Re: newbie

Gupta, Anu said:
> Is it a fact that going cross-browser / cross-platform necessarily
> results in a disproportiate cost increase ? What factors affects the
> ratio of compatibility:cost ?

Speaking as someone who's been doing HTML since 1995 and who's currently
doing a second all-CSS redesign in two years -- the simple answer is yes,
supporting multiple browsers adds costs, sometimes substantial ones.
(That's one reason I co-founded the Web Standards Project -- not
one-ups-man-ship, just trying to provide some background....)

I realize this is somewhat off-topic, but it just illustrates the value of
knowing something about development -- since sometimes seemingly simple
differences in screen layout can significantly change the amount of work
needed to build it.

Unfortunately, even "modern" browsers, like IE5+, Netscape, Gecko, Safari
and Opera still have a significant number of bugs and other
incompatibilities among them. So while coding to the spec and having it
work is an ideal, it's still only an ideal.

For example, I literally spent two days trying to wrangle out problems
with a basic three-column layout -- and ended up resorting to a
server-side browser detect to feed alternate code to one unfixable
browser.

As far as the relationship between compability and cost, it's pretty
simple: the more browsers and the more versions, the more it costs. A few
specific examples:

* Until IE6, IE used a non-standard (although arguably more sensible) way
of calculating the height and width of layout boxes. So you have to have
duplicate code to get things to work the same between IE and
standard-compliant browsers like NS, Gecko and Safari.

* WinIE5.0 still poses some significant difficulties to developers, which
go away if you only support WinIE6 (or even WinIE5.5). It's the browser I
had to go to a server-side solution because it failed to handle something
correctly that WinIE5.5+ had no problems with.

* While MacIE5 is pretty good, it's showing it's age. I had to change the
layout to feed MacIE5 a particular set of fixed width values instead of
flexible width values because of a particular unfixable bug.

* There were similar bugs to work around in Netscape, Opera and Safari
(which at least has an excuse because it's still beta).

So trying to account for this sort of thing across seven widely-used
major/minor browsers (WinIE6, WinIE5.5, WinIE5.0, MacIE5, Netscape 7,
Opera and Safari) takes effort.

    --------------------------------------------------------------
     POSTINGS (in plain text): [email protected]
           SUBSCRIPTION CHANGES:  http://lists.syntagm.co.uk
            (or send email to mailto:[email protected])
    --------------------------------------------------------------