aware, all other  
popular freeware CMS systems - cannot. (Not without a third party add- 
on, anyway.)


Mark Howells-Mead
- www.permanenttourist.ch
- www.flickr.com/photos/mhowells/

From struan at exo.org.uk  Thu Nov 15 09:02:35 2007
From: struan at exo.org.uk (Struan Donald)
Date: Thu, 15 Nov 2007 15:02:35 +0000
Subject: [thelist] SQL multiple Joins?
In-Reply-To: <8b20a7490711150625g7622c1a5kc33be11c529c3b7b@mail.gmail.com>
References: <8b20a7490711150625g7622c1a5kc33be11c529c3b7b@mail.gmail.com>
Message-ID: <20071115150233.GV10789@bollo.hoodee.co.uk>

* at 15/11 14:25 +0000 Tris said:
> Is it possible to have multiple joins..

Yes 



> FROM `users`
> 
> RIGHT JOIN `activities` as `activitiesSep` on
> 	activitiesSep.UserID = users.ID
> WHERE activitiesSep.`Timestamp` >= '1188604800' AND
> activitiesSep.`Timestamp` <= '1191196799'
> 
> RIGHT JOIN `activities` as `activitiesOct` on
> 	activitiesOct.UserID = users.ID
> WHERE activitiesOct.`Timestamp` >= '1191196800' AND
> activitiesOct.`Timestamp` <= '1193875199'
> 
> GROUP BY users.ID

Unless you're database is using a particularly mutant form of SQL you
probably want to do this instead:

> RIGHT JOIN `activities` as `activitiesSep` on
> 	activitiesSep.UserID = users.ID
> RIGHT JOIN `activities` as `activitiesOct` on
> 	activitiesOct.UserID = users.ID
> WHERE activitiesOct.`Timestamp` >= '1191196800' AND
> activitiesOct.`Timestamp` <= '1193875199' AND
> activitiesSep.`Timestamp` >= '1188604800' AND
> activitiesSep.`Timestamp` <= '1191196799'

Struan

From evolt_org at striderweb.com  Thu Nov 15 09:40:34 2007
From: evolt_org at striderweb.com (Stephen Rider)
Date: Thu, 15 Nov 2007 09:40:34 -0600
Subject: [thelist] Domain masking and $_SERVER['HTTP_HOST']
In-Reply-To: <381C9C56-6F9C-4EC5-A788-4133DCA36EAF@striderweb.com>
References: <381C9C56-6F9C-4EC5-A788-4133DCA36EAF@striderweb.com>
Message-ID: <7108A453-2371-42C4-B856-8253A53E33A1@striderweb.com>

Let's make the question a bit more short-winded. :)

I have a PHP application that does something different based on what  
domain it's in.  That is, I can point multiple domains to the same  
application, and the app looks up the domain and acts accordingly.

It's primary determining factor is $_SERVER['HTTP_HOST'].

How much of a security risk is this?  I don't want somebody to be  
able to use, for example, domain masking to point some totally  
different domain at my app and piggyback on it.

Yes, this could be locked down with a config file that limits it, but  
one of my goals is to make this virtually configuration-free for the  
user.

Would I be better off using $_SERVER['SERVER_NAME'] ?  Is this whole  
thing simply a bad idea?

Incidentally, I DID do a test run with a masked domain, and there was  
no problem, but I'm no expert on DNS, and one test doesn't prove  
much. :)

Thanks,
Stephen


If you want to use the SUP tag in a way that looks the same cross  
browser, AND doesn't mess up line heights, try putting this in your  
CSS file:

sup {
	position: relative;
	top: -3px;
	vertical-align: top;
	font-size: 90%;
}





On Nov 14, 2007, at 5:16 PM, Stephen Rider wrote:

> I'm working on a multiblog system that allows you to run multiple
> blogs of a single install of the blog platform.  (Currently it's only
> for WordPress, but should be modifiable to work with others).
>
> The basic concept is that you install WordPress -- that's your first
> blog.  Then you make symbolic links to that directory, and those are
> you other blogs.  My system calls a different set of configuration
> files depending on what "directory" it thinks it's being called from.
>
> In my attempts to make it _really_ easy to configure, I have code in
> it that auto-detects which directory it's in, or if it's the root
> directory, what domain it is.  It then looks for (or creates) tables
> in the database based on the directory or domain.  Different set of
> tables == different blog.
>
> The problem:
>
> I see a potential security problem here.  What would stop a person
> from pointing their own domain at my site and thus auto-creating
> their own blog?  Without access to my hosting, they can't directly
> point a domain to my directories, but what about domain masking?
>
> I used domain masking to point a different (unhosted) domain to my
> site, in an attempt to test it.  My regular site (and thus "proper"
> blog) showed, up.  I was somewhat expecting it to allow me to install
> an auto-created blog based on the masked domain.
>
> The Question:
>
> Why _didn't_ the domain masking work?  The auto-config works based on
> looking at $_SERVER['HTTP_HOST'].  When using masking does this
> variable see the "real" domain and not the masked one?
>
> Are there other security problems I should be looking for?
>
> Regards,
> Stephen Rider
>

From dan at virtuawebtech.co.uk  Thu Nov 15 11:02:42 2007
From: dan at virtuawebtech.co.uk (Dan Parry)
Date: Thu, 15 Nov 2007 17:02:42 -0000
Subject: [thelist] Floaty CSS
Message-ID: <061601c827a9$567bf660$0373e320$@co.uk>

Hi all

I'm having an issue with CSS that's probably really easy and elementary to
fix, but CSS isn't friendly to me... I have a holder (div) that contains 2
floated divs as per this diagram:

+--------------------------------+
|+-------+ +--------------------+|
||       | |                    ||
||   1   | |          2         ||
||       | |                    ||
|+-------+ +--------------------+|
+--------------------------------+

Now, how do I make sure that the divs (1 and 2) are always the same height?
They contain dynamic data and have their own background colours... If 2 is
larger than 1 then 1 finishes early (can't find a better way of saying that)
allowing 2 to expand vertically, eventually wrapping underneath 1...
diagram:

+--------------------------------+
|+-------+ +--------------------+|
||       | |                    ||
||   1   | |          2         ||
||       | |                    ||
|+-------+ |                    ||
|+---------+                    ||
||                              ||
|+------------------------------+|
+--------------------------------+

So how do I get the height of 1 to mirror that of 2? Conversely, 1 may be
longer than 2 so also needs the same constraints

I hope that makes sense

TIA!

Dan

-- 
Dan Parry

No virus found in this outgoing message.
Checked by AVG Free Edition. 
Version: 7.5.503 / Virus Database: 269.15.32/1131 - Release Date: 14/11/2007
16:54
 


From morrison.ben at gmail.com  Thu Nov 15 11:17:39 2007
From: morrison.ben at gmail.com (ben morrison)
Date: Thu, 15 Nov 2007 17:17:39 +0000
Subject: [thelist] Floaty CSS
In-Reply-To: <061601c827a9$567bf660$0373e320$@co.uk>
References: <061601c827a9$567bf660$0373e320$@co.uk>
Message-ID: <6073aef90711150917j157f52f3q5d08bba2de590b28@mail.gmail.com>

On Nov 15, 2007 5:02 PM, Dan Parry  wrote:
> Hi all
>
> I'm having an issue with CSS that's probably really easy and elementary to
> fix, but CSS isn't friendly to me... I have a holder (div) that contains 2
> floated divs as per this diagram:
>
> +--------------------------------+
> |+-------+ +--------------------+|
> ||       | |                    ||
> ||   1   | |          2         ||
> ||       | |                    ||
> |+-------+ +--------------------+|
> +--------------------------------+
>
> Now, how do I make sure that the divs (1 and 2) are always the same height?
> They contain dynamic data and have their own background colours... If 2 is
> larger than 1 then 1 finishes early (can't find a better way of saying that)
> allowing 2 to expand vertically, eventually wrapping underneath 1...
> diagram:

There is no actual way of doing it in CSS but you can *pretend* , you
give them both a width, and to give the appearance of the same height,
you apply a repeating background to the parent div...

