545343 - Port |Bug 526451 - allow to build gconf and gnomevfs/gio support independently| to comm-central

Status: RESOLVED FIXED

(MailNews Core :: Build Config, defect)

Description

[Serge Gautherie (:sgautherie)]

I'm loath to add ifdef everywhere.

Comment 1

[Serge Gautherie (:sgautherie)]

Attachment: (Av1) Port (the useful part of) it, Remove some unused vars

Comment 2

[Justin Wood (:Callek)]

Comment on attachment 444883 [details] [diff] [review]
(Av1) Port (the useful part of) it, Remove some unused vars

>From: Serge Gautherie <[e-mail]>
>             AC_DEFINE(MOZ_ENABLE_GNOMEVFS)
>         ],[
>-            if test "$MOZ_ENABLE_GNOMEVFS" = "force"
>-            then
>-                AC_MSG_ERROR([* * * Could not find gnome-vfs-module-2.0 >= $GNOMEVFS_VERSION])
>-            fi
>             MOZ_ENABLE_GNOMEVFS=
>         ])

Did not do a full review yet; but m-c still has this part, why are we removing it?

Also it does have some slight bitrot (purely with context due to 1.9.2 ifdefs) Can you please explain this part and attach a new patch; I'll be faster with review.

Comment 3

[Serge Gautherie (:sgautherie)]

Attachment: (Av2) Port (the useful part of) it, Remove some unused vars

Av1, with comment 2 suggestion(s).

(In reply to comment #2)
> m-c still has this part, why are we removing it?

Because of bug 513709...

Comment 4

[Justin Wood (:Callek)]

(In reply to comment #3)
> (In reply to comment #2)
> > m-c still has this part, why are we removing it?
> 
> Because of bug 513709...

Still not clear here, there is a lot of tangential activity on Bug 513709. Can you please be more explicit and descriptive on the why?

Holding off on full review pending that answer. [sorry for delayed reply as well]

Comment 5

[Serge Gautherie (:sgautherie)]

Attachment: (Av2a) Port (the useful part of) it, Remove some unused vars

Av2, simply (context) unbitrotted.


(In reply to comment #2)
> >-            if test "$MOZ_ENABLE_GNOMEVFS" = "force"
> 
> m-c still has this part, why are we removing it?

I think erroring out in m-c should be enough in this case.

Comment 6

[Justin Wood (:Callek)]

(In reply to comment #5)
> (In reply to comment #2)
> > >-            if test "$MOZ_ENABLE_GNOMEVFS" = "force"
> > 
> > m-c still has this part, why are we removing it?
> 
> I think erroring out in m-c should be enough in this case.

Short term I want to error on the side of caution with us removing things. I do have a general target in mind about mass-removing lots of this stuff, in favor of m-c. But with the LARGE number of changes between m-c and c-c atm here I don't want to just remove stuff "because we don't think we'll need/want it"

If you heavily disagree you can get Standard8 or KaiRo to review this for you if you want, since they likely know much more about GCONF than I do, but I doubt they will disagree.

Comment 7

[Serge Gautherie (:sgautherie)]

Attachment: (Av3) Port (the useful part of) it, Remove some unused vars [Checked in: Comment 10]

Av2a, with comment 2 suggestion(s).

Comment 8

[Justin Wood (:Callek)]

Comment on attachment 492142 [details] [diff] [review]
(Av3) Port (the useful part of) it, Remove some unused vars
[Checked in: Comment 10]

Either I don't get this well. or my mind is just not in sorts right. Either way, I'm retargetting at KaiRo for now.

Comment 9

[Robert Kaiser]

Comment on attachment 492142 [details] [diff] [review]
(Av3) Port (the useful part of) it, Remove some unused vars
[Checked in: Comment 10]

r=me, though I still wonder which of these we are really using.

Just FYI, because it was discussed in this bug, if we are actually using a var, then I'm all for having the check and error condition on our side. Where we are not using it in c-c, we probably should not even define it (and that's what I'm still wondering about a bit here).

Comment 10

[Serge Gautherie (:sgautherie)]

Comment on attachment 492142 [details] [diff] [review]
(Av3) Port (the useful part of) it, Remove some unused vars
[Checked in: Comment 10]

http://hg.mozilla.org/comm-central/rev/ce7efebfdbde


(In reply to comment #9)
> I still wonder which of these we are really using.

c-c needs (all) this for packaging (only):
{
http://mxr.mozilla.org/comm-1.9.2/search?string=MOZ_ENABLE_GNOME_COMPONENT&case=1&find=%2Finstaller%2Fpackage-manifest%5C.in%24
/mail/installer/package-manifest.in
    * line 677 -- #ifdef MOZ_ENABLE_GNOME_COMPONENT
/suite/installer/package-manifest.in
    * line 382 -- #ifdef MOZ_ENABLE_GNOME_COMPONENT
}