As of gaim-2.0.0beta4 gaim offers a CLI in addition to the standard gtk+ interface. Maybe useflags should be added to enable the user to choose which interface(s) he prefers to be compiled.
*** Bug 152201 has been marked as a duplicate of this bug. ***
(In reply to comment #1) > *** Bug 152201 has been marked as a duplicate of this bug. *** > argh sry for double posting
This will happen later it requires major changes in the ebuild to make it practical.
Well, we already have gtk USE flag, so we can utilize this whether to build gaim w/ gtk support or just text-version of gaim.
Created attachment 102206 [details, diff] gaim-newflags.patch What about this patch, I've been using it for a while now with my gaim svn ebuilds. Summary: - ncurses new use flag, new dep, enables consoleui, warns if ncurses is w/o unicode - gtk new use flag, conditional dep, enables gtkui - spell, xscreensaver, eds, startup-notification make these DEPENDs conditional on gtk - session-management new use flag, new dep (conditional on gtk), enables "sm" configure switch Unrelated to this bug: - sasl new use flag, new dep, enables cyrus-sasl support - nas remove the use flag, not supported it seems
Thank you, should be an adjust -r1 available shortly :)
Surely it'd make sense to fall back to the gtk version if neither USE=gtk nor console is specified (in the interests of having sensible defaults and non-interactivity): this is what gaim's always been in the past, and it's what most people will expect (and probably what they want)... It's kind of annoying when an upgrade fails for this reason *shrug*
(In reply to comment #7) > Surely it'd make sense to fall back to the gtk version if neither USE=gtk nor > console is specified (in the interests of having sensible defaults and > non-interactivity): this is what gaim's always been in the past, and it's what > most people will expect (and probably what they want)... I'd mostly agree with what you're saying, but regardless, it is reasonable to assume that there are people who are compiling something that is not in portage but that utilizes the gaim library '(and nothing more than the gaim library, no gtk interface nor ncurses such) to communicate with Yahoo! ICQ etc. peers. So, my humble vote would be to let users shoot themselves in the feet as much as they like and not be over generous with the band aid.
Dieing based on USE flag combinations is not allowed. If one interface is required but neither flag is specified, please pick a default one and fall back to that.
2.0.0_beta5 works here, 2.0.0_beta5-r1 stays forever without opening a socket of showing any user interface. seems to build all right. Any clue? I'm using amd64
(In reply to comment #7) > Surely it'd make sense to fall back to the gtk version if neither USE=gtk nor > console is specified (in the interests of having sensible defaults and > non-interactivity): this is what gaim's always been in the past, and it's what > most people will expect (and probably what they want)... > > It's kind of annoying when an upgrade fails for this reason *shrug* > console is default now ... so your emerge will not die if neither console or gtk use flag are set.
By the way, shouldn't the ncurses dependency be explicitly listed? ncurses is not in any system profile after all. It is true that bash pulls ncurses, which is in turn pulled by python, but hey, maybe there will be a time when portage is not required.
Is there any specific reason why the GUI is enabled by an useflag named after the library (gtk) but the console interface is enabled by an useflag named after the functionality (console) and not the library (ncurses)?