Bug 15056 - REGRESSION (r21472): Digg Podcasts Episodes Render "Digg" counter incorrectly
Summary: REGRESSION (r21472): Digg Podcasts Episodes Render "Digg" counter incorrectly
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: Page Loading (show other bugs)
Version: 523.x (Safari 3)
Hardware: Mac OS X 10.4
: P1 Normal
Assignee: Nobody
URL: http://digg.com/podcasts/Diggnation
Keywords: InRadar, NeedsReduction, Regression
Depends on:
Blocks:
 
Reported: 2007-08-22 17:29 PDT by Neil Daniels
Modified: 2007-08-25 05:36 PDT (History)
3 users (show)

See Also:


Attachments
Reduction (148 bytes, text/html)
2007-08-23 01:35 PDT, mitz
no flags Details
Single-block test case (120 bytes, text/html)
2007-08-23 01:54 PDT, mitz
no flags Details
Don't make inner empty inlines (11.32 KB, patch)
2007-08-24 01:59 PDT, mitz
hyatt: review+
Details | Formatted Diff | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description Neil Daniels 2007-08-22 17:29:04 PDT
This is kinda hard to explain so here is a screenshot: http://myskitch.com/neild/digg_podcasting_episode_bug-20070822-171844/

Basically, the digg counter on any podcast's episode list becomes misaligned. It seems that it _might_ be a problem with the way that it handles CSS. It should not be pushed down like it is.

This bug occurs on r25168, along with r25145; it may go further back but I haven't tested it that far, as I never noticed this before. This bug DOES NOT occur on Safari Version 3.0.3 (522.12.1).
Comment 1 mitz 2007-08-23 01:04:59 PDT
Regressed in <http://trac.webkit.org/projects/webkit/changeset/21472>.
Comment 2 mitz 2007-08-23 01:35:06 PDT
Created attachment 16091 [details]
Reduction
Comment 3 mitz 2007-08-23 01:39:25 PDT
WebKit creates an <a href="1"></a> between the <li> and the <a href="2">, whereas Firefox and html5lib don't. See also bug 13712 (although in the case given there, html5lib agrees that the empty inline should be created).
Comment 4 mitz 2007-08-23 01:54:24 PDT
Created attachment 16092 [details]
Single-block test case

This test case demonstrates the bug in the single intermediate block case. r21472 did not change the behavior in this case, only extended it to cases with multiple intermediate blocks.
Comment 5 mitz 2007-08-23 05:03:48 PDT
Even simpler case:
<span><i>foo</span><div> </i>bar</div>

Comment 6 Ian 'Hixie' Hickson 2007-08-23 13:21:44 PDT
Whether or not to create inlines for spaces is a known bug. I want to make them not be created, but I haven't quite worked out what the spec should say.
Comment 7 mitz 2007-08-23 13:33:33 PDT
(In reply to comment #5)
> Even simpler case:
> <span><i>foo</span><div> </i>bar</div>

OK, that was a borderline case. However WebKit also creates an empty <i> inside the div for

<span><i>foo</span><div></i>bar</div>

which is what this bug is about.
Comment 8 Ian 'Hixie' Hickson 2007-08-23 19:56:04 PDT
Ah yeah, for that case I don't think the inline should be created per spec. Glad to know the spec is on the side of compatibility here!
Comment 9 mitz 2007-08-24 01:59:33 PDT
Created attachment 16106 [details]
Don't make inner empty inlines
Comment 10 Dave Hyatt 2007-08-24 15:28:52 PDT
Comment on attachment 16106 [details]
Don't make inner empty inlines

r=me
Comment 11 Mark Rowe (bdash) 2007-08-25 05:36:11 PDT
Landed in r25236.
Comment 12 Mark Rowe (bdash) 2007-08-25 05:36:33 PDT
<rdar://problem/5437711>