Hi Ken, I merged and pushed on debian git my local changes. For separation I didn't succeed in making it work, so I disabled it (otherwise you won't get new work if you submit too many computation error units)
in order to make it work clone the milkyway code https://github.com/Milkyway-at-home/milkywayathome_client clone the debian directory http://anonscm.debian.org/gitweb/?p=pkg-boinc/boinc-app-milkyway.git make a tarball of the source package ./debian/rules tarball dpkg-buildpackage install, enjoy! report here in case of troubles thanks G. > Il Domenica 12 Gennaio 2014 10:39, Gianfranco Costamagna > <costamagnagianfra...@yahoo.it> ha scritto: > > Hi Ken, I had more luck than you in my boinc-app-milkyway from github. > I can run and validate successfully every milkyway_nbody task but I get a > failure on milkybody_separation one. > > I don't have the nbody build now, I can give it to you tomorrow if you are > intrested. > > Anyway for making it build: > install boinc, boinc-app-dev, boinc-dev > clone the repository > git submodule init > git submodule update > delete the boinc folder (in this way we will use _our_ boinc libraries) > apply the two pull requests > https://github.com/Milkyway-at-home/milkywayathome_client/pulls (or just use > the > patches in debian directory) > > clone the debian directory from here > http://anonscm.debian.org/gitweb/?p=pkg-boinc/boinc-app-milkyway.git > dpkg-buildpackage should do the trick > > On monday I'll update the debian git repository, it won't work right now > > after you create the package (I'll change the debian/rules for building only > the nbody) you will be able to run milkyway. > > As soon as I get more feedbacks, and if I don't figure out what is wrong > with separation, I'll update milkyway using only nbody platform (and of > course after I'm sure this is the right client to ship). > > EDIT: I see that maybe separation isn't just supported > http://milkyway.cs.rpi.edu/milkyway/apps.php > http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=3394 > http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=3407 > > > bests, > > G. > > >> Il Domenica 12 Gennaio 2014 9:27, Ken Sharp > <imwellcushtymel...@googlemail.com> ha scritto: >> > I'm not sure if the code you referenced is the correct code. > According >> to the forums the original code didn't have a licence and the newer > code >> hasn't been released yet, at least that's what I got from the > reading >> the threads. There really is nothing made clear on the subject. >> >> I've asked for a little clarity but with no response so far: >> http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=576 >> >> I tried building the code from github but it can't find libraries, uses > >> old libraries.... a bit of a nightmare. Probably isn't much point in >> using that code until someone officially tells us what is going on. >> >> >> On 03/01/14 00:28, Gianfranco Costamagna wrote: >>> Hi Ken >>> thanks for your report. >>> >>> Unfortunately I tried some time ago to package again milkyway with the > new >> code hosted here [1] >>> >>> Code that I think is the official client one. >>> >>> Anyway I'm just stuck with some errors when I run the package with > >> boinc, and I get only "computation errors" messages. >>> >>> I also asked two pull requests [2], to fix a build failure and a > potential >> security issue. >>> >>> Nobody replied so far. >>> What can I do? Fix everything without upstream support? >>> I can, but I don't have time and man power for making the package >> stable without looking really deeply at the code. >>> >>> I'll try again with the last milkyway code, can you please try to > have >> a better contact with upstream? >>> >>> thanks >>> >>> [1] https://github.com/Milkyway-at-home/milkywayathome_client >>> >>> [2] https://github.com/Milkyway-at-home/milkywayathome_client/pulls >>> >>> G. >>> >>> >>>> Il Giovedì 2 Gennaio 2014 20:48, Ken Sharp >> <imwellcushtymel...@googlemail.com> ha scritto: >>>>> Package: boinc-app-milkyway >>>> Version: 0.18d-4 >>>> >>>> The current version supplied by boinc-app-milkyway appears to run > fine >>>> on armel but: >>>> >>>> 1. The estimated runtime is always way off. >>>> 2. The results are always invalid. >>>> 3. It's a really old version: there will be many changes > between >> 0.18 >>>> and the current upstream 1.x branch which will fix a lot of bugs. >>>> >>>> If there is anything further needed then please let me know. >>>> >>>> Origin: >> http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=3440#60666 >>>> >>>> -- >>>> pkg-boinc-devel mailing list >>>> pkg-boinc-de...@lists.alioth.debian.org >>>> > http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-boinc-devel > >>>> >> > > -- > pkg-boinc-devel mailing list > pkg-boinc-de...@lists.alioth.debian.org > http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-boinc-devel > -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org