Re: [mer-general] [MER AND NEMO MIDDLEWARE] RFC: to merge updated and additional components into Mer from Nemo Middleware project
On Sunday, April 13, 2014 12:55:56 PM Carsten Munk wrote: https://github.com/nemomobile/libiodata.git Timed code, with a bit too generic name - group with Time Daemon https://github.com/nemomobile/libprofile-qt.git https://github.com/nemomobile/profiled.git I wrote the original N900 profiled and apart from the addition of the qt api there have been no major changes since then - I could take care of these two - Simo Piiroinen (spiiroin)
Re: [mer-general] [MER AND NEMO MIDDLEWARE] RFC: to merge updated and additional components into Mer from Nemo Middleware project
Hello, On 13/04/2014 12:55, Carsten Munk wrote: * Updated Systemd to 208 (Primary maintainer, Pekka Lundstroem) http://github.com/mer-packages/systemd https://github.com/mer-packages/kmod.git https://github.com/nemomobile/systemd-user-session-targets.git (new) I don't fully agree with this point, this would increase minimum kernel version and it would make hard to boot on 2.6.35 - 3.0 kernels. 2.6.37 is used on several OMAP based devices while 2.6.35 is used on several Freescale i.MX based devices and backporting kernel features is not easy at all. Bye, Davide Bettio,
Re: [mer-general] [MER AND NEMO MIDDLEWARE] RFC: to merge updated and additional components into Mer from Nemo Middleware project
On 04/13/2014 01:55 PM, Carsten Munk wrote: ... * Updated Systemd to 208 (Primary maintainer, Pekka Lundstroem) http://github.com/mer-packages/systemd https://github.com/mer-packages/kmod.git https://github.com/nemomobile/systemd-user-session-targets.git (new) Should be noted that this will drop support for hw adaptations that have kernel older than 3.0 by default. There might be way to backport some patches to the respective hw adaptation kernels to maintain those adaptations. * Project config enablers to make X11-less builds in Mer, which will be default off (i.e. X11 enabled) in Mer QA builds and general package updates/build system adjustments: Will this mean that libX* packages and dependencies will be dropped from mer core? https://github.com/mer-packages/boardname.git boardname should be dropped in favor of ssu and oneshot. There is already ongoing work to do this but not sure if it was finalized properly yet. ... Package Management daemon and bindings: https://github.com/nemomobile-packages/PackageKit-Qt.git https://github.com/nemomobile-packages/PackageKit.git https://github.com/nemomobile-packages/polkit-qt-1.git (Primary maintainer: Thomas Perl) Will we continue supporting zypper in the future? Synchronization framework: (not obligatory) https://github.com/nemomobile/buteo-mtp.git https://github.com/nemomobile/buteo-sync-plugins-google-simple.git https://github.com/nemomobile/buteo-sync-plugins-google.git https://github.com/nemomobile/buteo-sync-plugins.git https://github.com/nemomobile/buteo-syncfw.git https://github.com/nemomobile/buteo-syncml.git I would say we should get these in as synchronization is one of the main features in any mobile device. ... Primary maintainer: Thomas Perl (thp) Media indexing framework (optional): https://github.com/nemomobile/libqtsparql.git https://github.com/nemomobile-packages/tracker.git https://github.com/nemomobile/tumbler.git [and likely a boatload of the dependencies below] I think most of the deps are already listed after the rest of the things would be merged. I would propose that we evaluate this after everything else is merged to see how much it actually takes as maintenance load. In the end I don't think it is much if we just limit it to media indexing. ... Additional packages for discussion if they're alive or not and if other packages above depend on them and if they are not more part of UI. From this list I would mainly pick the ones that are dependencies for everything else. Couple of comments though: ... https://github.com/nemomobile/profiled.git Maybe this should be taken as part of mer? ... https://github.com/nemomobile/sd-utils.git We are missing sd card and mass storage handling in general from Mer. ... If you claim meritocratic maintainership to any of these packages, please let me know and we'll set up access rights for merging and review rights accordingily after 1st of May 2014 if proposal holds. I can take some maintainership of some of the generic packages that I've done general updates in the past as well. But lets see what is left, we should probably collect wiki page about all the packages with primary maintainer listed. Protesting to the architectural and location move of your component is perfectly valid to do, as well. I believe from an architectural and project point of view that this move will strengthen Mer and related Nemo middleware components as a open source mobile platform in it's own right and make contribution and participation easier to understand, explain and do. With best regards, Carsten Munk Mer project architect Very nicely put. Thank you for the effort. Best Regards, Marko Sage Saukko One of the Mer maintainers that just stopped having enough free time :/
