On Tue, Jun 26, 2012 at 1:02 AM, Peeters Simon <peeters.si...@gmail.com> wrote: > > Sorry that i forgot to give credit where credit is due: > The code in this program is majorly inspired by: > - nodm (this is actually a fork/adaption of an unfinished rewrite of nodm) > - user-session by Auke Kok > - various tutorials and man page examples > > Even though this programs functionality is comparable to user-session, > it was mostly written from scratch without knowledge about the > existence of user-session. > > to Auke: > > ugh > > > > this is starting to duplicate almost all the functionality that I've put in > > xorg-launch-helper and user-session-units. > > well, i had written almost everything before i knew these (and i still > have to take a closer look at them) > > the main difference i think is that i do all xorg preparation (which > sometimes requires root) in the main daemon so all pam variables can > be set accordingly (auto-detect next free vt, next available > address,...) > > > Not to mention the code looks... well, pretty much the same as we've > > had in uxlaunch/user-session for almost 3 years. Some attribution would > > be nice. > > I indeed forgot to put some atribution in the readme, (fixed this in git) > but i have to add a comment to what you say here: > - more than 75% of the code was written before i even looked at the > code of user-session > - large parts of the code were part of my (unfinished) effort to > rewrite nodm half a year ago. (options, pam stuff, most xserver stuff) > - as far as i remember only two features came in based on your code, > the --dpi= option, and support for non-suid Xorg (which are both quiet > trivial afaik) > > > But aside from that, we should not let that go much further. Working > > besides each other is just silly. (although I'll flat-out refuse to work > > with CMake, sorry). > > The CMake part is mainly because i needed a buildsystem fast, and i > had no experience with any buildsystem (from a developer side). and i > did not have the time to learn autotools (it appears pretty complex), > but will do in the future, and the i'll see what i do with this. (feel > free to fork on github and add autotools stuff) > > Besides that, feel free to use any of the code in your projects. > > >> In the gnome directory you can find a daemon i use to catch/forward > >> logout, shutdown and reboot requests from shutdown. > >> If you plan on creating such a DBus forward daemon for your DE, take a > >> look at the DBus oriented pseudo language i created in dbus_lang.h. > > > > That's cool. We certainly will want to make that work for other desktops > > as well. Standardization should be key here. > > well, ideally the interface should talk directly to logind an systemd, > but since this requires changes in packages i don't like changing a > lot, i just made this translation daemon for compatibility. > > I did write this as an exercise for myself, and initially did not > intend to share it, but then you asked for it in the other thread, so > here you have it.
well cool - I'm glad you actually shared it. It's also a good time to actually stick our heads together and come up with a decent plan going forward. One of the things we will want to avoid is to make non-upstreamable code for the parts where it makes sense, for instance, so implementation details do matter. thanks for writing back quickly - I really appreciate it. We also need to figure out a way to get your dbus patch accepted upstream, and perhaps this is something I can further help with. Auke _______________________________________________ systemd-devel mailing list systemd-devel@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/systemd-devel