Its sometimes called faux columns

http://www.alistapart.com/articles/fauxcolumns/

ben
-- 
Ben Morrison
http://www.benjaminmorrison.com

From dan at virtuawebtech.co.uk  Thu Nov 15 11:32:25 2007
From: dan at virtuawebtech.co.uk (Dan Parry)
Date: Thu, 15 Nov 2007 17:32:25 -0000
Subject: [thelist] Floaty CSS
In-Reply-To: <6073aef90711150917j157f52f3q5d08bba2de590b28@mail.gmail.com>
References: <061601c827a9$567bf660$0373e320$@co.uk>
	<6073aef90711150917j157f52f3q5d08bba2de590b28@mail.gmail.com>
Message-ID: <063b01c827ad$7d02f370$7708da50$@co.uk>

Ben Morrison wrote
> There is no actual way of doing it in CSS but you can *pretend* , you
> give them both a width, and to give the appearance of the same height,
> you apply a repeating background to the parent div...
> 
> Its sometimes called faux columns

Oh that's sneaky! I like it

I really wanted to avoid tables more than anything (evil... spawn of
Santa... that's right isn't it?)

> http://www.alistapart.com/articles/fauxcolumns/
> 
> ben
> --

Thanks again, Ben... It isn't the first time you've saved my worthless hide
:)

Dan
-- 
Dan Parry
 

No virus found in this outgoing message.
Checked by AVG Free Edition. 
Version: 7.5.503 / Virus Database: 269.15.32/1131 - Release Date: 14/11/2007
16:54
 


From pandar at fast-panda.com  Thu Nov 15 12:30:27 2007
From: pandar at fast-panda.com (megan cauley)
Date: Thu, 15 Nov 2007 13:30:27 -0500 (EST)
Subject: [thelist] Floaty CSS
Message-ID: <18210.207.245.75.194.1195151427.squirrel@webmail.fast-panda.com>

Dan,

Not sure if this has been answered yet, I get the list digest...

I have used this JavaScript, I found somewhere,...
to keep 2 divs equal heights. I have also called this function multiple
times in one document.

**   

...

**   

Whatever your div id's are set to that is where ('left') and ('right')
would be....


hth,

megan

---------------------------------

Message: 22
Date: Thu, 15 Nov 2007 17:02:42 -0000
From: "Dan Parry" 
Subject: [thelist] Floaty CSS
To: 
Message-ID: <061601c827a9$567bf660$0373e320$@co.uk>
Content-Type: text/plain;        charset="windows-1250"

Hi all

I'm having an issue with CSS that's probably really easy and elementary to
fix, but CSS isn't friendly to me... I have a holder (div) that contains 2
floated divs as per this diagram:

+--------------------------------+
|+-------+ +--------------------+|
||       | |                    ||
||   1   | |          2         ||
||       | |                    ||
|+-------+ +--------------------+|
+--------------------------------+

Now, how do I make sure that the divs (1 and 2) are always the same height?
They contain dynamic data and have their own background colours... If 2 is
larger than 1 then 1 finishes early (can't find a better way of saying that)
allowing 2 to expand vertically, eventually wrapping underneath 1...
diagram:

+--------------------------------+
|+-------+ +--------------------+|
||       | |                    ||
||   1   | |          2         ||
||       | |                    ||
|+-------+ |                    ||
|+---------+                    ||
||                              ||
|+------------------------------+|
+--------------------------------+

So how do I get the height of 1 to mirror that of 2? Conversely, 1 may be
longer than 2 so also needs the same constraints

I hope that makes sense

TIA!

Dan

-- 
Dan Parry


From rudy at r937.com  Thu Nov 15 13:53:23 2007
From: rudy at r937.com (r937)
Date: Thu, 15 Nov 2007 14:53:23 -0500
Subject: [thelist]  SQL multiple Joins?
Message-ID: <020a01c827c1$3c7ee660$9a7ba8c0@curly>

tris, one thing you might do is learn the differences between the
different types of joins

"X right join Y" means you want all Ys, with or without any matching Xs

so when you have this --

  FROM `users`RIGHT JOIN `activities` as `activitiesSep`
      on activitiesSep.UserID = users.ID

this suggests that you want all activities with or without the matching user

this doesn't make sense

how can you have an activity for a user that doesn't exist?

most likely you want INNER JOIN


rudy



From Paul.Bennett at wcc.govt.nz  Thu Nov 15 14:20:10 2007
From: Paul.Bennett at wcc.govt.nz (Paul Bennett)
Date: Fri, 16 Nov 2007 09:20:10 +1300
Subject: [thelist] Ruby on Rails: why?
In-Reply-To: <1FBD8D01-0E5A-4644-8276-5D23BBAF7475@mountain.ch>
References: <20E8922C-F68B-4022-B3E6-90E6528FE235@mountain.ch>
	<1FBD8D01-0E5A-4644-8276-5D23BBAF7475@mountain.ch>
Message-ID: 

Hi Mark,

Thanks for your responses. I think it is valuable to challenge the 'latest and greatest' thing and thinking about your points has caused me to challenge and clarify some of my thoughts around MVC vs. 'library' development.

In response:
MVC differs from library development in that the structure of the framework explicitly separates business logic, data extraction & manipulation and presentation. By the sounds of things, your own library may be closer to the spirit of MVC than you may think.

MVC frameworks do include a set of helpful classes / helpers to abstract away such things as data insertion / extraction, validation (although cakePHP's inbuilt validation model often needs tweaking / extending), printing html content etc. As such, you're correct in saying they may need extending or altering to fit your project needs. In a way I'm sure this is similar for your own library - it gets extended or altered according to the needs of the application you're working on.

One benefit of MVC frameworks (apart from logical separation) for 'newer' developers is that the function 'library' is already there, so they don't have to roll their own.

I'm not sure I understand your point regarding production and Wordpress. Wordpress is free open-source software, but apart from that isn't really related to MVC frameworks. Can you clarify?

As an aside, a few negatives of MVC I've found in cakePHP development are (feel free to correct me):
- code reuse
MVC isn't great for promoting code reuse. Things like behaviours and components aid in this but sharing classes between models isn't as simple as I'd like.

- validation
I've found the validation model in cakePHP to be buggy / hard to use. I usually build my own.

- documentation
cakePHP is a great framework but the learning curve could be lessened by better documentation. RoR has a definite edge here, as 37Signals and the rails core group have done a great job or not only promoting Rails but providing clear documentation and help resources. cakePHP's material to me is representative of a 'typical open source PHP project' - a great piece of software poorly promoted, poorly presented and with less than fantastic documentation...