Re: [mer-general] [MER AND NEMO MIDDLEWARE] RFC: to merge updated and additional components into Mer from Nemo Middleware project
2014-04-13 14:14 GMT+02:00 Marko Saukko marko.sau...@gmail.com: On 04/13/2014 01:55 PM, Carsten Munk wrote: ... * Updated Systemd to 208 (Primary maintainer, Pekka Lundstroem) http://github.com/mer-packages/systemd https://github.com/mer-packages/kmod.git https://github.com/nemomobile/systemd-user-session-targets.git (new) Should be noted that this will drop support for hw adaptations that have kernel older than 3.0 by default. There might be way to backport some patches to the respective hw adaptation kernels to maintain those adaptations. * Project config enablers to make X11-less builds in Mer, which will be default off (i.e. X11 enabled) in Mer QA builds and general package updates/build system adjustments: Will this mean that libX* packages and dependencies will be dropped from mer core? Inverse, what I meant was that we add enablers for doing X11-less builds, but we don't drop X11 in default Mer builds. https://github.com/mer-packages/boardname.git boardname should be dropped in favor of ssu and oneshot. There is already ongoing work to do this but not sure if it was finalized properly yet. ... Package Management daemon and bindings: https://github.com/nemomobile-packages/PackageKit-Qt.git https://github.com/nemomobile-packages/PackageKit.git https://github.com/nemomobile-packages/polkit-qt-1.git (Primary maintainer: Thomas Perl) Will we continue supporting zypper in the future? Yes, zypper is a frontend to libzypp in practice. Synchronization framework: (not obligatory) https://github.com/nemomobile/buteo-mtp.git https://github.com/nemomobile/buteo-sync-plugins-google-simple.git https://github.com/nemomobile/buteo-sync-plugins-google.git https://github.com/nemomobile/buteo-sync-plugins.git https://github.com/nemomobile/buteo-syncfw.git https://github.com/nemomobile/buteo-syncml.git I would say we should get these in as synchronization is one of the main features in any mobile device. ... Primary maintainer: Thomas Perl (thp) Media indexing framework (optional): https://github.com/nemomobile/libqtsparql.git https://github.com/nemomobile-packages/tracker.git https://github.com/nemomobile/tumbler.git [and likely a boatload of the dependencies below] I think most of the deps are already listed after the rest of the things would be merged. I would propose that we evaluate this after everything else is merged to see how much it actually takes as maintenance load. In the end I don't think it is much if we just limit it to media indexing. ... Additional packages for discussion if they're alive or not and if other packages above depend on them and if they are not more part of UI. From this list I would mainly pick the ones that are dependencies for everything else. Couple of comments though: ... https://github.com/nemomobile/profiled.git Maybe this should be taken as part of mer? I'm unsure if it's even used at all. But I'd like to be proven wrong. ... https://github.com/nemomobile/sd-utils.git We are missing sd card and mass storage handling in general from Mer. ... If you claim meritocratic maintainership to any of these packages, please let me know and we'll set up access rights for merging and review rights accordingily after 1st of May 2014 if proposal holds. I can take some maintainership of some of the generic packages that I've done general updates in the past as well. But lets see what is left, we should probably collect wiki page about all the packages with primary maintainer listed. Protesting to the architectural and location move of your component is perfectly valid to do, as well. I believe from an architectural and project point of view that this move will strengthen Mer and related Nemo middleware components as a open source mobile platform in it's own right and make contribution and participation easier to understand, explain and do. With best regards, Carsten Munk Mer project architect Very nicely put. Thank you for the effort. Best Regards, Marko Sage Saukko One of the Mer maintainers that just stopped having enough free time :/
Re: [mer-general] [MER AND NEMO MIDDLEWARE] RFC: to merge updated and additional components into Mer from Nemo Middleware project
Carsten Munk carsten.m...@gmail.com writes: The mission and goals of Mer is amongst others: Mer is an open, mobile-optimised, core distribution aimed at device manufacturers; powered by Qt/QML and HTML5 - openly developed, inclusive, and meritocratically governed. Provide a mobile device oriented architecture Primary customers are device vendors - not end-users. Over time, the traditional mobile device focus has been done mostly as part of the Nemo Middleware project, with Mer focusing on base system, with a large overlap in contributor, vendor participation, use and contributions, even infrastructure. The barrier between Mer and Nemo Middleware and what goes where has often been confusing and let to big architectural challenges and difficulty in contributing to Mer and especially Nemo MW based projects, with differing methods, bugtrackers, requirements, etc - despite the overlap of people, bugtrackers, even if they were on the same infrastructure. In