243345 - [FIX]bugzilla request dropdowns show strange character

Status: RESOLVED FIXED

(Core :: Layout: Form Controls, defect, P2)

Description

[Christian :Biesinger (don't email me, ping me on IRC)]

nightly 2004051006, gtk1 (no xft), linux/x86

The request dropdowns on bugzilla, when no flag is selected, show a strange
character.
looks like a rectangle, with a dotted outline. will attach a screenshot.

Comment 1

[Christian :Biesinger (don't email me, ping me on IRC)]

Attachment: screenshot

Comment 2

[Boris Zbarsky [:bzbarsky]]

Regressed between 2004-05-07-08 and 2004-05-08-08.  Checkins:

http://bonsai.mozilla.org/cvsquery.cgi?treeid=default&module=all&branch=HEAD&branchtype=match&dir=&file=&filetype=match&who=&whotype=match&sortby=Date&hours=2&date=explicit&mindate=2004-05-07+08%3A00%3A00&maxdate=2004-05-08+08%3A00%3A00&cvsroot=%2Fcvsroot

Possible culprits:

Bug 237760, caillon's checkin for style stuff, bug 242734.

Note that this is not a "strange character" -- it's just the focus rect being
mis-sized.

Comment 3

[Christian :Biesinger (don't email me, ping me on IRC)]

(In reply to comment #2)
> Regressed between 2004-05-07-08 and 2004-05-08-08.  Checkins:

whoa bz, you are too fast for me... I was still downloading a build to hunt for
the regression date

> Note that this is not a "strange character" -- it's just the focus rect being
> mis-sized.

oh. heh. cool. :)

going to mark major, as this is quite visible...

Comment 4

[Boris Zbarsky [:bzbarsky]]

Attachment: Patch

I was wrong.  It's a strange character... ;)

Comment 5

[Boris Zbarsky [:bzbarsky]]

Attachment: Patch for real

Comment 6

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 148241 [details] [diff] [review]
Patch for real

jst, this is all your fault, man!  The null terminator is _not_ part of the
string!

Comment 7

[Christian :Biesinger (don't email me, ping me on IRC)]

(i.e. this was caused by bug 242734)

Comment 8

[David Baron :dbaron:]

The implementation of SetText (in nsGenericDOMDataNode.cpp) assumes its argument
is null-terminated.

Comment 9

[Boris Zbarsky [:bzbarsky]]

Hmm.. so it does, when it does the nsDependentString thing.  We should probably
fix that, since the call signature makes it look that no such assumptions, and
neither does the API (or document the assumption and change this patch accordingly).

Comment 10

[Johnny Stenback (:jst)]

Attachment: Don't assume aBuffer is null-terminated.

Sorry for not replying here before, I had some major email problems all day
yesterday and missed this discussion.

How about this, in addition to bz's patch?

Comment 11

[Johnny Stenback (:jst)]

Comment on attachment 148241 [details] [diff] [review]
Patch for real

Duh :-) r+sr=jst

Comment 12

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 148346 [details] [diff] [review]
Don't assume aBuffer is null-terminated.

What about the char* version of SetText?  It has the same issue, no?

Comment 13

[Johnny Stenback (:jst)]

Yeah, duh, fixed locally. Checking in (r+sr=bz over IRC).

Comment 14

[Johnny Stenback (:jst)]

Fixed.