:)
Paul

From evolt_org at striderweb.com  Thu Nov 15 14:37:02 2007
From: evolt_org at striderweb.com (Stephen Rider)
Date: Thu, 15 Nov 2007 14:37:02 -0600
Subject: [thelist] Ruby on Rails: why?
In-Reply-To: 
References: <20E8922C-F68B-4022-B3E6-90E6528FE235@mountain.ch>
	
	<1FBD8D01-0E5A-4644-8276-5D23BBAF7475@mountain.ch>
	
	
Message-ID: <1E45D936-C4DF-4E41-9843-09A857E9B6CA@striderweb.com>


On Nov 15, 2007, at 9:04 AM, Mark Howells-Mead wrote:
>>>
>>> the high profile shareware/freeware CMS systems, like WordPress  
>>> for example, don't properly support multilingualism. There are of  
>>> course plug-ins, but the majority of them are likely to be way  
>>> less than they could be, because they're based on WordPress itself.
>>
>> WordPress is most definitely multi-lingual.
>> 
>
> That's to say, I can't use the default installation of WordPress to
> power a site where the visitor can switch between (say) English,
> French, German and Italian content at will. That requires an input
> environment which is capable of handling content in several languages
> concurrently, which WordPress - and, as far as I am aware, all other
> popular freeware CMS systems - cannot. (Not without a third party add-
> on, anyway.)

The general philosophy in WordPress is KISS -- put the major  
universal functions in core and leave the bloat for plugins, so users  
can install _only_ what they're going to use.  As the _vast majority_  
of bloggers aren't writing in multiple languages simultaneously, what  
you describe is definitely plugin territory.

HOWEVER: regarding quality of code and such -- a great many plugins  
are written by the same people writing the core.  I am on a mailing  
list for people writing WP core and plugins, and I have even seen  
instances where core was changed to make it easier in future to  
integrate certain types of plugins.

:)

Does WordPress have faults?  Of course.  But it's pretty robust, and  
designed from the start to be capable of significant flexibility via  
plugins.

Stephen

From dan at virtuawebtech.co.uk  Thu Nov 15 17:36:05 2007
From: dan at virtuawebtech.co.uk (Dan Parry)
Date: Thu, 15 Nov 2007 23:36:05 -0000
Subject: [thelist] Floaty CSS
In-Reply-To: <18210.207.245.75.194.1195151427.squirrel@webmail.fast-panda.com>
References: <18210.207.245.75.194.1195151427.squirrel@webmail.fast-panda.com>
Message-ID: <071801c827e0$4ab40b10$e01c2130$@co.uk>

> -----Original Message-----
> From: thelist-bounces at lists.evolt.org [mailto:thelist-
> bounces at lists.evolt.org] On Behalf Of megan cauley
> Sent: 15 November 2007 18:30
> To: thelist at lists.evolt.org
> Subject: Re: [thelist] Floaty CSS
> 
> Dan,
> 
> Not sure if this has been answered yet, I get the list digest...
> 
> I have used this JavaScript, I found somewhere,...
> to keep 2 divs equal heights. I have also called this function multiple
> times in one document.
> 
> **   
> 
> ...
> 
> **   
> 
> Whatever your div id's are set to that is where ('left') and ('right')
> would be....
> 
> 
> hth,
> 
> megan

Thanks Megan I did have a similar idea but it was going to be class-based
using a DOM collection... I decided against the method because I can be a
bit purist ;)... I was sure CSS alone could do it so decided to stick with
that route (well, I say route... It was more of a meander ending in a deep
river)

I do prefer to keep JS away from pure layout if I can at all help it :)

Again thanks... Can never have enough alternative methods

Dan
-- 
Dan Parry
 

No virus found in this outgoing message.
Checked by AVG Free Edition. 
Version: 7.5.503 / Virus Database: 269.15.32/1131 - Release Date: 14/11/2007
16:54
 


From martin at easyweb.co.uk  Thu Nov 15 18:17:43 2007
From: martin at easyweb.co.uk (Martin Burns)
Date: Fri, 16 Nov 2007 00:17:43 +0000
Subject: [thelist] Ruby on Rails: why?
In-Reply-To: 
References: <20E8922C-F68B-4022-B3E6-90E6528FE235@mountain.ch>
	
	<1FBD8D01-0E5A-4644-8276-5D23BBAF7475@mountain.ch>
	
Message-ID: 


On 15 Nov 2007, at 14:59, Stephen Rider wrote:

> Your phrasing is confusing -- don't know if you intended the double-
> negative, BUT just for the record...
>
> WordPress is most definitely multi-lingual.
> 

As is Plone, and Drupal has pretty strong multi-lingual support.

...and there's no shame in supporting multilingualism via plugin, if  
you have an architecture that has a very small core.

Cheers
Martin

> On Nov 15, 2007, at 6:27 AM, Mark Howells-Mead wrote:
>
>> For example, none of the high
>> profile shareware/freeware CMS systems, like WordPress for example,
>> don't properly support multilingualism. There are of course plug-ins,
>> but the majority of them are likely to be way less than they could
>> be, because they're based on WordPress itself.

--
 > Spammers: Send me email -> yumyum at easyweb.co.uk to train my filter
 > http://dspam.nuclearelephant.com/






From martin at easyweb.co.uk  Thu Nov 15 18:43:39 2007
From: martin at easyweb.co.uk (Martin Burns)
Date: Fri, 16 Nov 2007 00:43:39 +0000
Subject: [thelist] Ruby on Rails: why?
In-Reply-To: 
References: <20E8922C-F68B-4022-B3E6-90E6528FE235@mountain.ch>
	
Message-ID: 


On 14 Nov 2007, at 19:00, Paul Bennett wrote:

> Poor code? Bear in mind that most popular MVC frameworks are also  
> *very* popular open-source projects with many intelligent developers  
> not only extending but also securing and optimising the code, as  
> well as identifying and fixing bugs. Also some pretty high profile,  
> high traffic and high-load applications run on MVC frameworks (i.e.:  
> all of 37Signals stuff) so they've passed the production environment  
> test.

Welllll, yes and no.

As you'll see from testing URLs like:
http://www.basecamphq.com?=PHPE9568F35-D428-11d2-A769-00AA001ACF42
(ie add that query string to a PHP site and it will normally return  
the Zend logo), there's a lot more reliance on PHP in production code  
than some of the RoR cheerleaders might admit.

And I've picked up a theme that the rapidly developed code from using  
the RoR scaffolding shouldn't be your final production code; it's   
almost functional proof of concept code that of course you should  
optimise.

But this isn't a fault of the generic MVC design pattern, which will  
save you vastly in terms of coding effort in both initial development  
and ongoing maintenance - money that can easily be ploughed into  
improved hardware if performance becomes an issue.

Still, some other MVC type frameworks that might be useful:

http://www.djangoproject.org (if you like your Python)
http://struts.apache.org/ (if you like your Java)
http://www.phpmvc.net/ (if you like your PHP)
http://labs.adobe.com/wiki/index.php/Cairngorm (if you like your Java  
and your Flash/Flex)

Cheers
Martin

--
 > Spammers: Send me email -> yumyum at easyweb.co.uk to train my filter
 > http://dspam.nuclearelephant.com/






From prosayist at inbox.com  Fri Nov 16 08:00:45 2007
From: prosayist at inbox.com (Dan Stockton)
Date: Fri, 16 Nov 2007 09:00:45 -0500
Subject: [thelist] Link problems on html on desktop
In-Reply-To: <000a01c8270d$1d71bb30$0501a8c0@pc00012345678>
References: <000a01c8270d$1d71bb30$0501a8c0@pc00012345678>
Message-ID: <473DA28D.9060905@inbox.com>

> Ok, here's the story.
> 
> I was showing the kid how to..  make links to swf games.. on.. his laptop..
> .. Firefox .. Opened the page from IE.. Has anyone ever run
> across this before or better yet know a solution? 
> 
> Thanks, Breezy.

