On Thu, Oct 2, 2014 at 4:22 PM, Huidae Cho <[email protected]> wrote: > On Wed, Oct 01, 2014 at 05:49:08PM -0400, Vaclav Petras wrote: >> On Wed, Oct 1, 2014 at 4:43 PM, Markus Metz >> <[1][email protected]> wrote: >> >> On Tue, Sep 30, 2014 at 8:16 PM, Huidae Cho <[2][email protected]> >> wrote: >> > Hmm... I "remove"d g.list/g.remove and "rename"d >> g.mlist/g.mremove. Maybe, >> > there is a better way? >> Did you "remove" or "svn remove"? Same for "rename". I guess you did >> "svn [remove|rename]", and some other people did not "make >> distclean" >> before "svn up". FWIW, I had no problems with your changes, doing >> "make distclean" before "svn up". >> Whoever compiles GRASS from source and wants the latest and greatest >> (svn up), must clean the source code first with "make distclean" >> before recompiling. This applies to all 4 branches. This is not >> GRASS >> specific but the way how svn works. >> >> make distclean or just make clean, this might be it. The message from >> SVN was not much helpful, although it changed from the previous >> version. >> But usually you don't have to do make *clean, so how to know these >> things ahead (except for looking at Trac)? I don't know. Or is there >> some option in SVN (from the options you are given) to use "theirs all" >> as it was in previous SVN version (I don't know which)? > > I guess that when you "svn up", it doesn't clean the OBJ directory in > the old g.list/g.remove directories and old object files conflict with > the new source files?
Exactly. The OBJ directory is not under version control, so svn does not touch it. That means, svn up can not remove a directory if it contains anything not under version control, and this can lead to conflicts. Whoever compiles GRASS from source and wants the latest and greatest (svn up), must clean the source code first with "make distclean" before updating with svn up. This applies to all 4 branches. This is not GRASS specific but the way how svn works. I did this on different systems with different svn clients and had no problems. Markus M > > Huidae > >> >> Markus M >> >> > >> > On Tue, Sep 30, 2014 at 1:19 PM, Vaclav Petras >> <[3][email protected]> wrote: >> >> >> >> >> >> >> >> On Tue, Sep 30, 2014 at 12:02 PM, Markus Neteler >> <[4][email protected]> >> >> wrote: >> >>> >> >>> On Tue, Sep 30, 2014 at 5:25 PM, Vaclav Petras >> <[5][email protected]> >> >>> wrote: >> >>> ... >> >>> > Confirmed, I had to do the same in all copies too some time ago. >> SVN is >> >>> > not >> >>> > really good at renaming and deleting things, it is always >> confused and >> >>> > I >> >>> > really don't understand why I need to do revert. >> >>> >> >>> As this almost never happened, it may have been this specific >> commit >> >>> being different from others. >> >>> Whatever, the solution we have. >> >>> >> >> Oh, now I hope it was not because I recommended some Git-like >> procedure to >> >> Huidae which is not really what SVN wanted. Anyway, it was easy to >> solve >> >> although it required manual intervention to automatic builds (at >> least the >> >> one for tests). >> >> >> >> >> >>> >> >>> Markus >> >> >> >> >> >> >> >> _______________________________________________ >> >> grass-dev mailing list >> >> [6][email protected] >> >> [7]http://lists.osgeo.org/mailman/listinfo/grass-dev >> > >> > >> > >> > _______________________________________________ >> > grass-dev mailing list >> > [8][email protected] >> > [9]http://lists.osgeo.org/mailman/listinfo/grass-dev >> >> References >> >> 1. mailto:[email protected] >> 2. mailto:[email protected] >> 3. mailto:[email protected] >> 4. mailto:[email protected] >> 5. mailto:[email protected] >> 6. mailto:[email protected] >> 7. http://lists.osgeo.org/mailman/listinfo/grass-dev >> 8. mailto:[email protected] >> 9. http://lists.osgeo.org/mailman/listinfo/grass-dev _______________________________________________ grass-dev mailing list [email protected] http://lists.osgeo.org/mailman/listinfo/grass-dev