that sense, it makes sense that we try to remove the artifical gap between Mer and Nemo Middleware to ensure a productized and working, tested mobile core distribution. This is a RFC for Mer vendors, remaining advisory board members and maintainers, with idea of implementation happening after 1st of May 2014. We have already begun (a long time ago goal) to try and make Mer contribution easier, through setting up GitLab and migrating the core repositories from Gerrit to GitLab, hosted in Mer infrastructure, authenticated with Mer bugtracker accounts. In addition to that, we have set up a OBS _service based Mer build in https://build.merproject.org/project/monitor?project=mer-core%3Ai486%3Adevel and https://build.merproject.org/project/monitor?project=mer-core%3Adevel This uses the tar_git web service with option 'dumb' to utilize old Mer repository formats. Due to previous inability for Mer to accept contributions and release sufficiently fast for Mer-based products, a method was devised to make sure contribution could happen faster, by overlaying Mer with a faster-moving set of packages, based on git repositories where actual development happened. The lessons from that has been learnt and were part of the reason for the GitLab move and _service based mentioned in the previous three paragraphs. It led to bad situations before that the factual bleeding edge Mer was not in Mer Gerrit, but in Mer GitHub repositories. It is also worth noting that it is easy, and possible to only use a subset of Mer source packages in vendor builds for specific purposes. The original thought that Mer vendors would typically use full binary releases of Mer has been disproven and most uses are using their own Mer builds, done on own OBSes, and hence it makes more sense to see Mer as a set of software components that can be combined in different ways and built in a couple of different configurations for QA. The intended components to be updated and replacing current versions in Mer (currently Qt 5.0 based): * Mobile-productized Qt5.1 packaging and modifications from Nemo's Qt fork: (Primary maintainer, Robin Burchell/w00t) https://github.com/mer-packages/qtdocgallery https://github.com/mer-packages/qtwayland https://github.com/mer-packages/qtdeclarative https://github.com/mer-packages/qtsystems https://github.com/mer-packages/qtbase https://github.com/mer-packages/qtmultimedia https://github.com/mer-packages/qtlocation https://github.com/mer-packages/qtscenegraph-adaptation https://github.com/mer-packages/qtwebkit5 https://github.com/mer-packages/qttranslations https://github.com/mer-packages/qtjsbackend https://github.com/mer-packages/qtsvg https://github.com/mer-packages/qtquick1 https://github.com/mer-packages/qtsensors https://github.com/mer-packages/qtxmlpatterns [some may be missing] With updated 5.2 packages in a staging project while it gets productized, https://github.com/mer-qt * Updated ConnMan 1.23 (Primary maintainer, Lorn Potter/lpotter/ljp) http://github.com/mer-packages/connman https://github.com/mer-packages/crda.git https://github.com/mer-packages/wireless-regdb.git https://github.com/mer-packages/wpa_supplicant.git https://github.com/mer-packages/mobile-broadband-provider-info.git (APN list) https://github.com/nemomobile/libconnman-qt.git (new) https://github.com/nemomobile/connectionagent.git (new needed for UI enablement of ConnMan) * Updated PulseAudio 4.0 (Primary maintainer, Juho Hammilinen/jusa) http://github.com/mer-packages/pulseaudio https://github.com/mer-packages/sbc.git https://github.com/nemomobile/pulseaudio-modules-nemo.git (new) https://github.com/nemomobile-packages/libcanberra.git (new) * Updated Systemd to 208 (Primary maintainer, Pekka Lundstroem) http://github.com/mer-packages/systemd https://github.com/mer-packages/kmod.git https://github.com/nemomobile/systemd-user-session-targets.git (new) * Updated versions of Qt5 ports of
Re: [mer-general] [MER AND NEMO MIDDLEWARE] RFC: to merge updated and additional components into Mer from Nemo Middleware project
On 13/04/2014 8:55 pm, Carsten Munk wrote: oFono with ability to connect with Android RILD stack and bindings: https://github.com/nemomobile-packages/ofono (update) https://github.com/nemomobile/libqofono.git (new) https://github.com/nemomobile/tone-generator.git (new) I reckon these are up my alley, even though I only have worked on libqofono.
Re: [mer-general] [MER AND NEMO MIDDLEWARE] RFC: to merge updated and additional components into Mer from Nemo Middleware project
On 13/04/2014 8:55 pm, Carsten Munk wrote: oFono with ability to connect with Android RILD stack and bindings: https://github.com/nemomobile-packages/ofono (update) https://github.com/nemomobile/libqofono.git (new) https://github.com/nemomobile/tone-generator.git (new) I reckon these are up my alley, even though I've only worked on libqofono. - Lorn
Re: [mer-general] [MER AND NEMO MIDDLEWARE] RFC: to merge updated and additional components into Mer from Nemo Middleware project
On 13/04/2014 8:55 pm, Carsten Munk wrote: QML Bindings for common middleware tasks: https://github.com/nemomobile/nemo-qml-plugin-connectivity.git this one too please.