[thelist] XHTML and language="javascript"

Shawn K. Quinn skquinn at frogger.kicks-ass.net
Thu Jun 19 01:24:01 CDT 2003


On Wednesday June 18 2003 23:42, Charlie Griefer wrote:
> I don't know what kind of situations each of us are in
> professionally...but I work for a non profit organization that is
> woefully under staffed. Additionally, I pick up as many freelance
> gigs as I possibly can outside of work.  In both situations, I simply
> don't possess the resources to test each page I write in every
> browser on every platform.  I don't have the time or the manpower or
> the facilities.

That's what validators are for. That's why we have standards for HTML, 
CSS, image formats, etc. Not that testing in a browser is a bad thing; 
it's not, but beyond a certain point testing should not be necessary. 
(Though it should be noted Internet Explod^Hrer has a long track record 
of non-standards-compliant behavior, in particular, second-guessing an 
HTTP Content-Type of text/plain when the specification explicity 
forbids it, and a half-baked implementation of PNG image support.)

> Of course, in every task we do, we want to look at ROI.  Every
> statistic I've seen (and yes, I know that 86% of all statistics are
> meaningless) say that IE 5+ has over 90% of the market (I want to say
> 97%, but since I don't have any numbers in front of me to cite, I'd
> rather err on the side of caution).  Even the remaining percentage
> (be it NS or Mozilla or Opera) were on higher version browers, IIRC.

There's no way to reliably gauge just how many different people 
accessing your site are using a given browser. The HTTP User-Agent 
header can be spoofed and often is due to boneheaded "browser sniffing" 
that rejects anything that doesn't appear to be Netscape Navigator or 
Internet Explod^Hrer.

If your site happens to turn away a wealthy donor or a group of willing 
volunteers (since this is a non-profit organization you're talking 
about, I'm assuming both are potential visitors) because the donor or 
group leader is using some oddball browser you've never heard of that 
can't handle a particular coding mistake you've made, you may never 
hear of it. Even worse is inaccessibility for the people your site is 
trying to help (assuming that applies to your organization, obviously 
if it's an SPCA chapter or something similar that doesn't really apply 
nearly as much) or journalists.

> The question that leaps into my mind is...why expend such resources
> to satisfy approximately 5% of the population? 

With proper site authoring/design this should never be an issue. Just 
don't expend the effort to keep that percentage out.

> Isn't there, at some point, an 'acceptable level of loss'?  Instead of
> spending the resources to ensure that those 5% (as well as the 
> remaining 95%) have an 'adequate' experience...couldn't (or shouldn't)
> I invest that time in trying to ensure that the 95% have as memorable
> an experience as possible?  

If that 5% includes any significant number of volunteers (not sure if 
the "woefully under-staffed" above implies a lack of volunteers as 
well, assuming your organization has some and most NPOs do), one 
high-profile journalist, one filthy rich potential donor (or a 
comparable number of not-so-filthy-rich potential donors), or any 
significant number of people in need (who might be making do with 
Windows 3.1 or even DosLynx!), that's definitely not what I'd call an 
"acceptable level of loss." The horrifying thing is you may never know 
just who you've turned away, and you certainly won't know until it's 
too late. (Yes, I've sent e-mail to webmaster addresses emphatically 
stating my dissatisfaction with their respective sites. In at least one 
case, there was not even a webmaster address at the domain, which is 
perhaps one of the oldest Web-related conventions in existence.)

It's a bit different in the case of for-profit corporations, but the 
principle is the same.

> And no, I'm not talking about 'bleeding edge' in-your-face
> flash animations all over the site.  And yes, I will readily concede
> that content is the most important part of a web site (people surf
> the Web to find information).  However, the Web *is* a visual medium,

The Web is not exclusively a visual medium. Speech browsers and Braille 
terminals are out there. Lynx, a text-only Web browser and one of the 
first Web browsers in existence, is still in active development; see 
<URL:http://lynx.isc.org/current/> which shows a new development build 
released only 18 days ago (and note that some sites still have rude 
messages similar to "please come back with a browser from this decade", 
thankfully this is less common now that people realize that's exactly 
what the search engine indexes, too).

> and I believe that if Company B is offering the same content that I'm
> offering...but in a prettier package...most would opt to return to
> Company B's site over mine.

"Company B's site uses *way* too many images, I'm sick of waiting for it 
to load" (click back button)

"Another lame Flash-based intro at Company B, that's so 1998" (click 
back button)

"Oh, real nice, images for text that are so small I can barely read 
them, so much for Company B" (click back button)

"*%$&^#@!!! Company B's site crashed my browser, and they really think 
they're going to get *my* money now?" (restart browser, no need to 
click back button)

> Let's face it.  There *is* competition.  And with IE 5.5+ and NS 6+
> becoming (arguably) defacto standards...I need to keep up with the
> competition and present my information in as memorable a fashion as
> possible.  Quite honestly, I can't imagine anybody surfing the Web on
> NS 4.x anymore...

