On Mon, 30 Apr 2018 11:35:53 +0000 Mike Blumenkrantz <michael.blumenkra...@gmail.com> said:
> Hm it may be the case that g_main_loop_run apps are not using EFL events. I > will dig into this a bit more to confirm. That'd be good, but at least the data says it's not being used. > Regardless of this usage, however, I'm still not sure it's a good idea to > be dropping support for this type of integration just because a recent > refactor has broken it. Dropping support should probably be done over the > period of a couple releases, with deprecation in between, to ensure there > is no confusion and that there really are no users who will step in to > maintain this feature. Just FYI, we haven't done that before e.g. dropping PS3 support. The general question of "is anyone using this" was floated etc. etc. much like this thread. This is one of those features I suspect is just not used at all today. Given this thread so far I still think that is true. Looking at https://github.com/intel/wayland-fits/issues/6 shows it seemingly has had some bugs that cause hangs (well no fixes inside efl for that and the commit that seems to fix it there in wayland-fits is so huge that it doesn't look like a fix, more by luck and workaround, so not sure it is an efl bug or something else, but it kind of smells like it is a problem with efl's g-main-loop enabled). at least i smell there being bugs in that code path because of its ifdef compile time option, no testing worth talkling about nature and rare-if-ever need. i'm just saying that even if kept it's probably buggy and problematic anyway. > On Mon, Apr 30, 2018 at 6:50 AM Mike Blumenkrantz < > michael.blumenkra...@gmail.com> wrote: > > > Okay, in this case I have had direct confirmation within the past week > > that g_main_loop_run is indeed used for some EFL Tizen apps. It seems to me > > that the best option would be to continue maintaining this feature for now. > > > > On Sat, Apr 28, 2018 at 1:15 AM Carsten Haitzler <ras...@rasterman.com> > > wrote: > > > >> On Fri, 27 Apr 2018 18:37:41 +0000 Mike Blumenkrantz > >> <michael.blumenkra...@gmail.com> said: > >> > >> > --enable-g-main-loop is needed to use g_main_loop_run(), yes? > >> > >> yes. i covered that with the maemo case. using a gtk/g* based app with > >> the ecore > >> main loop running on top of that (which doesn't actually work well going > >> forward > >> given EFL_MAIN etc. as efl is now basically just wanting to be the > >> framework). > >> > >> glib support allows you to use all the glib infra (sources, timeouts > >> etc.) with > >> an ecore main loop without using the actual g main loop under it all, and > >> that is on by default. --with-glib=always does that and avoids needing to > >> call > >> an api call to enable this integration as it turns it on always. > >> > >> > On Fri, Apr 27, 2018 at 2:33 PM Carsten Haitzler <ras...@rasterman.com> > >> > wrote: > >> > > >> > > On Fri, 27 Apr 2018 15:32:49 +0000 Mike Blumenkrantz > >> > > <michael.blumenkra...@gmail.com> said: > >> > > > >> > > > I have actually used this in projects before, and I also know that > >> Tizen > >> > > > definitely uses it. I'm strongly opposed to any removals here, and > >> tests > >> > > > >> > > tizen absolutely does not use it. i checked the spec files. i checked > >> both > >> > > the > >> > > old ecore separated one and the latest efl merged one. what makes you > >> > > think it > >> > > does? here's the enable/disable configure options, none of which are > >> > > --enable-g-main-loop: > >> > > > >> > > %autogen \ > >> > > --disable-static \ > >> > > --disable-doc \ > >> > > --with-glib=always \ > >> > > --disable-xim \ > >> > > --disable-scim \ > >> > > --disable-neon \ > >> > > --disable-wayland-text-input \ > >> > > --disable-gesture \ > >> > > --with-tests=none \ > >> > > --enable-fb \ > >> > > --disable-tslib \ > >> > > %if %{with wayland} > >> > > --enable-ecore-wayland \ > >> > > --enable-wayland \ > >> > > --enable-egl \ > >> > > --with-opengl=es \ > >> > > --disable-rpath \ > >> > > --disable-ibus \ > >> > > --enable-tbm \ > >> > > %endif > >> > > %if %{with x} > >> > > --with-opengl=es \ > >> > > --disable-gesture \ > >> > > %else > >> > > --with-x11=none \ > >> > > --disable-rpath \ > >> > > %endif > >> > > --disable-physics \ > >> > > --disable-cxx-bindings \ > >> > > --enable-lua-old \ > >> > > --enable-ecore-buffer \ > >> > > --disable-gstreamer1 \ > >> > > --enable-harfbuzz \ > >> > > --enable-hyphen \ > >> > > --with-dictionaries-hyphen-dir=/usr/share/hyphen/ \ > >> > > --disable-cserve \ > >> > > --disable-poppler \ > >> > > --disable-spectre \ > >> > > --disable-librsvg \ > >> > > --disable-libraw \ > >> > > --disable-systemd \ > >> > > --disable-cserve \ > >> > > --with-elementary-base-dir="share/.elementary" \ > >> > > > >> > > > >> --enable-i-really-know-what-i-am-doing-and-that-this-will-probably-break-things-and-i-will-fix-them-myself-and-send-patches-abb > >> > > # --enable-systemd \ > >> > > # --enable-drm \ > >> > > # --enable-gl-drm \ > >> > > > >> > > so that makes me wonder if you are thinking of the right option. > >> there is > >> > > --enable-g-main-loop which is where ecore literally sits on top of > >> the glib > >> > > loop, and there is --with-glib=yes/no/always which is the glib > >> integration. > >> > > > >> > > > should likely be added to ensure that this continues to function as > >> > > > expected. > >> > > > > >> > > > To be clear, I am talking about both components of the integration: > >> both > >> > > > calling glib events from the ecore main loop and using > >> g_main_loop_run() > >> > > > for the overall main loop. > >> > > > >> > > i think not as if you say tizen definitely uses it... and it doesn't, > >> then > >> > > i > >> > > would think you are confused between the integration and literally > >> sitting > >> > > on > >> > > top of g main loop. > >> > > > >> > > so if you've used it, where have you used it and why? tizen surely > >> doesn't. > >> > > the reason the g-main-loop option existed actually was for maemo > >> which is > >> > > long > >> > > dead and buried to allow efl code to run within an existing gtk/glib > >> using > >> > > app > >> > > which is what that platform was centered around (and maemo never > >> shipped > >> > > efl > >> > > as a core requirement etc.). since tizen never used this then i know > >> of no > >> > > platforms that need or use this (as it has to be solved at the > >> platform > >> > > level > >> > > because it's a build option and not a standard efl api) or actual use > >> > > cases. > >> > > it's not an api or abi guarantee. it's a build feature e.g. like > >> > > supporting the > >> > > ps3 which we removed since it's now a "dead platform". > >> > > > >> > > > On Thu, Apr 26, 2018 at 4:19 AM Carsten Haitzler < > >> ras...@rasterman.com> > >> > > > wrote: > >> > > > > >> > > > > so we have integration for glib/main loop in regular efl which > >> allows > >> > > glib > >> > > > > stuff to work with efl, but --enable-g-main-loop basically puts > >> the efl > >> > > > > ecore > >> > > > > loop on top of the glib main loop. > >> > > > > > >> > > > > is anyone actually using this? is the regular glib loop > >> integration not > >> > > > > enough? > >> > > > > > >> > > > > -- > >> > > > > ------------- Codito, ergo sum - "I code, therefore I am" > >> > > -------------- > >> > > > > Carsten Haitzler - ras...@rasterman.com > >> > > > > > >> > > > > > >> > > > > > >> > > > > > >> > > > >> ------------------------------------------------------------------------------ > >> > > > > Check out the vibrant tech community on one of the world's most > >> > > > > engaging tech sites, Slashdot.org! http://sdm.link/slashdot > >> > > > > _______________________________________________ > >> > > > > enlightenment-devel mailing list > >> > > > > enlightenment-devel@lists.sourceforge.net > >> > > > > https://lists.sourceforge.net/lists/listinfo/enlightenment-devel > >> > > > > > >> > > > > >> > > > >> ------------------------------------------------------------------------------ > >> > > > Check out the vibrant tech community on one of the world's most > >> > > > engaging tech sites, Slashdot.org! http://sdm.link/slashdot > >> > > > _______________________________________________ > >> > > > enlightenment-devel mailing list > >> > > > enlightenment-devel@lists.sourceforge.net > >> > > > https://lists.sourceforge.net/lists/listinfo/enlightenment-devel > >> > > > > >> > > > >> > > > >> > > -- > >> > > ------------- Codito, ergo sum - "I code, therefore I am" > >> -------------- > >> > > Carsten Haitzler - ras...@rasterman.com > >> > > > >> > > > >> > >> > >> -- > >> ------------- Codito, ergo sum - "I code, therefore I am" -------------- > >> Carsten Haitzler - ras...@rasterman.com > >> > >> -- ------------- Codito, ergo sum - "I code, therefore I am" -------------- Carsten Haitzler - ras...@rasterman.com ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