By default, files with SWF extension are opened with IE.
Firefox menu says "Save Link As..." (with 'k' being the accesskey)

try SWF Catcher extension for Firefox;

http://www.sothinkmedia.com/swf-catcher-firefox/

I prefer SWF associated with SWF-Opener, great tool for playing SWF with 
association available from the program's view menu (click 
'View/Associate SWF files with SWF-Opener')

http://browsertools.net/SWF-Opener/

Try the best Free Download Manager;

http://freedownloadmanager.org/

it's open-source now and the newest version has capabilities for FLV 
download from certain sites (YouTube, Google, etc.. )

Drag/Drop the link to the 'DropBox' to have FDM download the file to the 
place of your choosing, click the file to open with SWF Opener, play the 
game. Repeat as necessary.

-Dan

____________________________________________________________
FREE 3D MARINE AQUARIUM SCREENSAVER - Watch dolphins, sharks & orcas on your desktop!
Check it out at http://www.inbox.com/marineaquarium

From mwarden at gmail.com  Fri Nov 16 09:59:43 2007
From: mwarden at gmail.com (Matt Warden)
Date: Fri, 16 Nov 2007 10:59:43 -0500
Subject: [thelist] Truth about Outer Joins vs. Union All in Oracle?
Message-ID: 

listers:

There are rumors that outer joins should be avoided in Oracle in favor
of a semantically equivalent union all. I have sort of taken this as
truth, but the more I think about it the more I am not convinced. I
have done some googling, but have not found what I'm looking for.

The following are two versions of the same query, one with a left
outer join and the other using a union all. After each query is the
execution plan chosen by the cost-based optimizer.

Could someone with better experience reading execution plans confirm
that the union all should perform better in general (and an
explanation why, if possible)?

Probably want to view this in fixed font...

select nbr_ssn
from data_exch.t_exch_data_header d
left join data_exch.t_exch_match_header m
            on d.idn_header_data_exch = m.idn_header_data_exch;

Operation                Name                   Rows    Bytes   Cost
 CPU Cost   IO Cost
SELECT STATEMENT                                35M     640M    113506
 16G        112293
 HASH JOIN RIGHT OUTER                          35M     640M    113506
 16G        112293
  INDEX FAST FULL SCAN   PK_T_EXCH_MATCH_HEADER 1752K   10M     1722
 251M       1703
  PARTITION HASH ALL                            35M     438M    61548
 8G         60932
    TABLE ACCESS FULL    T_EXCH_DATA_HEADER     35M     438M    61548
 8G         60932


select nbr_ssn
from data_exch.t_exch_data_header d
inner join data_exch.t_exch_match_header m
            on d.idn_header_data_exch = m.idn_header_data_exch
union all
select nbr_ssn
from data_exch.t_exch_data_header d
where not exists (
            select *
            from data_exch.t_exch_match_header m
            where d.idn_header_data_exch = m.idn_header_data_exch
);

Operation                  Name                 Rows    Bytes
Cost	CPU Cost   IO Cost
SELECT STATEMENT                                35M     641M    227011
 16G        112293
  UNION-ALL
   HASH JOIN                                    1752K   31M     113506
 16G        112293
    INDEX FAST FULL SCAN   PK_T_EXCH_MATCH_HEAD 1752K   10M     1722
 251M       1703
    PARTITION HASH ALL                          35M     438M    61548
 8G         60932
     TABLE ACCESS FULL     T_EXCH_DATA_HEADER   35M     438M    61548
 8G         60932
   HASH JOIN RIGHT ANTI                         33M     609M    113506
 16G        112293
     INDEX FAST FULL SCAN  PK_T_EXCH_MATCH_HEAD 1752K   10M     1722
 251M       1703
     PARTITION HASH ALL                         35M     438M    61548
 8G         60932
      TABLE ACCESS FULL    T_EXCH_DATA_HEADER   35M     438M    61548
 8G         60932


All I can really take out of this is that the I/O cost of the second
is about twice the first (I assume because of the second table scan of
the table with  > 30 million records in it!). But I'm sure I'm only
considering one small piece of the puzzle. Any thoughts?

Is this a special case where the outer join is better? What about the
general case?

Thanks!

-- 
Matt Warden
Cincinnati, OH, USA
http://mattwarden.com


This email proudly and graciously contributes to entropy.

From mwarden at gmail.com  Fri Nov 16 10:07:24 2007
From: mwarden at gmail.com (Matt Warden)
Date: Fri, 16 Nov 2007 11:07:24 -0500
Subject: [thelist] Truth about Outer Joins vs. Union All in Oracle?
In-Reply-To: 
References: 
Message-ID: 

On 11/16/07, Matt Warden  wrote:
> The following are two versions of the same query, one with a left
> outer join and the other using a union all. After each query is the
> execution plan chosen by the cost-based optimizer.

Yikes, that wrapped badly. I've uploaded a txt file to my server to
save you guys the pain of reformatting:

http://mattwarden.com/explainplan.txt

-- 
Matt Warden
Cincinnati, OH, USA
http://mattwarden.com


This email proudly and graciously contributes to entropy.

From pturmel-webdev at turmel.org  Fri Nov 16 10:42:34 2007
From: pturmel-webdev at turmel.org (Phil Turmel)
Date: Fri, 16 Nov 2007 11:42:34 -0500
Subject: [thelist] Truth about Outer Joins vs. Union All in Oracle?
In-Reply-To: 
References: 
Message-ID: <473DC87A.2040504@turmel.org>

Matt Warden wrote:
> listers:
> 
> There are rumors that outer joins should be avoided in Oracle in favor
> of a semantically equivalent union all. I have sort of taken this as
> truth, but the more I think about it the more I am not convinced. I
> have done some googling, but have not found what I'm looking for.
> 
> The following are two versions of the same query, one with a left
> outer join and the other using a union all. After each query is the
> execution plan chosen by the cost-based optimizer.
> 
> Could someone with better experience reading execution plans confirm
> that the union all should perform better in general (and an
> explanation why, if possible)?
> 
> Probably want to view this in fixed font...
> 
> select nbr_ssn
> from data_exch.t_exch_data_header d
> left join data_exch.t_exch_match_header m
>             on d.idn_header_data_exch = m.idn_header_data_exch;
> 
> Operation                Name                   Rows    Bytes   Cost
>  CPU Cost   IO Cost
> SELECT STATEMENT                                35M     640M    113506
>  16G        112293
>  HASH JOIN RIGHT OUTER                          35M     640M    113506
>  16G        112293
>   INDEX FAST FULL SCAN   PK_T_EXCH_MATCH_HEADER 1752K   10M     1722
>  251M       1703
>   PARTITION HASH ALL                            35M     438M    61548
>  8G         60932
>     TABLE ACCESS FULL    T_EXCH_DATA_HEADER     35M     438M    61548
>  8G         60932
> 
> 
> select nbr_ssn
> from data_exch.t_exch_data_header d
> inner join data_exch.t_exch_match_header m
>             on d.idn_header_data_exch = m.idn_header_data_exch
> union all
> select nbr_ssn
> from data_exch.t_exch_data_header d
> where not exists (
>             select *
>             from data_exch.t_exch_match_header m
>             where d.idn_header_data_exch = m.idn_header_data_exch
> );
> 
> Operation                  Name                 Rows    Bytes
> Cost	CPU Cost   IO Cost
> SELECT STATEMENT                                35M     641M    227011
>  16G        112293
>   UNION-ALL
>    HASH JOIN                                    1752K   31M     113506
>  16G        112293
>     INDEX FAST FULL SCAN   PK_T_EXCH_MATCH_HEAD 1752K   10M     1722
>  251M       1703
>     PARTITION HASH ALL                          35M     438M    61548
>  8G         60932
>      TABLE ACCESS FULL     T_EXCH_DATA_HEADER   35M     438M    61548
>  8G         60932
>    HASH JOIN RIGHT ANTI                         33M     609M    113506
>  16G        112293
>      INDEX FAST FULL SCAN  PK_T_EXCH_MATCH_HEAD 1752K   10M     1722
>  251M       1703
>      PARTITION HASH ALL                         35M     438M    61548
>  8G         60932
>       TABLE ACCESS FULL    T_EXCH_DATA_HEADER   35M     438M    61548
>  8G         60932
> 
> 
> All I can really take out of this is that the I/O cost of the second
> is about twice the first (I assume because of the second table scan of
> the table with  > 30 million records in it!). But I'm sure I'm only
> considering one small piece of the puzzle. Any thoughts?
> 
> Is this a special case where the outer join is better? What about the
> general case?
> 
> Thanks!
> 
Hi Matt,

