652055 - Seven of eight menu drop-downs ok, but eight one doesn't work.

Status: RESOLVED INVALID

(Core :: DOM: HTML Parser, defect)

Description

[Sandi]

User-Agent:       Mozilla/5.0 (Windows NT 5.1; rv:2.0.1) Gecko/20100101 Firefox/4.0.1
Build Identifier: Mozilla/5.0 (Windows NT 5.1; rv:2.0.1) Gecko/20100101 Firefox/4.0.1

A page on a forum application I coded for Invision Power "Board has 7 drop down menus to select four start hours/minutes and four stop hours/minutes, for a total of 8 drop down menus.

After 7 menu drop-downs Firfox v4.01 will stop displaying them correctly. The drop-down tool missing and the contents spilled down the page.

Reproducible: Always

Steps to Reproduce:
1. Load a page with more than seven drop down menus. It will happen on any webpage regardless of server OS.

Actual Results:  
The drop-down tool missing, but the drop down minutes contents are present and spilled down the page. See image below.

http://codinglair.com/image/ff4error.png

Expected Results:  
Drop down tool not missing and working.

This multiple drop-down application works as intended on various forums. It works with previous Firefox versions, Internet Explorer, Safari and Chrome.

Comment 1

[Sandi]

Correct first sentence to read: "A page on a forum application I coded for Invision Power "Board has 8 drop down menus to select four start hours/minutes and four stop hours/minutes, for a
total of 8 drop down menus."

Comment 2

[(mostly gone) XtC4UaLL [:xtc4uall]]

Do you get any relevant Error Console Output?

Is your Issue reproducible without Extensions in Safe-Mode and/or using a new Profile?
https://support.mozilla.com/en-US/kb/Safe+Mode
https://support.mozilla.com/en-US/kb/Managing+profiles#Creating_a_profile

Comment 3

[Sandi]

Not familiar with the Error Console.

Results are the same in Safe-Mode: have not made another profile. If necessary I will make a new one.

Comment 4

[Henrik Skupin [:whimboo][⌚️UTC+2]]

Does that also happen for you with Firefox 3.6.x?

Comment 5

[Henrik Skupin [:whimboo][⌚️UTC+2]]

Also it would be very helpful if you could minimize the code so we could have a testcase for it.

Comment 6

[Sandi]

Henrik, I'm sorry about getting back late on this, I've been very busy.

Yes it works with Firefox 3.6.x. As I stated above "It works with previous Firefox versions, Internet Explorer, Safari and Chrome."

I have made a stripped down version of the pertinent code, and will use the 'Add an attachment' link above.

The attached code simulates acting on settings normally queried from a database by storing typical data in an array called $config. Other than the simulated data the functions are similar to my application's functions.

This application is an admin control panel for a battle game I code for Invision IP.Board. This attached code will reproduce the Firefox 4.1 bug everytime.

Comment 7

[Sandi]

Attachment: A zip file with code for a test case.

Comment 8

[Sandi]

Besides using the testcase attachment, I have also uploaded it to my development server.

http://codinglair.com/test/ff41test.php

Comment 9

[Henrik Skupin [:whimboo][⌚️UTC+2]]

Thanks for the testcase, but could you please create a static version of it? A simple stripped down HTML file should be enough here.

Given your feedback that this is a regression could you please test the beta releases of Firefox 4 and check which beta has been introduced this issue? You can find the beta releases here:

http://stage.mozilla.org/pub/mozilla.org/firefox/releases/

Thanks

Comment 10

[Henrik Skupin [:whimboo][⌚️UTC+2]]

Can you please disable the hardware acceleration in preferences / advanced? Does that solve the issue? It works fine for me so I think it could be related to your graphics card. Further please tell us the graphic details from about:support.

Comment 11

[Sandi]

I have two systems.

1) Mainbaord: Biostar TA790GXB3 | CPU: AMD Phenom II X2 545 | Display: NVIDIA GeForce 9500 GT 1GB video ram

2) Mainbaord: PCchips A13G | AMD Athlon 64 3500 | Display: Geforce 8400 GS 256MG video ram

Disabling hardware acceleration on the 9500 made no difference. Cannot disable it on the 8400 (if that old beater even has it).

