201528 - browser ignores MARQUEE BEHAVIOR = "SLIDE"

Status: RESOLVED FIXED

(Core :: Layout, defect, P4)

Description

[Joey Bowles]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.4b) Gecko/20030407
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.4b) Gecko/20030407

<marquee behavior="slide"> does not work as designed. Instead of sliding to the
left and stopping, the SLIDE command is ignored and the text continues to scroll
to the left.

For a reference only, check out any verison of IE.

Reproducible: Always

Steps to Reproduce:
1. create the HTML page and open it


Actual Results:  
The text continued to scroll across the screen, time after time.

Expected Results:  
The text should scroll across the screen (right to left) and stop when it hits
the left edge.

My HTML:


<font color="orange" face="verdana" size=4>Default:: <marquee>This text should
continuously slide to the left</marquee></font>
 
<font color="blue" face="verdana" size=4>Slide:: <marquee behavior="slide">This
text should slide to the left then STOP</marquee></font>
 
<font color="green" face="verdana" size=4>Alternate:: <marquee
behavior="alternate">This text should slide to the left then to the
right</marquee></font>

Comment 1

[Josh Birnbaum]

-> doron

Marquee is not standard, and Mozilla only has a partial implementation.  This
might be a wontfix

Comment 2

[Joey Bowles]

It may not be "standard" HTML", but it is widely used. Most people don't know 
the difference (and some don't care) what is standard and what is not.

Don't mark it as "WON'T FIX" you should at least leave it open so someone can 
work on it if they want. Marquee is two thirds complete, why prevent someone 
from finishing it?

Comment 3

[Doron Rosenberg (IBM)]

Anyone who wants to do it, feel free.  I only implemented the most used commands.

Comment 4

[Mats Palmgren (inactive)]

-> Layout

Comment 5

[Rowan]

(In reply to comment #2)
> It may not be "standard" HTML", but it is widely used. Most people don't know 
> the difference (and some don't care) what is standard and what is not.
> 
> Don't mark it as "WON'T FIX" you should at least leave it open so someone can 
> work on it if they want. Marquee is two thirds complete, why prevent someone 
> from finishing it?

I agree, this should be fixed.

Comment 6

[Jo Hermans]

*** Bug 299169 has been marked as a duplicate of this bug. ***

Comment 7

[Martijn Wargers (dead)]

Updating url, the previous didn't exist anymore.
Another url with a behavior="slide" marquee:
http://www.mountaindragon.com/html/marquee.htm

Comment 8

[Martijn Wargers (dead)]

Attachment: patch1

This patch is contaminated with other stuff, but in principle this seems to
work just fine for me.

Comment 9

[Martijn Wargers (dead)]

Attachment: testcase

Comment 10

[Martijn Wargers (dead)]

Attachment: patch2

I also did some testing with:
https://bugzilla.mozilla.org/attachment.cgi?id=202035 (from bug 239919)
Dynamic changing behavior values isn't equal to IE6 with this patch. I could fix that, but I doubt the usefulness. I would need to add code for those cases and IE6's handling seems also quite buggy in particular cases, anyway.

Comment 11

[Doron Rosenberg (IBM)]

Comment on attachment 203174 [details] [diff] [review]
patch2

Would it not be better to also cancel the setTimeout once we finish the slide?

Comment 12

[Martijn Wargers (dead)]

(In reply to comment #11)
> (From update of attachment 203174 [details] [diff] [review] [edit])
> Would it not be better to also cancel the setTimeout once we finish the slide?
You're right.

I'm afraid my previous patch from bug 239919, which was checked in, caused some regressions:
- https://bugzilla.mozilla.org/attachment.cgi?id=201562 (testcase4 from bug 314369)
- https://bugzilla.mozilla.org/attachment.cgi?id=202035 (testcase1 from bug 239919) : Press "set behavior=alternate" and then "direction=up", you'll notice that the marquee speed has increased.

I'll attach a patch that fixes these issues.

Comment 13

[Martijn Wargers (dead)]

Attachment: patch3

This patch fixes this bug and the regressions I mentioned before.
A lot more of the code just needs to be in the DOMAttrModified handler.
The extra truespeed checks, I added in bug 239919, caused the first regression. I've removed those (since they are only useful for truespeed=0 and truespeed='' when dynamically changing truespeed values, so not important for now at least, imho).

Comment 14

[Martijn Wargers (dead)]

The Microsoft example doesn't show up completely 'right', btw.
The top marquee gets partly hidden with the patch at the end.
But I think that's a different issue. It happens because of the use of margin inside of the marquee.
A new bug should be filed about that.

Comment 15

[Johnny Stenback (:jst)]

Comment on attachment 203525 [details] [diff] [review]
patch3

sr=jst

Comment 16

[:Gavin Sharp [email: [e-mail]]]

mozilla/layout/style/xbl-marquee/xbl-marquee.xml; new revision: 1.21;

Comment 17

[Ria Klaassen (not reading all bugmail)]

*** Bug 359507 has been marked as a duplicate of this bug. ***