I don't see in the execution plan any indication that your join 
column is indexed (idn_header_data_exch) in the data table 
(idn_header_data_exch).  (?!?!) Unless I've misread something, 
that forces a full table scan in both halves of the union.

FWIW, I don't bother to convert outer joins to unions, ever 
(yet).  The loss of maintainability is just too expensive.  If I 
ever run into a tough case that absolutely needs this, I'd 
arrange for the FK column to be NULL when appropriate, and scan 
on that in the single-table half of the union.

Phil
-- 
Need to contact me offlist?
   Drop -webdev or you probably won't get through.

From mwarden at gmail.com  Fri Nov 16 11:31:29 2007
From: mwarden at gmail.com (Matt Warden)
Date: Fri, 16 Nov 2007 12:31:29 -0500
Subject: [thelist] Truth about Outer Joins vs. Union All in Oracle?
In-Reply-To: <473DC87A.2040504@turmel.org>
References: 
	<473DC87A.2040504@turmel.org>
Message-ID: 

On 11/16/07, Phil Turmel  wrote:
> I don't see in the execution plan any indication that your join
> column is indexed (idn_header_data_exch) in the data table
> (idn_header_data_exch).  (?!?!) Unless I've misread something,
> that forces a full table scan in both halves of the union.

Hi Phil, that column is the pk of both tables and you can see the use
of its index here:

INDEX FAST FULL SCAN   PK_T_EXCH_MATCH_HEADER

You can probably read the txt file version better than my mangled
inline version: http://mattwarden.com/explainplan.txt

The index is scanned twice in the union all version.

> FWIW, I don't bother to convert outer joins to unions, ever
> (yet).  The loss of maintainability is just too expensive.  If I
> ever run into a tough case that absolutely needs this, I'd
> arrange for the FK column to be NULL when appropriate, and scan
> on that in the single-table half of the union.

Well that would only work if the relationship is just right, where the
table you are joining on is on the right side of a 1-M, such that the
FK is on the driving table.

Thanks for your input, Phil. I agree with your assessment of the
maintainability impact.

-- 
Matt Warden
Cincinnati, OH, USA
http://mattwarden.com


This email proudly and graciously contributes to entropy.

From pturmel-webdev at turmel.org  Fri Nov 16 12:19:48 2007
From: pturmel-webdev at turmel.org (Phil Turmel)
Date: Fri, 16 Nov 2007 13:19:48 -0500
Subject: [thelist] Truth about Outer Joins vs. Union All in Oracle?
In-Reply-To: 
References: 	<473DC87A.2040504@turmel.org>
	
Message-ID: <473DDF44.1010808@turmel.org>

Matt Warden wrote:
> On 11/16/07, Phil Turmel  wrote:
> 
>>I don't see in the execution plan any indication that your join
>>column is indexed (idn_header_data_exch) in the data table
>>(idn_header_data_exch).  (?!?!) Unless I've misread something,
>>that forces a full table scan in both halves of the union.
> 
> 
> Hi Phil, that column is the pk of both tables and you can see the use
> of its index here:
> 
> INDEX FAST FULL SCAN   PK_T_EXCH_MATCH_HEADER

I did see this.

> 
> You can probably read the txt file version better than my mangled
> inline version: http://mattwarden.com/explainplan.txt
> 
> The index is scanned twice in the union all version.

What I don't see is an index scan of PK_T_EXCH_DATA_HEADER
                                                ^^^^
I have no idea why.  I would have expected to see it in BOTH 
versions of your query.  Could you share your table and index 
definitions?

What I also see is "HASH JOIN RIGHT ANTI" as the implementation 
of your "where not exists ()" clause.  Oracle chose to convert 
your subquery into a join.

>>FWIW, I don't bother to convert outer joins to unions, ever
>>(yet).  The loss of maintainability is just too expensive.  If I
>>ever run into a tough case that absolutely needs this, I'd
>>arrange for the FK column to be NULL when appropriate, and scan
>>on that in the single-table half of the union.
> 
> Well that would only work if the relationship is just right, where the
> table you are joining on is on the right side of a 1-M, such that the
> FK is on the driving table.

It's not so much that it's 1-M, but rather that the table in the 
single-table portion of the union can be filtered on its own 
content.  Your example needs a subquery, which is computationally 
equivalent to a join (typically).

So your query plan doubled because you went from "outer join" to 
"(inner join) union (outer join)".

HTH,

Phil

-- 
Need to contact me offlist?
   Drop -webdev or you probably won't get through.

From Brian at hondaswap.com  Fri Nov 16 13:30:37 2007
From: Brian at hondaswap.com (Brian Cummiskey)
Date: Fri, 16 Nov 2007 14:30:37 -0500
Subject: [thelist] regex help
Message-ID: <473DEFDD.7070400@hondaswap.com>

Hi guys,

I'm looking for a regex for some textareas.  Basically, anything goes, 
so I can't strip out everything.
On the contrary, I'm ALLOWING things instead.

Here's what i've come up with so far:


Function cleanup(str)
    str =Trim(str)
    Dim regEx
    Set regEx = New RegExp

    ' find ALL matching substrings, instead of just the first instance.
    regEx.Global = true
    regEx.Pattern = "[^0-9a-zA-Z\s-.]"

    'replace the junk with ''
    output = regEx.Replace(str, "")

    cleanup = output

End function


This gets me everything i need except single quotes and double quotes, 
but they need to be escaped.

any suggestions?



From joel at streamliine.com  Fri Nov 16 14:29:28 2007
From: joel at streamliine.com (Joel D Canfield)
Date: Fri, 16 Nov 2007 12:29:28 -0800
Subject: [thelist] web and db server local time
Message-ID: <72E9FAA171D63B48AAC707C72900E6B4D89F90@ireland.spinhead.com>

sanity check: I'm mucking with my homemade blog tool. inserting
timestamps is putting in a time one hour in the future. I'm in Pacific
Time; web/db servers are in AZ at Crystal Tech.