Comment 12

[Sandi]

Graphics details from about:support

[MAIN]
Adapter Description		NVIDIA GeForce 9500 GT
Vendor ID			10de
Device ID				0640
Adapter RAM			Unknown
Adapter Drivers			nv4_disp
Driver Version			6.14.11.9107
Driver Date			9-27-2009
Direct2D Enabled			Blocked on your graphics driver. Try updating your graphics driver to version 257.21 or newer.
DirectWrite Enabled		false (0.0.0.0, font cache n/a)
WebGL Renderer			(WebGL unavailable)
GPU Accelerated Windows		0/1

[Old beater backup]
Adapter Description		NVIDIA GeForce 8400 GS (Microsoft Corporation - WDDM v1.1)
Vendor ID			10de
Device ID				0404
Adapter RAM			256
Adapter Drivers			nvd3dum nvwgf2um
Driver Version			8.15.11.8593
Driver Date			5-14-2009
Direct2D Enabled			Blocked on your graphics driver. Try updating your graphics driver to version 257.21 or newer.
DirectWrite Enabled		false (6.1.7600.16699, font cache n/a)
WebGL Renderer			(WebGL unavailable)
GPU Accelerated Windows		0/1

Comment 13

[Henrik Skupin [:whimboo][⌚️UTC+2]]

You are using outdated graphics drivers on both machines. Can you please update both and test again?

Comment 14

[Sandi]

Both card drivers updated.

[9500]
Driver Version	6.14.12.7061
Driver Date	4-7-2011

[8400]
Driver Version	8.17.12.7061
Driver Date	4-7-2011

No different in results.

Comment 15

[Henrik Skupin [:whimboo][⌚️UTC+2]]

Can you please test in a fresh profile? As given in comment 3 you haven't done it yet.

If it still happens please test the beta releases as given in comment 9. That should give us a first insight and possible candidate builds.

Comment 16

[Sandi]

Created a new profile and started Firefox.

Problem still exists running under a new profile. btw I use FF4.1 on this main computer, but also dual boot it into Win7. On win7 I am using FF4.0 and the problems exists as well.

I would like to download and install several of the beta builds as suggested earlier, but I am busy behind schedule on a couple coding projects, so will not be until I find a bit more time.

Comment 17

[Henrik Skupin [:whimboo][⌚️UTC+2]]

This really sounds like an issue with the graphic card. Bas, do you have any idea?

Comment 18

[[not reading bugmail]]

Attachment: This code works in Firefox 6a1

I think bug is invalid.   All I did was add a closing </select> tag at the end of that drop down select which was not there, and it worked and renders fine.

Comment 19

[[not reading bugmail]]

Looks to me like user error ->WFM.

Comment 20

[Boris Zbarsky [:bzbarsky]]

It's not obvious to me that this is invalid.

The question is whether <select> should imply </select>.  I'm guessing the HTML5 spec says it shouldn't (in which case either this is invalid or the spec needs changing), but Henri should confirm.

Comment 21

[Henri Sivonen (:hsivonen)]

The spec says:
A start tag whose tag name is "select"

    Parse error. Act as if the token had been an end tag with the tag name "select" instead.

Thus, when the hour menu is missing an end tag, what the author intended to be the start tag of the minute menu becomes the end tag of the hour menu and the options for minutes end up outside a select. Chrome and Opera Ragnarök agree. DOM-wise, but default to treating the selectless options as display: none;. IE6, IE8 and IE9 (in whatever mode the test case triggers) agree as well on the point of how select closes but then ignore the selectless option tags (which leads to the content of the options showing on the page).

(In reply to comment #0)
> This multiple drop-down application works as intended on various forums. It
> works with previous Firefox versions, Internet Explorer, Safari and Chrome.

I believe the above statement isn't accurate about Chrome and IE. (I didn't test Safari or an old version of Firefox.)

This is not a bug in the parser. bz, this might be considered to be showing a UA style sheet or frame constructor bug (since the selectness options don't behave as display: none;), even though I can't cite a relevant normative sentence from the spec.

Comment 22

[Boris Zbarsky [:bzbarsky]]

There is no such normative sentence.  The spec is silent on the subject of <option> outside <select>, so we don't do any magic for it.