493589 - about:config rendering problem

Status: VERIFIED FIXED

(Firefox for Android Graveyard :: General, defect)

Description

[Doug Turner (:dougt)]

Attachment: screenshot

load m.cnn.com
load about:config
search for "network" (probably anything really)
pan to end of page
notice that you see m.cnn.com at the bottom of the page.

Comment 1

[Vivien Nicolas (:vingtetun) (:21) - (NOT reading bugmails, needinfo? please)]

Attachment: Patch to correct the rendering for xul pages

This bug probably affect many xul document. In fact if the document is not an HTML one or a SVG one we keep as size the actual canvas size (CanvasBrowser.js#660).
This patch handle XUL document in .

Comment 2

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

Hmm, we got rid of this code in bug 458740, since bug 111034 seemed to make it unnecessary. If we really do need it, we should just add it as an || to the existing condition and keep the two code paths in common, but I'd like to know why scrollHeight/Width is not accurate here.

Comment 3

[Vivien Nicolas (:vingtetun) (:21) - (NOT reading bugmails, needinfo? please)]

(In reply to comment #2)
> Hmm, we got rid of this code in bug 458740, since bug 111034 seemed to make it
> unnecessary. If we really do need it, we should just add it as an || to the
> existing condition and keep the two code paths in common, but I'd like to know
> why scrollHeight/Width is not accurate here.

The problem here is that cdoc.body did not exist for XULDocument and so Math.max return NaN as the max value and this force to return the canvas size, but otherwise you're right with bug 111034 scrollWidth/Height is completely accurate for XUL doc and I've not see that.

Comment 4

[Vivien Nicolas (:vingtetun) (:21) - (NOT reading bugmails, needinfo? please)]

Attachment: Updated patch with take care of bug 111034

Comment 5

[Mark Finkle (:mfinkle) (use needinfo?)]

Attachment: simple patch to stop NaN from bubbling out of Math.max

I think this is more like Gavin had in mind. If Math.max() propagates the Nan, then let's set a reasonable lower limit.

Comment 6

[Vivien Nicolas (:vingtetun) (:21) - (NOT reading bugmails, needinfo? please)]

(In reply to comment #5)
> I think this is more like Gavin had in mind. If Math.max() propagates the Nan,
> then let's set a reasonable lower limit.

Yes, this is what I've understand but I just had the feeling it was a kind of workaround to do it that way (let a variable called body and do some Math.max when there is no reason to do it).
If we do it that way can we just remove the two isNan from the last test?

Comment 7

[Vivien Nicolas (:vingtetun) (:21) - (NOT reading bugmails, needinfo? please)]

After thinking a bit, I've just realized SVGDocument is the exception and you don't want to had exception, right?

Comment 8

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

I think we need the SVGDocument exception still. I think Mark's patch is probably right, though I'd probably just add "|| 0" to the Math.max parameters rather than specifying a dummy object.

Comment 9

[Mark Finkle (:mfinkle) (use needinfo?)]

http://hg.mozilla.org/mobile-browser/rev/88b57569da6b

Comment 11

[Al Billings [:abillings - ex-MoCo]]

I find it annoying when we dupe the original bug to the one that came weeks later. :-P

Comment 12

[Al Billings [:abillings - ex-MoCo]]

Now I cannot scroll about:config at all on OS X. Should I open a new bug?

Comment 13

[Aakash Desai [:aakashd]]

verified FIXED:

Mozilla/5.0 (Windows; U; WindowsCE 5.2; en-US; rv:1.9.3a1pre) Gecko/20090821 Fennec/1.0a3pre