I'm a little foggy today, but I certainly don't remember building
timezone adjustments into web apps before. If I'm using forms in a
browser to insert/update info, shouldn't the web server be aware of my
local time via the browser?

yeah, I'm that foggy

joel

From bobm at dottedi.biz  Fri Nov 16 14:57:01 2007
From: bobm at dottedi.biz (Bob Meetin)
Date: Fri, 16 Nov 2007 13:57:01 -0700
Subject: [thelist] web and db server local time
In-Reply-To: <72E9FAA171D63B48AAC707C72900E6B4D89F90@ireland.spinhead.com>
References: <72E9FAA171D63B48AAC707C72900E6B4D89F90@ireland.spinhead.com>
Message-ID: <473E041D.8040305@dottedi.biz>

Joel, 

Not sure about the browser part.  However here is some php code I added 
to display the currentDate or currentTime on a page.  You have to adjust 
the TimeDiff according to the timezone of the hosting service.

-Bob

$TimeDiff = +2+date('');
$TimeZoneEpoc = time() - ($TimeDiff*60*60);
$currentDate = date('F j, Y',$TimeZoneEpoc);
$currentTime = date('g:i A',$TimeZoneEpoc);

Joel D Canfield wrote:
> sanity check: I'm mucking with my homemade blog tool. inserting
> timestamps is putting in a time one hour in the future. I'm in Pacific
> Time; web/db servers are in AZ at Crystal Tech.
>
> I'm a little foggy today, but I certainly don't remember building
> timezone adjustments into web apps before. If I'm using forms in a
> browser to insert/update info, shouldn't the web server be aware of my
> local time via the browser?
>
> yeah, I'm that foggy
>
> joel
>   


-- 
Bob Meetin
dotted i - Internet Strategies & Solutions
www.dottedi.biz
303-926-0167



From rjmolesa at consoltec.net  Fri Nov 16 15:15:27 2007
From: rjmolesa at consoltec.net (Jon Molesa)
Date: Fri, 16 Nov 2007 16:15:27 -0500
Subject: [thelist] web and db server local time
In-Reply-To: <72E9FAA171D63B48AAC707C72900E6B4D89F90@ireland.spinhead.com>
References: <72E9FAA171D63B48AAC707C72900E6B4D89F90@ireland.spinhead.com>
Message-ID: <20071116211527.GE17112@jenna.rjmolesa.homelinux.net>

*On Fri, Nov 16, 2007 at 12:29:28PM -0800 Joel D Canfield  wrote:

AFAIK, the time is local to the server.  Browsers don't report their
time without some kind of javascript.  Even if they did, server side
would have to make the adjustment and it isn't automatic.  In your
blogging tool are you pulling the time from the client system or the
server?

> Date: Fri, 16 Nov 2007 12:29:28 -0800
> From: Joel D Canfield 
> Subject: [thelist] web and db server local time
> To: thelist at lists.evolt.org
> 
> sanity check: I'm mucking with my homemade blog tool. inserting
> timestamps is putting in a time one hour in the future. I'm in Pacific
> Time; web/db servers are in AZ at Crystal Tech.
> 
> I'm a little foggy today, but I certainly don't remember building
> timezone adjustments into web apps before. If I'm using forms in a
> browser to insert/update info, shouldn't the web server be aware of my
> local time via the browser?
> 
> yeah, I'm that foggy
> 
> joel
> -- 
> 
> * * Please support the community that supports you.  * *
> http://evolt.org/help_support_evolt/
> 
> For unsubscribe and other options, including the Tip Harvester 
> and archives of thelist go to: http://lists.evolt.org 
> Workers of the Web, evolt ! 

-- 
Jon Molesa
rjmolesa at consoltec.net
if you're bored or curious
http://rjmolesa.com

From Ron.Luther at hp.com  Fri Nov 16 15:38:00 2007
From: Ron.Luther at hp.com (Luther, Ron)
Date: Fri, 16 Nov 2007 21:38:00 +0000
Subject: [thelist] web and db server local time
In-Reply-To: <20071116211527.GE17112@jenna.rjmolesa.homelinux.net>
References: <72E9FAA171D63B48AAC707C72900E6B4D89F90@ireland.spinhead.com>
	<20071116211527.GE17112@jenna.rjmolesa.homelinux.net>
Message-ID: 

Joel D Canfield asked:

>>shouldn't the web server be aware of my local time via the browser?


Hi Joel,


Not automagically.  And in a hand-rolled app, this is onna them things you hafta hand-roll.  ;-)

For a bloggy thing, one idea might be to lose the timestamp, just use date, and enter that as a field with your commentary.

Another idea might be to store everything in PST and add a "Dateline Left Coast" beneath the title in your display markup.

Thoidly, you might consider saving everything in the db in 'standard' Greenwich (or Wickedwich) time.  So if you move the server to another host in another time zone ... or start entering blog entries during your daily commutes between the Emerald Isle and Timbuktu or grow so big you need a distributed server farm large enough to shame inktomi ... you'll be able to figger out when is when ... if not where is where!


For more ideas, please insert another nickel!

HTH,
RonL.

When I had a site hosted in Australia (the currency conversion rate at the time made it toooo cheap to pass up) I left the display at server-time and put some kind of cheesey "in Sydney, Australia" next to the display. 

From anthony at baratta.com  Fri Nov 16 15:59:41 2007
From: anthony at baratta.com (Anthony Baratta)
Date: Fri, 16 Nov 2007 13:59:41 -0800
Subject: [thelist] regex help
Message-ID: 

What language are you using? Is this VBS/ASP?

-----Original message-----
From: Brian Cummiskey Brian at hondaswap.com
Date: Fri, 16 Nov 2007 11:30:37 -0800
To: "thelist at lists.evolt.org" thelist at lists.evolt.org
Subject: [thelist] regex help

> Hi guys,
> 
> I'm looking for a regex for some textareas.  Basically, anything goes, 
> so I can't strip out everything.
> On the contrary, I'm ALLOWING things instead.


From joel at streamliine.com  Fri Nov 16 16:06:42 2007
From: joel at streamliine.com (Joel D Canfield)
Date: Fri, 16 Nov 2007 14:06:42 -0800
Subject: [thelist] web and db server local time
References: <72E9FAA171D63B48AAC707C72900E6B4D89F90@ireland.spinhead.com><20071116211527.GE17112@jenna.rjmolesa.homelinux.net>
	
Message-ID: <72E9FAA171D63B48AAC707C72900E6B4D89F92@ireland.spinhead.com>

> Not automagically.  And in a hand-rolled app, this is onna 
> them things you hafta hand-roll.  ;-)

darn. too long on intranets instead of the real world.

yeah; I'll timestamp local and adjust in the display.

d'oh.

10Q Bob, Jon, Ron, and anyone else who's *not* laughing at me right now.

joel

From Brian at hondaswap.com  Fri Nov 16 16:24:24 2007
From: Brian at hondaswap.com (Brian Cummiskey)
Date: Fri, 16 Nov 2007 17:24:24 -0500
Subject: [thelist] regex help
In-Reply-To: 
References: 
Message-ID: <473E1898.3050004@hondaswap.com>

Anthony Baratta wrote:
> What language are you using? Is this VBS/ASP?
>   
yup, classic ASP


From rjmolesa at consoltec.net  Fri Nov 16 16:55:30 2007
From: rjmolesa at consoltec.net (Jon Molesa)
Date: Fri, 16 Nov 2007 17:55:30 -0500
Subject: [thelist] web and db server local time
In-Reply-To: <72E9FAA171D63B48AAC707C72900E6B4D89F92@ireland.spinhead.com>
References: 
	<72E9FAA171D63B48AAC707C72900E6B4D89F92@ireland.spinhead.com>
Message-ID: <20071116225530.GF17112@jenna.rjmolesa.homelinux.net>

No laughing here.  You did say you were running on fumes.  I had an
issue with people recieving my emails in the future or the far distant
past.  My system kept loosing the correct time.  It took me forever to
solve.  The simple solution was a daily sync with an internet time
server.

