[Talk] Web Browser Exclusion
mgoodhew at pcug.org.au
Wed Dec 25 19:26:59 EST 2002
On Sun, 8 Dec 2002, David J N Begley wrote:
> Has anyone had any success in convincing sites to change their exclusory
> practices when it comes to Web browser detection? By this I mean when I site
This may be re-enlivening an old and well-beaten thread for many (sorry),
but I thought I might have some vaguely interesting thoughts to add and
I'd like to digest the previous discussion first.
My background: I've just returned from Cambridge, UK and the employ of ANT
limited (www.antlimited.com - don't think this is much of a commercial
plug, somehow I doubt anyone on this list will me in their target
market). Their primary product at the moment is a small, solid, portable
web browser for embedded applications called Fresco. I was the product
development team leader there shortly before I left. I have worked on
general bug fixing and implementing a sound subsystem, MP3 decoder
interface and helped incorporate the code for Macromedia Flash 5.
The thing is we were constantly beaten-up by our customers about
"incompatibility" with various web pages (the most common of which was
Hotmail, not surprisingly. Every now and then we had to augment our
default UA string to get around this). When this was a "Please upgrade to
IE X" case, it was a relatively straightforward response to reject the
error report (which may actually have helped us as we wouldn't then
actually experience any bugs on the following pages and thus have to fix
them :-). Often however, the errors were due to some other conflict in the
designer's expectations of how things are supposed to behave and how they
actually perform in all implementations.
One suggestion raised in this discussion (sorry, using pine over telnet,
so it's not convenient to check who said it) was that web browser
manufacturers should write code to be compatible with all standards,
defacto or otherwise. Having aimed at this goal in my last job, I can tell
you that it's a lot harder to achieve than you think (even when working on
Fresco, you were constantly reminded of how difficult it could be
sometimes). Some standards (usually the "real" W3C ones and some kind of
defacto - usually IE) are directly incompatible.
engine, but has "a DOM 0" (i.e. no more standardised than anyone else's
DOM 0 - there is work in progress to bring this up to "real" DOM
standards though). More than one customer had assumed that "oils is
oils" with ECMAScript and then raised a bazillion bug reports when they
found paces with scripts that didn't work with our DOM.
Another point from the discussion is that web designers tend to come from
a graphic design background. This is partially true in my experience
(graphic design and general publishing/layout), but a large proportion of
the web design community seem (or at least seemed) to be "cowboys" with no
relevant professional background at all ("I used to be a brickie's
labourer, now I'm a webmaster!"). It was suggeted in the discussion that
designers are deliberately targetting IE as a "best experience for a broad
base" platform. I think this credits them with more forethought than
they're capable. Typically I think web designers just design for their own
browser and are blissfuly unaware of many more (usually the response to a
suggestion to improve a web page is initially confusion and then anger to
mask their ignorance and insecurity).
Now onto the productive side:
It was suggested that a media campaign would be a possible way to address
this problem. I think this is a good idea and would back it to the extent
that I can help write/edit press releases or whatnot to this end. A little
while ago there was a bit of a thread about how AUUG has somewhat "lost
its way" and is seeking new directions. Also, there's been the "Open
source in government" discussion and submissions. I think that AUUG's
collective experience and knowlege would be well utilised in helping the
community in general by emitting press releases like this from time to
time. AUUG's general media profile could certainly only improve from media
Anyway, that's enough drivel from me for the time being. I'll see if I can
draft something up by the end of january to post to the list and in the
meantime, I'd be keen on hearing your thoughts on the issue.
Miles Goodhew - Unix system programming guru extraordinaire.
email: mgoodhew at pcug.org.au.
- You can finger this address for PGP public key.
More information about the Talk