[thelist] pixel perfect requirements and web standards

[Robert Gormley]

Shawn K. Quinn wrote:

>
>True, but you don't have to use GIF in order to put images on the Web in
>modern times. The void left by Unisys asserting its LZW patent was
>problematic for a time, but that's why we have PNG now (and IMO of
>course, PNG is a much better standard than GIF could ever have been).
>  
>
The only 'issue' I have, with reference to your previous remark, is that
PNG is, I've found, a lot less efficient. Even 8 bit PNGs I've found are
often 3-4 times larger than the GIFs. It is definitely a better standard
though.

>Which BSDs specifically? I only know of a native GNU/Linux plug-in and
>no end of problems running it under FreeBSD. Is there in fact a *native*
>FreeBSD plug-in? (note: *not* running the GNU/Linux plug-in under
>FreeBSD's Linux emulation)
>  
>
Actually, I'll grant, that may well have been the case.

>Also, I don't think the latest Flash plug-in is available for some of
>those, and some sites insist upon the latest plug-in.
>  
>
Is that inherently Flash's fault, though? Netscape 4 has horrible
support for CSS - is it Netscape's 'fault' for supporting the JS
necessary (in the case of client side, for one) for allowing sites to
browser-detect that "You must have the latest browser to use this site"?

>>I know of many free software packages that have less support - try compiling 
>>Gnome/GTK for HP/UX and see how much of a struggle it is.
>>    
>>
>
>With free software, you *can* make the changes to the source code
>required to get it running under HP/UX if needed. With binary-only
>software, you are pretty much limited to e-mailing the source and asking
>them for help getting it working.
>  
>
You can, with caveats. I think the quote I read was "Will never be
available, because some of the underlying core libraries just will not
support the functionality". When you start rewriting those core
libraries, and you go down, how far before you're rewriting the OS, (if
you even can), for the sake of this. While the potential is available,
is the effective end result any different to it being 'unavailable'?