In article <[EMAIL PROTECTED]>,
Scott I. Remick <[EMAIL PROTECTED]> wrote:
>Adding additional bloat to a browser to try and interpret bad code is
>not the right way to address the problem. Sure, in this case it might
>seem like a simple fix. Missing closing tags. So say Mozilla adds code
>for it. Tomorrow it'll be something else, and the person will say,
>"well, you supported bad html for missing closing tags for <FONT>, so
>you should support THIS bad html too". So Mozilla gives in and adds
>more code for that instance of bad HTML as well. Then a third
>situation, and a fourth, and so on.
Where do you get the idea that any extra interpretation is needed or
desired? I suggest you read the bugzilla notes about the issue that
I mentioned earlier--they're quite informative.
As a side note, the current algorithm gets very inefficient as the
tag nesting level increases. I bumped MAX_REFLOW_DEPTH from 200 to 2000
to see what'd happen, and layout gets slower and slower as the nesting
level increases. While I haven't studied the code, it feels O(n^2)-ish
to me. It may be the case that if the algorithm is changed to be
iterative rather than recursive, Mozilla will be faster and less bloated.
While your concerns about bloat and meeting Mozilla's shipping deadline,
whatever that might be, is commendable, they should be directed
elsewhere. Neither I, nor anyone else I've seen in this discussion has
said that this is a critical bug that must be fixed immediately. I
said that it'd be nice if Mozilla could handle this. As in when someone
gets around to it. Also, please note that Mozilla is what's known as an
"open source" project; the "Mozilla developers" you refer to are not a
fixed set of people. If someone decides to tackle this problem, it
won't slow down the rest of the development. As an example, development
didn't slow down because someone had to take time off from their busy
schedule to make Mozilla work on old 68K Macs running NetBSD--a platform
that very few people care about. An interested party (namely myself)
volunteered and did the missing XPCOM bits for it. This is the idea
behind open source; people who see a problem are free to fix it
themselves. Probably not me this time though, seeing that I'm pretty
busy these days.
--
Name: Dave Huang | Mammal, mammal / their names are called /
INet: [EMAIL PROTECTED] | they raise a paw / the bat, the cat /
FurryMUCK: Dahan | dolphin and dog / koala bear and hog -- TMBG
Dahan: Hani G Y+C 25 Y++ L+++ W- C++ T++ A+ E+ S++ V++ F- Q+++ P+ B+ PA+ PL++
