Martin Bochnig wrote: > #0) When you say "we have foo#N in this and that consolidation > already [...]", then you are talking about your distributor > (Sun). And now the "unfortunately": Not all consolidations interact > as well with the community, as yours does. Not all consolidations > provide all of their code. Some things are difficult to find. Some > things don't build. Some things are temorarily not available to > external 3rd party distributors. Other externally available things > may just be out of date. A while ago G11n's xkb has been such a case > (current status??). I believe that it is very good to have the > specialists from G11n maintain the xkb tree's code base. But I would > find it convenient (in the name of all external distributors) to > have xkb in the fox-gate, because without /usr/X11/lib/X11/xkb Xorg > is not usable whatsoever (except you use XMODMAP(1) in > ~/.xinitrc). The Distributor Sun could simply ignore the fox-gate's > xkb tree, which would then be part of it. Not build it and not build > the package, but instead use the packages created by G11n. External > individuals could enable it on demand, and in such a case the "make > download" target should fetch the most recent version of xkb from > G11n. In case such a link doesn't exist yet, we would ask G11n to > create it for us.
I was just explaining why they weren't in the gate before - I still think it would be better if you could work with the JDS & G11n teams to get your changes merged with theirs instead of maintaining separate copies, but it's up to you if you want to put in that effort. Certainly, the Sun distributions will be ignoring them - that's easy, since we don't use the FOX gate directly, but still maintain the Nevada master gate from which we push to FOX, and to which we pull changes from FOX after going through the required review processes, so we can just not pull those directories from FOX into the Nevada X gate. The G11n xkb code is at: http://src.opensolaris.org/source/xref/nv-g11n/g11n/src/xorg/keytables/ aka src/xorg/keytables in ssh://anon at hg.opensolaris.org/hg/nv-g11n/g11n As far as I know that's kept up to date, but you'd have to ask the g11n team about that. > But I don't understand the following: In > http://src.opensolaris.org/source/xref/fox/fox-gate/XW_NV/open-src/xserver/Makefile.inc > you set the server's "Build XDM-Auth-1 extension (default: auto)" > from "auto" to "disable". Why? Because XDM-Auth-1 uses Triple-DES encryption, which at one point meant we couldn't include it in Solaris and still ship to certain countries. Though I believe all the laws around that have changed to recognize that Triple-DES isn't serious encryption any more, I still have to sit down and do paperwork to send to our lawyers to submit to various governments before I can enable it in a Sun product, to comply with the various countries' encryption import/export/usage laws. Again, this is something that you can do in your builds that I could just not pull into the Nevada gates until that paperwork is done. > It is still possible to start the resulting server through xdm, but > you would always be greeted by a warning message by default, sating > "UNSECURE SESSION". You shouldn't get that for XDM-Auth-1 being disabled as long as you're using MIT-MAGIC-COOKIE instead. I haven't checked how the xdm Moinak put into FOX is configured though. I'll have to check that when I merge our existing Solaris xdm codebase with it. -- -Alan Coopersmith- alan.coopersmith at sun.com Sun Microsystems, Inc. - X Window System Engineering