Seb,
Am 31.03.2014 um 16:06 schrieb Sebastian Kuzminsky <s...@highlab.com>: > But i've spent the past year or so working on ja and ubc; fixing bugs, > adding missing required features, and generally trying to move them > towards release readiness. I must have missed your ubc contributions from last year, sorry - exactly which fixes were you referring to? That packaging cannot happen without you working on your buildbot as long as it is the only one is evident, and that is why packaging needs work for ubc3: but by and large in another repo. I suggested to you to install a buildbot VM with Xenomai and RT-preempt prerequisites at least a year ago, and you flatly refused back then. Do you by any chance to see any connection between your stance back then, and the fact that packaging isnt there now? so where are you on that? Did you change your stance? and where are the results? on g.l.o/unified-build-candidate-3: I see some contributions from you about packaging, and lots of work by John to get you going, and quite a bit of that ignored by you, for instance debian/configure. > There are well publicized open issues > (https://sf.net/p/emc/bugs/search/?q=ubc) that I and others ran into > while working on the branch. great. Let's go through these in turn and just make sure you read them in full first: #365 ubc doesn't build debs against our lucid rtai kernel: that is a fix for _your rtai package_, not ubc3. People run ubc3 regularly with RTAI 3.4. But if you come up with an actual bug report _about ubc_ then I will certainly look at it. #363 ubc posix flavor doesnt pass tests on armhf: pick https://github.com/mhaberler/linuxcnc/commit/85713a7d05d3221999bb57a47b7110c2f977f89a - other than your dysfunctional patch which I reverted - this fixes the actual problem #361 snprintf(.. "%lf"...) format interpretation change causes runtests failure: please read the text, the fix is already there: pick it and be done with it: http://git.mah.priv.at/gitweb?p=emc2-dev.git;a=shortlog;h=refs/heads/float-compare #356 ubc: 'make setuid' in RIP builds copies files outside the git repo: cosmetic shell script work - no RTAPI PhD required. And I understand that is fixed already by John. #355 ubc: logging to rsyslogd has problems: the proposed POC was more damage than help. This is fixed better and long since in https://github.com/mhaberler/linuxcnc/commit/8932a99e430747a1ba29e551e2c2e76755da6845 onwards. calling that a release stopper is a bit of a joke, isnt it? to summarize: If you cant pick from a repo, or refuse to fetch and merge forward because your dont like my repo or whatever the reasons are, then I cannot help you, but please stop the badmouthing and grandstanding about your 'contributions'. This isnt exactly what I took home from Wichita as the 'cooperate in good faith' commitment. If you are just dying to shift blame on the completely derailed 2.6 effort, please look elsewhere. seriously pissed, - Michael ------------------------------------------------------------------------------ _______________________________________________ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
