* Add support for installing the new Firefox brand icon by defining
WITH_NEW_ICON [1]
For all that's new, check out http://www.mozilla.org/products/firefox/releases/
PR: 71781 [1]
Submitted by: Radek Kozlowski <radek@raadradd.com> [1]
This is being done instead of the update to 1.0 PR 1 since we're in a ports
freeze, and too many big changes is not a good idea.
This update covers the following Mozilla bugs:
250862
255067
256316
Thanks to nectar for scraping all of these patches together.
Obtained from: Mozilla CVS
Approved by: portmgr (implicit)
* Fix a problem with the pkg-config files referencing mozilla components
and not firefox components
PR: 68221 [1]
Submitted by: Tom McLaughlin <tmclaugh@sdf.lonestar.org> [1]
to fix some threading issues in Firefox, but I have not personally encountered
them (at least not ones fixed by this patch).
See http://bugzilla.mozilla.org/show_bug.cgi?id=247204 for more details.
Submitted by: green
* Make sure optimized CFLAGS are actually respected [1]
* Frobnicate the mozconfig file a little bit to be more inline with typical
Linux builds
* Try a safer hack to ensure the install completes okay
Reported by: kris [1]
list of Known Issues, see:
http://www.mozilla.org/products/firefox/releases/0.9.html
This update also includes a new pkg-install script based on ports/65590
from Thierry Thomas <thierry@pompo.net>. However...
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!WARNING!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
After installing Firefox, you must su - to root, and run ``firefox'' before
running it as any other user. If you do not su - to root, you may lose some
of your user settings. If you try to run Firefox as a normal user without
first running it as root, the browser will not show up. This is a known issue
with Firefox, and they hope to address it in a future release.
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!WARNING!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
properly detect libiconv support [1]
* Take care of a few more pthread substitutions
* Bump PORTREVISION
Reported by: Andrew Muhametshin <andrew@trifon.ru>
worth)
* Fix up some path issues in the Makefile
* Minor fix to plug-in installation
* Restore Franz's fix for the LD_LIBRARY_PATH PREFIX problem
* Bump PORTREVISION
will be tried. There is one addition, though. JPI_LIST is now
overrideable. [1]
* Remove the bit about Perl version from pkg-message. We support all versions
of Perl now.
Requested by: ale
release of the Firebird web browser with a new name to avoid conflicts with
the Firebird database. This is version 0.8.
http://www.mozilla.org/products/firefox/releases/
* Update to 0.8 and change name to firefox [1]
* Make PLIST from a fake installation [1]
* Restore original credit and portname in Makefile [1]
* Optionally install headers and IDL files [2]
* Use PERL instead of REINPLACE_CMD in a few places [1]
* Use OPTIONS
* Enable the Inspector extension
* Fix another possible esound related problem
* Enable basic optimizations if WITH_DEBUG is not specified
PR: 62631 [2]
Submitted by: trevor [1] [2]
do with esound being installed, but not being used (e.g. when not running
under the GNOME Desktop). What happens now is sound support is disabled by
default. If you wish to enable Type Ahead Find sound support, edit your
prefs.js file, and add:
user_pref("accessibility.typeaheadfind.enablesound", true);
* Make portlint happy by not including anything after bsd.port.post.mk
* Clean up the pkg-message a bit to reflect reality with respect to Perl
Thanks to casaveli on BSDForums for finding it was esound causing the lock up
as well as testing the patch mentioned above.
I've taken a chance and split the sparc64 and alpha mega-diffs up because
sparc64 patched the result of alpha's patches, and amd64 needed to do the
same again - but it was sorted lexically before alpha/sparc64. I've
grouped the xptcall patches together since they are standalone and split
up and merged the other infrastructure jumbo-patches.
I also attempted to sync up the nspr vs firebird diffs for sparc64 and
alpha because they disagreed. HOWEVER... While this port 'depends'
on the nspr port, it doesn't actually *use* it. It should probably
be removed from the LIB_DEPENDS list.
Note that the amd64 xptcall code came from SuSE via the mozilla bugzilla
database.
