57554 - </sup> outside </p> causes all text after to render in superscript

Status: VERIFIED FIXED

(Core :: DOM: HTML Parser, defect, P3)

Description

[Cesar Eduardo Barros]

From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; Linux 2.4.0-test9-pre1 i686; en-US; m18)
Gecko/20001019
BuildID:    2000101908

The page has a font size which decreases when it shouldn't

Reproducible: Always
Steps to Reproduce:
Open the page

Actual Results:  The middle of the page gets small fonts, which get smaller, and
then smaller again.

Expected Results:  Showed a normal font size

<kirun> cesarb: tags are closed in wrong order
<kirun> e.g. <p>foo<sup>bar</p></sup>

Shouldn't quirks mode avoid that problem?

Comment 1

[James Lariviere]

IMO, mozilla is rendering this right.  You point out youself that this is
incorrect markup.

When would you draw the line for mozilla to render incorrect markup...?

IMHO, I would rather see programmers fixing incorrect mozilla renderings of
correct code than adding extra code to mozilla to render incorrect web authored
code.

Suggest WONTFIX.

Comment 2

[James Lariviere]

I am marking this INVALID because the problem is caused by tags closed in the
wrong order.

Reporter: if you think this is incorrect I would suggest evagelism to the web
site author.

Comment 3

[James Lariviere]

Reopening since other browsers render this as web author really intended even
though there is incorrect ending tag order. Also, adding keyword, and marking OS
ALL.

Changing summary. Old summary:"The ever-decreasing font size bug strikes again".

Reporter: when you say "bug strikes again" is this a bug you have reported before?

Thanks buster!

Comment 4

[James Lariviere]

Sorry about the spam! Confirming.

Comment 5

[Kevin McCluskey (gone)]

Dividing up Clayton's bugs to triage

Comment 6

[Cesar Eduardo Barros]

Yes, it is. Sorta. The previous one dealt with unclosed tags.

I can't search for "all bugs I've ever touched", and have a bug open on that
problem; if you somehow manage to do it, it was in a rfc-editor.org page IIRC
(I don't have more than 30 bugs with me as a submitter; just list them all).

The symptoms were exactly the same: the font size gets smaller and smaller.
Since it was "unclosed tags" then and now it's "reverse closed tags", I don't
really know if this is a regression or not.

Comment 7

[dcone (gone)]

triage

Comment 8

[buster]

this sounds a lot like a parser issue.  It's the parser's job to feed the layout
system rational input, even when the source document is irrational.  It would be
*great* if you could produce a tiny testcase for this bug.

Comment 9

[rickg]

Attachment: Min testcase...

Comment 10

[Cesar Eduardo Barros]

Found my previous bug, awful rendering in rfc-editor's page, at
http://bugzilla.mozilla.org/show_bug.cgi?id=22596

Comment 11

[rickg]

inline element closure rules were wrong in nested cases. All better now.

Comment 12

[Chris Petersen]

Marking verified per last comments.