Bugzilla Bug 213365

Compilation error for net-news/liferea USE=xulrunner (same error for all
versions, 1.2.23 to 1.4.13 -- 1.0.24's ebuild perm disables xulrunner):

In file included from mozsupport.cpp:36:
/usr/include/xulrunner-1.9/unstable/gtkmozembed.h:64:1: warning: this is the
location of the previous definition
In file included from /usr/include/xulrunner-1.9/unstable/nsrootidl.h:11,
                 from /usr/include/xulrunner-1.9/unstable/nsISupports.h:10,
                 from /usr/include/xulrunner-1.9/unstable/nsIWebBrowser.h:10,
                 from
/usr/include/xulrunner-1.9/unstable/gtkmozembed_internal.h:41,
                 from mozsupport.cpp:37:
/usr/include/xulrunner-1.9/unstable/nscore.h:208:1: warning: "NS_EXPORT_"
redefined
In file included from mozsupport.cpp:36:
/usr/include/xulrunner-1.9/unstable/gtkmozembed.h:63:1: warning: this is the
location of the previous definition
mozsupport.cpp:41:32: error: dom/nsIDOMKeyEvent.h: No such file or directory
mozsupport.cpp:46:28: error: necko/nsNetCID.h: No such file or directory
mozsupport.cpp: In function 'gint mozsupport_key_press_cb(GtkWidget*, void*)':
mozsupport.cpp:56: error: 'nsIDOMKeyEvent' was not declared in this scope
mozsupport.cpp:56: error: 'event' was not declared in this scope
mozsupport.cpp:56: error: expected primary-expression before ')' token
mozsupport.cpp:56: error: expected `;' before 'ev'
mozsupport.cpp:67: error: 'nsIDOMKeyEvent' is not a class or namespace
mozsupport.cpp: In function 'void mozsupport_set_offline_mode(gboolean)':
mozsupport.cpp:281: error: 'NS_IOSERVICE_CONTRACTID' was not declared in this
scope
make[4]: *** [liblihtmlm_la-mozsupport.lo] Error 1
make[4]: Leaving directory
`/var/tmp/portage/net-news/liferea-1.4.13/work/liferea-1.4.13/src/mozilla'
make[3]: *** [all-recursive] Error 1
make[3]: Leaving directory
`/var/tmp/portage/net-news/liferea-1.4.13/work/liferea-1.4.13/src'
make[2]: *** [all] Error 2
make[2]: Leaving directory
`/var/tmp/portage/net-news/liferea-1.4.13/work/liferea-1.4.13/src'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory
`/var/tmp/portage/net-news/liferea-1.4.13/work/liferea-1.4.13'
make: *** [all] Error 2
 * 
 * ERROR: net-news/liferea-1.4.13 failed.
 * Call stack:
 *               ebuild.sh, line   49:  Called src_compile
 *             environment, line  673:  Called gnome2_src_compile
 *             environment, line  576:  Called die
 * The specific snippet of code:
 *       emake || die "compile failure"
 *  The die message:
 *   compile failure
 * 
 * If you need support, post the topmost build error, and the call stack if
relevant.
 * A complete build log is located at
'/var/tmp/portage/net-news/liferea-1.4.13/temp/build.log'.
 * The ebuild environment file is located at
'/var/tmp/portage/net-news/liferea-1.4.13/temp/environment'.
 *

Reassigning to correct maintainer per reporters request

This is a known issue; upstream hasn't ported to xulrunner 1.9 yet.  This one
will have to wait for them, I believe.  If it starts to be a problem (ie,
xulrunner 1.9 is in the tree), I'll see about doing the port myself.

Ubuntu is already applying:
http://patches.ubuntu.com/l/liferea/extracted/xul1.9.code.dpatch
http://patches.ubuntu.com/l/liferea/extracted/xul1.9.dpatch

Maybe they could help

Created an attachment (id=146588) [details]
this ebuild fix issue for me

Current ebuild doesn't disable gecko module if xulrunner enabled.
I have been fix ebuild for myself, please look.

That's unrelated to this bug.  Please open a new bug for it.

liferea-1.4.13 tried to build against xulrunner-1.9 if both 1.8 and 1.9 were
installed. This was due to a check for using (tree-cleaned) www-client/mozilla
as a gecko backend which gave a false positive with xulrunner-1.9

liferea-1.4.14 reordered the logic completely (bug #217003), and the problem is
now gone since gecko is explicitly disabled when xulrunner is enabled. Hence
www-client/mozilla is never checked for, and there is no false positive.

Just FYI that bug #217003 was two birds with one stone ;)

I'd like to leave this open for xulrunner 1.9 support.

I can confirm this error on liferea-1.4.13 on amd64. But 1.4.15 compiles good
and runs without errors... Should it be stabilized for amd64? is there any
other limitant?

(In reply to comment #8)
> I can confirm this error on liferea-1.4.13 on amd64. But 1.4.15 compiles good
> and runs without errors... Should it be stabilized for amd64? is there any
> other limitant?
> 

(In reply to comment #8)
> I can confirm this error on liferea-1.4.13 on amd64. But 1.4.15 compiles good
> and runs without errors... Should it be stabilized for amd64? is there any
> other limitant?
> 

According to this changelog. 1.4.16 fixes this bug.

"What's New in 1.4.16 Stable Release:

· Fixed wrong language code of Czech translation. (reported by Martin Picek)
· Fixes SF #1920860: crash after memory corruption when parsing feed with
invalid encoding (reported by mumrel)
· Build support for XulRunner 1.9 (patch from Alexander Sack)
· Significant speed improvement for the cache conversion necessary when
upgrading from 1.4.9. Fixes 100% CPU during startup problem.
· Fixes comment removal bug.
· Drops all comments from cache to cleanup stale comments.
· Changes DB schema to better handle comments during item and subscription
deletion. This makes the cache version incompatible to 1.4.15!
"

1.4.15 is going stable (bug #228357)  It will use only xulrunner 1.8.  The plan
is to put in  1.4.16 using only xulrunner 1.9 (and maybe optionally webkit, if
I can get it to work well enough). It will go stable with xulrunner 1.9 some
time in july.

1.4.16b and 1.4.17 build against xulrunner 1.9