717969 - Use "browser.chromeURL" preference in Core tests, to support non-Firefox applications

Status: RESOLVED FIXED

(Core :: General, defect)

Description

[Serge Gautherie (:sgautherie)]

"Found 19 matching lines in 14 files"

There should be no reason to use "Firefox" URI in Core tests.
Document/fix them.

See bug 717753 as an example.

NB: I haven't checked (yet) how these tests behave in SeaMonkey, but I'm wondering...

Comment 1

[Serge Gautherie (:sgautherie)]

"Found 14 matching lines in 11 files"

Comment 2

[saran]

Hi I am new here and would like to help out. Could you please guide me on this ?

Comment 3

[Serge Gautherie (:sgautherie)]

(In reply to saran from comment #2)
> Could you please guide me on this ?

What do you want to know more precisely ?

Comment 4

[saran]

If I am right , does a check condition need to be added to check if the browser is firefox or seamonkey and set the URI accordingly ? like in  bug 717753 ?

Comment 5

[saran]

So should I add something like this ?

 var browserUri = "chrome://browser/content/browser.xul";//firefox by deafult

//check it is seamonkey and use the correct uri 

if (navigator.userAgent.match(/ SeaMonkey\//))
  browserUri = "chrome://navigator/content/navigator.xul";

And wherever the browser URI is hardcoded , I use the variable browserUri so it is not firefox by default always ?

Comment 6

[Abhishek Potnis [:avp]]

Hi,

I am interested in working on this bug, but seems like Serge has not been around for some time(I checked on IRC), can someone please mentor this bug instead?

Thanks.

Comment 7

[Mike Hoye [:mhoye]]

Abhishek, are you still interested in pursuing this?

Comment 8

[Abhishek Potnis [:avp]]

Hi Mike,

Yeah, I would like to work on this! How do I get started?

Comment 9

[Mike Hoye [:mhoye]]

I see you've got a patch in already, so your dev environment is already set up. Serge, can we get a DXR link for Abhishek to get him started?

Comment 10

[Abhishek Potnis [:avp]]

The default value of the preference "Browser.chromeURL" is set to "chrome://browser/content/". I am guessing, to fix this bug - to support non-Fx applications, I need to make use of that preference instead of having the entire URI entered as a string everytime.

The link in the URL field of the bug, gave me the occurences of "chrome://browser/content/browser.xul" in tests folders. But the occurences have a browser.xul appended to it. So if I change it to something like Browser.chromeURL+"/browser.xul", it will still be only for Firefox.

How do I proceed further? Please guide me.

Thanks.

Comment 11

[Serge Gautherie (:sgautherie)]

(In reply to saran from comment #4)
> If I am right , does a check condition need to be added to check if the
> browser is firefox or seamonkey and set the URI accordingly ? like in  bug
> 717753 ?

Ftr, bug 717753 used 'browser.chromeURL' instead of an 'if', eventually:
see
https://hg.mozilla.org/mozilla-central/rev/1a94821b4655


(In reply to Mike Hoye [:mhoye] from comment #9)
> Serge, can we get a DXR link for Abhishek to get him started?

(I'm not familiar with DXR. Maybe someone else can.)


(In reply to Abhishek Potnis [:avp] from comment #10)
> The link in the URL field of the bug, gave me the occurences of
> "chrome://browser/content/browser.xul" in tests folders. But the occurences
> have a browser.xul appended to it. So if I change it to something like
> Browser.chromeURL+"/browser.xul", it will still be only for Firefox.

Iirc, the "browser.xul" part is optionnal.
I suggest that you try and check whether tests still work fine without it.

Ftr, see
http://kb.mozillazine.org/Browser.chromeURL

Comment 12

[Patrick]

can someone please provide the file path for this bug?

Comment 13

[Mark Banner (:standard8)]

I think given the current state of SeaMonkey & Thunderbird, it'll be hard to know if this is actually worth fixing. Also, as Serge is not really around know, that makes knowing the value harder. I'd rather specific bugs be filed where they are still having issues.

I'll marked this as "fixed" as some bugs were previously fixed (see the dependent list).