*On Fri, Nov 16, 2007 at 02:06:42PM -0800 Joel D Canfield  wrote:

> Date: Fri, 16 Nov 2007 14:06:42 -0800
> From: Joel D Canfield 
> Subject: Re: [thelist] web and db server local time
> To: thelist at lists.evolt.org
> 
> > Not automagically.  And in a hand-rolled app, this is onna 
> > them things you hafta hand-roll.  ;-)
> 
> darn. too long on intranets instead of the real world.
> 
> yeah; I'll timestamp local and adjust in the display.
> 
> d'oh.
> 
> 10Q Bob, Jon, Ron, and anyone else who's *not* laughing at me right now.
> 
> joel
> -- 
> 
> * * Please support the community that supports you.  * *
> http://evolt.org/help_support_evolt/
> 
> For unsubscribe and other options, including the Tip Harvester 
> and archives of thelist go to: http://lists.evolt.org 
> Workers of the Web, evolt ! 

-- 
Jon Molesa
rjmolesa at consoltec.net
if you're bored or curious
http://rjmolesa.com

From kendsnyder at gmail.com  Fri Nov 16 17:26:50 2007
From: kendsnyder at gmail.com (Ken Snyder)
Date: Fri, 16 Nov 2007 16:26:50 -0700
Subject: [thelist] web and db server local time
In-Reply-To: <20071116211527.GE17112@jenna.rjmolesa.homelinux.net>
References: <72E9FAA171D63B48AAC707C72900E6B4D89F90@ireland.spinhead.com>
	<20071116211527.GE17112@jenna.rjmolesa.homelinux.net>
Message-ID: <473E273A.10701@gmail.com>

Jon Molesa wrote:
> AFAIK, the time is local to the server.  Browsers don't report their
> time without some kind of javascript.
> ...
Here's a way you can get browsers to report their timezone:



  
  
  
  
  




"minutes_offset" will then be posted to the server so that dates from 
and to the user can be converted.  "minutes_offset" is the offset from 
GMT/UTC time.  If the user has JavaScript disabled, then you can default 
to the server clock timezone or whatever.

- Ken Snyder

From rjmolesa at consoltec.net  Fri Nov 16 21:16:36 2007
From: rjmolesa at consoltec.net (Jon Molesa)
Date: Fri, 16 Nov 2007 22:16:36 -0500
Subject: [thelist] web and db server local time
In-Reply-To: <473E273A.10701@gmail.com>
References: <72E9FAA171D63B48AAC707C72900E6B4D89F90@ireland.spinhead.com>
	<20071116211527.GE17112@jenna.rjmolesa.homelinux.net>
	<473E273A.10701@gmail.com>
Message-ID: <20071117031635.GG17112@jenna.rjmolesa.homelinux.net>

Nice!  Thanks.

*On Fri, Nov 16, 2007 at 04:26:50PM -0700 Ken Snyder  wrote:

> Date: Fri, 16 Nov 2007 16:26:50 -0700
> From: Ken Snyder 
> Subject: Re: [thelist] web and db server local time
> To: "thelist at lists.evolt.org" 
> 
> Jon Molesa wrote:
> > AFAIK, the time is local to the server.  Browsers don't report their
> > time without some kind of javascript.
> > ...
> Here's a way you can get browsers to report their timezone:
> 
> 

>   
>   
>   
>   
>   
> 

> 
> 
> "minutes_offset" will then be posted to the server so that dates from 
> and to the user can be converted.  "minutes_offset" is the offset from 
> GMT/UTC time.  If the user has JavaScript disabled, then you can default 
> to the server clock timezone or whatever.
> 
> - Ken Snyder
> -- 
> 
> * * Please support the community that supports you.  * *
> http://evolt.org/help_support_evolt/
> 
> For unsubscribe and other options, including the Tip Harvester 
> and archives of thelist go to: http://lists.evolt.org 
> Workers of the Web, evolt ! 

-- 
Jon Molesa
rjmolesa at consoltec.net
if you're bored or curious
http://rjmolesa.com

From kasimir.k.lists at gmail.com  Sun Nov 18 07:38:05 2007
From: kasimir.k.lists at gmail.com (kasimir-k)
Date: Sun, 18 Nov 2007 13:38:05 +0000
Subject: [thelist] web and db server local time
In-Reply-To: <72E9FAA171D63B48AAC707C72900E6B4D89F92@ireland.spinhead.com>
References: <72E9FAA171D63B48AAC707C72900E6B4D89F90@ireland.spinhead.com><20071116211527.GE17112@jenna.rjmolesa.homelinux.net>	
	<72E9FAA171D63B48AAC707C72900E6B4D89F92@ireland.spinhead.com>
Message-ID: <4740403D.3040106@gmail.com>

Joel D Canfield scribeva in 16/11/2007 22:06:
> yeah; I'll timestamp local and adjust in the display.

I'd recommend saving time in UTC (Time formerly known as Greenwich Mean
Time (GMT)). That way you'll ever only to have figure out client's time
zone (with javascript or the user choosing their time zone). If you save
your times in any other than UTC, then you'll have to worry both the
server's and client's time zones. Also dealing with DST is much easier
if you store UTC.

.k


From joel at streamliine.com  Sun Nov 18 09:32:26 2007
From: joel at streamliine.com (Joel D Canfield)
Date: Sun, 18 Nov 2007 07:32:26 -0800
Subject: [thelist] web and db server local time
References: <72E9FAA171D63B48AAC707C72900E6B4D89F90@ireland.spinhead.com><20071116211527.GE17112@jenna.rjmolesa.homelinux.net>	<72E9FAA171D63B48AAC707C72900E6B4D89F92@ireland.spinhead.com>
	<4740403D.3040106@gmail.com>
Message-ID: <72E9FAA171D63B48AAC707C72900E6B4D89F99@ireland.spinhead.com>

> If you save
> your times in any other than UTC, then you'll have to worry both the
> server's and client's time zones. Also dealing with DST is much easier
> if you store UTC.

well, even though I'm the only user, and probably always will be, this
is better future-proofing. thanks for the nudge.

joel

From joshua at waetech.com  Sun Nov 18 15:32:50 2007
From: joshua at waetech.com (Joshua Olson)
Date: Sun, 18 Nov 2007 16:32:50 -0500
Subject: [thelist] regex help
In-Reply-To: <473DEFDD.7070400@hondaswap.com>
References: <473DEFDD.7070400@hondaswap.com>
Message-ID: <001801c82a2a$934eff00$7801a8c0@OLSONXPS>

> -----Original Message-----
> From: Brian Cummiskey
> Sent: Friday, November 16, 2007 2:31 PM
> 
> Hi guys,
> 
> I'm looking for a regex for some textareas.  Basically, 
> anything goes, so I can't strip out everything.
> On the contrary, I'm ALLOWING things instead.

Try:

regEx.Pattern = "[^0-9a-zA-Z\s-.""']"

