On Fri, 13 May 2011 11:18:01 +0200, Klaus Espenlaub wrote: > On 13.05.2011 00:08, Mario Lobo wrote: >> On Thursday 12 May 2011 18:58:57 Alexey Eromenko wrote: >> >> > Why no one sends patches upstream, here ? (ofc. they must not break >> > other platforms) >> >> Bernhard is doing a wonderful job with the ports. I'm sure he has his >> (good) reasons for doing it the way he has. It has been working >> perfectly because I've been following bluelife for quite a while now and >> the ports have always compiled and run flawlessly.
Well, it's usually not so smooth for us compared to all supported platforms because we usually start our big testing efforts with the community after a release was done. So it happens that bugs are found and fixed after a release and then we just include the patch in our port instead of waiting for the next release. There are also regularly patches that come from the FreeBSD community and we include in the port but usually at the same time we push them upstream. Additionally there are sometimes patches that fix a particular build problem for exotic configurations but the patch is just a hack to get it build. Such a patch will never be pushed upstream but we report the problem instead. We currently maintain 2 sets of virtualbox ports (in the future 3). These without any suffix are the current stable versions which is 4.0.6 right now. Those with -devel suffix are svn snapshots which is r36702 right now and is updated about once a month at least. >> Probably the few remaining code patches would break other OSs. > > Not sure about that in all cases. He pushes things upstream > regularly, when they are sufficiently tested. > > In the long run there should be no reason to have significant changes > in the ports. Yes, that is what we try to do. We don't want to maintain our own patches but we also cannot live without them. -- Bernhard Fröhlich http://www.bluelife.at/ _______________________________________________ vbox-dev mailing list [email protected] http://vbox.innotek.de/mailman/listinfo/vbox-dev
