279257 - Some of the designMode frames are not loaded when there are at least 2 frames and there are linked style sheets in at least one of them

Status: RESOLVED WORKSFORME

(Core :: DOM: Editor, defect)

Description

[Rotem Liss]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.5) Gecko/20041115 Firefox/1.0
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.5) Gecko/20041115 Firefox/1.0

When using designMode function, I tried to load 2 designMode areas. I wrote into
the frames a link to a style sheet. The first area wasn't a designMode area, but
a normal frame.

When I tried to use 3 designMode areas, the 2 firsts weren't loaded, but the
last was loaded.

When I tried not to use the linked style sheets, it has worked properly. When I
used a linked style sheet even in one frame, the firsts weren't loaded again.

Reproducible: Always

Steps to Reproduce:
1. Build a style sheet for the testing, and save it in a ".css" file.
2. Build a page with 2 (or more) designMode areas.
3. Write the "<style>" tag with "@import" to the style sheet, for one of the
designMode areas.
4. Load the page in your browser.

Actual Results:  
The first designMode area is as a normal frame - the cursor is not there, and I
cannot write in.

Expected Results:  
The two frames have a cursor in (when clicking on), and I can write in any of them.

Comment 1

[Prognathous]

(In reply to comment #0)
> Steps to Reproduce:
> 1. Build a style sheet for the testing, and save it in a ".css" file.
> 2. Build a page with 2 (or more) designMode areas.
...

How about attaching a minimal testcase, or linking to one?

Prog.

Comment 2

[Rotem Liss]

Attachment: Minimal Testcase

> How about attaching a minimal testcase, or linking to one?
> Prog.

I hadn't time, but now I have created one.

Comment 3

[Rotem Liss]

Attachment: A testcase, also has a "bold" button

I created a better testcase, which has a "bold" button, for you will see the
exception message of execCommand.

Comment 4

[Boris Zbarsky [:bzbarsky]]

So... that's not a testcase.  That's an application.

For example, why do you do the weird timeout thing before setting designMode?

It sounds like you're just ending up setting designMode on a document that's
still loading....

Comment 5

[Rotem Liss]

(In reply to comment #4)
> So... that's not a testcase.  That's an application.
> 
> For example, why do you do the weird timeout thing before setting designMode?
> 
> It sounds like you're just ending up setting designMode on a document that's
> still loading....

Well... I've tried to create an editor like this, then I saw it doesn't work -
so I inserted the timeout checking and it worked fine.

Yes, it is a part of an application, but you can see the bug also in the Minimal
Testcase, so it is not something I caused to in the "application".

Comment 6

[Boris Zbarsky [:bzbarsky]]

Again, the point is that you're setting designMode on a document that's still
loading...

Comment 7

[Rotem Liss]

The code is at(In reply to comment #6)
> Again, the point is that you're setting designMode on a document that's still
> loading...

The code is after the page, so it should being loaded after the page. What's
wrong with it?

By the way, can someone confirm it?

Comment 8

[Boris Zbarsky [:bzbarsky]]

The stylesheet takes time to load, no?

Comment 9

[Rotem Liss]

Again: can anyone confirm it with the testcases?

Comment 10

[Boris Zbarsky [:bzbarsky]]

> Again: can anyone confirm it with the testcases?

I cannot.  For me, I cannot write in either iframe, and both buttons throw the
same exceptions when clicked.  Hence the questions, given that the testcase is
not showing the bug it claims to be showing, and given that I can't figure out
exactly what the testcase is trying to do and why.

Comment 11

[Rotem Liss]

Attachment: Minimal Testcase - loading in "load" event of window

Trying loading the frames in "load" event of the window.

Comment 12

[Rotem Liss]

Attachment: The testcase also has a "bold" button - loading in "load" event of the window

Comment 13

[Rotem Liss]

For me, the second set of testcases works exactly as the first - the first frame
is not loaded, the second is loaded. Can anyone confirm?

Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.7) Gecko/20050415
Firefox/1.0.3

Comment 14

[Asaf Romano (gone)]

Rotem, can you retest it using a trunk build? 1.7.x is ancient.

Comment 15

[Asaf Romano (gone)]

Comment on attachment 182963 [details]
The testcase also has a "bold" button - loading in "load" event of the window

WFM

Comment 16

[Rotem Liss]

Mmm... all of the testcases work for me in the Trunk release:

Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8b2) Gecko/20050516
Firefox/1.0+

Should I close the bug as Worksframe? Or should someone else do that?

Comment 17

[Boris Zbarsky [:bzbarsky]]

Either way works.

Comment 18

[Rotem Liss]

It indeed works in Firefox 1.5.