[Webkit-unassigned] [Bug 11031] Another crazy counters bug

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=11031





--- Comment #17 from Shinichiro Hamaji <hamaji at chromium.org>  2009-11-16 06:00:52 PST ---
It seems that Carol's patch would still need some more work. For example, Darin
suggested me to invalidate subsequent counters instead of creating counter when
a render object with counter directives is added, but this patch seems to do
the latter way. So, even if we decide to ask Carol to fix this bug, I guess it
would make sense to land Bug 30505 first so that reviewers don't need to say
the same comments on Carol's patch. What do you think?

By the way, I'd like two more issues/requests in the patch:

> +    if (temp = node->parent())

I cannot compile your patch on my Mac due to this line. This line produces a
GCC warning and WebKit is using -Werror on Mac.

I think it would be better to have more smaller tests so that we can easily
distinguish the cause of failures in future. Specifically, I think this bug can
be divided into 4 smaller issues which are related each other:

1. dynamic addition of increment node
2. dynamic deletion of increment node
3. dynamic addition of reset node
4. dynamic deletion of reset node

Once these four issues are fixed, I think it's very easy to fix style change
case. So, I would like to have four tests. I was planning to fix this bug with
4 steps, and Bug 30505 was the first step.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.