Here's a quote from a recent post, which I've modified to make a point.
I've removed the attribution, as my intention is not to single out any
person in particular, but to make an observation about something I've seen
with disappointing regularity:
> I recognize the frustration with browser A. Who here hasn't finisheda
> beautiful layout in browser B, and then screamed obscenities and
> frightened coworkers and family when you saw how browser A mangledyour
> beautiful and validated work? I know I have!
Browsers A and B can be *any* browser. The problem is the work process by
which a design is only checked in one, maybe two browsers while the
development proceeds, then when everything is "finished" it is checked in
Software engineering has been slowing moving towards a maxim of "test
early, test often". It's taken many years to get there, but unit testing
is now a common practice. Perhaps it's because a lot of web designers are
more "artist" than "engineer" that that doesn't seem to be happening in
this field? Not that I have anything against artists! :-)
Most people will have a list of browsers they'd like their sites to work
with. The exact list is not important. What is important is that if you're
going to check your site in browser X, then you start checking right from
the beginning, not at the very end!
As a user of a minority browser (Opera), I perhaps see the results of lack
of testing more often than others. I've lost count of the number of times
I've found a one or two-line fix for a web designer oversight.
Many studies have found that the earlier bugs are found, the cheaper,
quicker, and easier they are to fix. That applies to web design too.
I hope more of you might consider earlier testing. I expect the result
would be happier web designers, and more robust cross-browser designs.
Thanks for reading!
Andrew Gregory, <URL: [EMAIL-REMOVED] >
<URL: http://www.scss.com.au/family/andrew/ >
List wiki/FAQ -- http://css-discuss.incutio.com/
Supported by evolt.org -- http://www.evolt.org/help_support_evolt/