Joshua

<><><><><><><><><><>
Joshua L. Olson
WAE Technologies, Inc.
http://www.waetech.com/
Phone: 706.210.0168
Fax: 413.812.4864
Private Enterprise Number: 28752

Monitor bandwidth usage on IIS6 in real-time:
http://www.waetech.com/services/iisbm/ 


From santilal at scorpioneng.co.nz  Sun Nov 18 15:34:52 2007
From: santilal at scorpioneng.co.nz (Santilal Parbhu)
Date: Mon, 19 Nov 2007 10:34:52 +1300
Subject: [thelist] Mysql connection failed
In-Reply-To: <09DEC79B6F995248B402303C86CA4912237AAF@scarlet.ActiveIDE.local>
Message-ID: 

Thanks Guys,

It is a strange problem.  It has re-occurred a couple of times over the last
weeks and worked normally a couple of times as well.  I can't find any
pattern.  I have tried the username of 'admin'@'%', but that didn't seem to
help.

I have had two thoughts about this:

The first is that I recently installed WAMP, because I was going to look at
upgrading to PHP5.  WAMP has Apache Server (which I have disabled) and also
PHP5 and MySQL.  Could it be that on some occasions it is calling the wrong
MySQL?  How is the MySQL address specified?  (Note that these problems only
started after I loaded WAMP so this looks like a possibility.)

The other thought I had was that when starting my laptop (on which I am
doing all the development) I get a couple of errors:

1.	Explorer.exe - Could not locate framedyn.dll.
2.	Rundll - Could not locate c:/WINDOWS/system32\xcnkljrv.dll.

I have no idea what these errors mean.

Thanks.

Santilal


-----Original Message-----
From: thelist-bounces at lists.evolt.org
[mailto:thelist-bounces at lists.evolt.org] On Behalf Of Chris Anderson
Sent: Thursday, 15 November 2007 10:24 a.m.
To: thelist at lists.evolt.org
Subject: Re: [thelist] Mysql connection failed

> You mention 'improving security' - any chance you've tightened up
> database security? It looks as if your database access is failing due
> to an incorrect username or password, or the user account doesn't have
> access to the database.

+1 - but also check the host because of the MySql way of treating
admin at localhost and admin at devmachine as separate accounts.

If it's intermittent, it *may* even be the localhost thing - perhaps
MySql is sometimes seeing the host as localhost and sometimes as
servername, so check you have 'admin'@'localhost' *and*
'admin'@'servername' (or just define 'admin'@'%' to cover all hosts if
you are not worried about third party access)

Chris
-- 

* * Please support the community that supports you.  * *
http://evolt.org/help_support_evolt/

For unsubscribe and other options, including the Tip Harvester 
and archives of thelist go to: http://lists.evolt.org 
Workers of the Web, evolt ! 


From meredith at pintsize.com  Sun Nov 18 16:59:22 2007
From: meredith at pintsize.com (Meredith Tupper)
Date: Sun, 18 Nov 2007 17:59:22 -0500
Subject: [thelist] looking for a Drupal developer
Message-ID: <4740C3CA.3060009@pintsize.com>

I have a simple 5 or 6 page site which I'd like to transfer to 
Drupal for easy CMS.  I have played with a bit but my steep learning 
curve is hindering progress.  If you're interested, please reply off 
list.

Thanks,
Meredith


From Chris at activeide.com  Sun Nov 18 18:45:39 2007
From: Chris at activeide.com (Chris Anderson)
Date: Mon, 19 Nov 2007 00:45:39 -0000
Subject: [thelist] Mysql connection failed
References: 
Message-ID: <09DEC79B6F995248B402303C86CA4912237AB3@scarlet.ActiveIDE.local>

> The first is that I recently installed WAMP, because I was going to
> look at
> upgrading to PHP5.  WAMP has Apache Server (which I have disabled) and
> also
> PHP5 and MySQL.  Could it be that on some occasions it is calling the
> wrong
> MySQL?  How is the MySQL address specified?  (Note that these problems
> only
> started after I loaded WAMP so this looks like a possibility.)

Can you configure the new instance of MySQL?

If so, try adding the same user at host that you use on the "proper"
instance, and see if the error changes to not being able to find the
database ;-)

Chris

From santilal at scorpioneng.co.nz  Sun Nov 18 21:10:46 2007
From: santilal at scorpioneng.co.nz (Santilal Parbhu)
Date: Mon, 19 Nov 2007 16:10:46 +1300
Subject: [thelist] Mysql connection failed
In-Reply-To: <09DEC79B6F995248B402303C86CA4912237AB3@scarlet.ActiveIDE.local>
Message-ID: 

Thanks Chris,

I haven't had a chance to look into configuring the new instance of MySQL,
but I did have a look to see if there was any data in the new database.  To
my amazement, all of the tables from my earlier instance of MySQL and also
the users, passwords and grants were the same.  Somehow, it must have copied
them.  I don't know enough about how MySQL works to understand why or how
this happened.  So it does seem possible that your suggestion might be
worthwhile, even though you would expect that if the same users and
permissions are copied onto the new database, then access should not be a
problem.  I will keep on it.  Thanks again.

Santilal

-----Original Message-----
From: thelist-bounces at lists.evolt.org
[mailto:thelist-bounces at lists.evolt.org] On Behalf Of Chris Anderson
Sent: Monday, 19 November 2007 1:46 p.m.
To: thelist at lists.evolt.org
Subject: Re: [thelist] Mysql connection failed

> The first is that I recently installed WAMP, because I was going to
> look at
> upgrading to PHP5.  WAMP has Apache Server (which I have disabled) and
> also
> PHP5 and MySQL.  Could it be that on some occasions it is calling the
> wrong
> MySQL?  How is the MySQL address specified?  (Note that these problems
> only
> started after I loaded WAMP so this looks like a possibility.)

Can you configure the new instance of MySQL?

If so, try adding the same user at host that you use on the "proper"
instance, and see if the error changes to not being able to find the
database ;-)

Chris
-- 

* * Please support the community that supports you.  * *
http://evolt.org/help_support_evolt/

For unsubscribe and other options, including the Tip Harvester 
and archives of thelist go to: http://lists.evolt.org 
Workers of the Web, evolt ! 


From mwarden at gmail.com  Sun Nov 18 21:17:35 2007
From: mwarden at gmail.com (Matt Warden)
Date: Sun, 18 Nov 2007 22:17:35 -0500
Subject: [thelist] Truth about Outer Joins vs. Union All in Oracle?
In-Reply-To: <473DDF44.1010808@turmel.org>
References: 
	<473DC87A.2040504@turmel.org>
	
	<473DDF44.1010808@turmel.org>
Message-ID: 

Phil,

Thanks again for your response. See below:

On 11/16/07, Phil Turmel  wrote:
> What I don't see is an index scan of PK_T_EXCH_DATA_HEADER
>                                                 ^^^^
> I have no idea why.  I would have expected to see it in BOTH
> versions of your query.  Could you share your table and index
> definitions?

I wouldn't expect that the DB would use the index. It has to check
every row of T_EXCH_DATA_HEADER one-by-one; I would think converting
to and from rowid would just be extra computation in this case.

> What I also see is "HASH JOIN RIGHT ANTI" as the implementation
> of your "where not exists ()" clause.  Oracle chose to convert
> your subquery into a join.
>
...
> So your query plan doubled because you went from "outer join" to
> "(inner join) union (outer join)".

So, when people talk about outer join vs. union all, are they talking
about unioning an uncorrelated subquery and an IN clause?

-- 
Matt Warden
Cincinnati, OH, USA
http://mattwarden.com


This email proudly and graciously contributes to entropy.