86259 - widget/src/xlib/nsAppShell.cpp uses non-standard for init scope

Status: VERIFIED FIXED

(Core :: XUL, defect)

Description

[John Mark Vandenberg]

widget/src/xlib/nsAppShell.cpp uses c++ coding standard regarding for
initialisation variable scope which differs from rest of Mozilla source.

MIPSpro is farily particular about ansi for initialisation scope ... and the 
rest of the Mozilla code base compiles.  This is the first piece of code which 
has broken on this.  

http://www.mozilla.org/hacking/portable-cpp.html#variables_in_for

Comment 1

[John Mark Vandenberg]

Attachment: widget/src/xlib/nsAppShell.cpp : Patch to use ansi standard for init scope

Comment 2

[cls]

Bouncing to XP Toolkit/Widgets.

Comment 3

[John Morrison]

I'm not sure exactly who owns the xlib code these days, but that code is 
cvs blame to timeless. It's a trivial fix, so it should just be checked in.
Heck, I think even I can say r=jrgm :-]. blizzard, sr=?

Comment 4

[timeless]

Attachment: change second usage from i to j.

Comment 5

[timeless]

yes I actually own that code, attached is a counter patch (we should talk via 
email since sparing in bugzilla is a waste of public bits). if it works i'd 
prefer to use it. but i think i need to talk to shaver and brendan and people 
about this specific compiler because I really don't like its scoping features.

Comment 6

[John Morrison]

Well, r=jrgm for whichever assuming the latter also works under MIPSpro.

But, by the way, that redeclaration of 'int i' would also fail with VC++.

Comment 7

[John Mark Vandenberg]

timeless,
The compiler is not the problem, as MIPSpro can allow this non-ansi code with -
LANG:ansi-for-init-scope=OFF

However the ansi standard has been revised to make this code illegal, and the 
mozilla coding standard follows this.  The entire mozilla repository follows 
this coding standard, excepting this bug and bug 86036 , and possibly some 
other hard to find code that is not built in a tinderbox .

I agree sparing is useless, both patches make the code follow the standard ... 
lets just get one of them in.

Comment 8

[tor]

sr=tor

Comment 9

[Asa Dotzler [:asa]]

a= [e-mail] for checkin to the trunk.
(on behalf of drivers)

Comment 10

[timeless]

fix checked in.

Comment 11

[Chad Austin]

verified checked in