This I can agree with somewhat, as Navigator 4.x's parser/layout engine 
is terminally ill with regard to handling even basic CSS.

> or surfing with JS turned off. 

I do, every day. There's a very short list of sites that can run scripts 
on my computer, and it's very rare that I add to it. In some cases, I 
make another user account for using a specific site.

> It just seems to me that a *vast* majority of the sites out there
> today adhere to the aforementioned defacto standard.

I see more of them that are trying to get it right by the W3C's specs.

> If you're still reading...great (means you haven't yet hit the
> 'reply' button to tell me I'm nuts). 

:-)

> But there is one other reason I take this 'stance', as it were.  The
> lack of standards has been the nightmare of the Web developer for...7
> years now (brief history on me...my first professional HTML
> 'webmaster' gig was in 1996 when I had to check compatiblity between
> NS and Mosaic...so yes, I've been in the trenches).  We cry and moan
> about the lack of standards, and angrily point fingers at the browser
> manufacturers...

It's not the "lack of standards" per se, but the inability or 
unwillingness of programmers (mainly at Microsoft, where making 
software incompatible with standards on purpose is a routine practice) 
to implement what is in the standard and not what they feel like 
writing today.

> but I'd go so far as to say they're not the problem.  We are.  The 
> collective Web development community.  Those of us who
> jump thru hoops and bend over backwards making sure that no matter
> how crappy a browser is...that we've got a condition or object
> detection or some sort of code voodoo to make sure that our code will
> run properly in it.  We give these companies no reason to try and
> change.  No reason whatsoever to adhere to standards.  Why should
> they?  We're puppets on strings and we'll accomodate them.

I would agree that in certain instances, such as the refusal of most 
sites to use transparent PNGs because IE doesn't support them properly, 
or the many cases in which proper MIME types aren't sent because IE 
violates the spec and guesses based on either the content returned or 
the character sequence after the last dot in the URL (which, for the 
unenlightened, is *NOT* a filename extension because URLs aren't 
filenames), that the Web admins / authors / designers / developers / 
gurus / whatever you want to call them are at least somewhat at fault.
Of course, the real problem is a lack of awareness among the computing 
public running Windows that they do, in fact, have a choice of what 
browser to use, combined with Microsoft doing what they feel maximizes 
their profits as opposed to what is in the best interest of the Web and 
their customers. (When a company can ignore customers at will, they are 
a monopoly or near-monopoly, and need to be broken up; sadly, the US 
DOJ and court system missed their chances to slam the brakes on this 
train wreck waiting to happen.)

> Is it too idealistic or naive to think that if enough of us do take a
> stance, that perhaps these companies might take notice?  If we all
> write standards compliant code, and it runs in browser A, but not
> browsers B, C, or D...that people might stop using browsers B, C, and
> D?  And maybe browser manufacturers B, C, and D will take notice and
> work harder to ensure that their product displays compliant code in
> the expected manner?

Maybe, maybe not. Public awareness of the problem definitely needs to be 
raised, that's for sure. I mean, campaigns like Cari Burstein's at 
<URL:http://www.anybrowser.org/campaign/> have been going on for years. 
The specific example of browser-specific HTML tags are less of a 
concern now (though now we have proprietary CSS extensions which are 
almost as bad, and who knows what kinds of scripting incompatibilities; 
I don't do much with scripting so wouldn't elaborate too much there).

> (pausing for breath)
>
> That's all I've got.  I hope I haven't offended anybody, as that was
> not at all my intention.  I only post this to solicit input.  If I
> didn't have a great deal of respect for the folks on this list...I
> wouldn't have bothered. But I'm curious.  Am I the only one who feels
> this way?  I can't possibly be the only one out here who's tired of
> the wild inconsistencies between even different releases of the same
> browser.

The Web would not have gotten this far without some standardization. I 
share the belief that the entire HTML 3.2 debacle (basically, the 
complete ignorance of the CSS level 1 recommendation in favor of 
presentational attributes of HTML) pushed the state of the Web back 3 
to 4 years behind where it should be, and it's possible we're losing 
even more ground so today's Web may not be where we should have been in 
1999 or 2000, but closer to what we should have had in 1998 or so, 
which is admittedly a hugely chilling prospect. Just this week there 
was someone on another list asking about HTML 3.2 <body> attributes, 
deprecated in HTML 4.0 which is now over 5 years old! I was under the 
impression a "transitional" period was supposed to last maybe 2-3 years 
tops, after which it was pretty safe to write strict.

> I do hope to hear back from a few of you.  If the list is not the
> appropriate place for this particular discussion, feel free to e-mail
> me offlist.  Or if there is a more appropriate venue (an online forum
> perhaps?), lead the way and I'll be more than happpy to follow.

This seems to be appropriate enough unless I'm missing something. 
There's always the comp.infosystems.www.authoring.html newsgroup on 
Usenet.

-- 
Shawn K. Quinn



More information about the thelist mailing list