>> From: Alan W. Irwin (on cmake list) > > Yes, [svnversion] is the whole point! It is much easier > than using individual > > revision nubmers for each file. > > Although not nearly as informative. I just tried svnversion, > and it only > gives a range of Revision numbers for the source tree. That's > a non-unique > way of identifying source trees
How so? As I understand the range is the range of project revision nubmers at which the most recent changes for files has occurred. But I am fairly new to SVN after switching from CVS, so I would like to know how identical svnversion output from different source trees would come about? Maybe I should ask that on the svn list as well . >However, simplicity that usually might work often trumps correct complexity Well, I never said I was NOT already doing the other, in fact i said I WAS doing it! I had some of the aformentioned hair-tearing to get the '.h' files to record correct revision, in the executable, without a great deal of repetitive output, as I recall. Complexity that is hard to get correct often results in incorrect complexity :-/ I can even think of a few ways that 'correct' method fails. Like modifying the source without checking in. IN svnversion, the all-important 'M' indicator appears in that case. Of course the dreaded 'M' would only tell you that there's little hope of reconstructing the exact source, that someone failed to use the revision control suitably -- which would be good to know if true. The whole thing can help undertake proper record keeping but I don't think it should quite take the place of proper record keeping right?. It'd be nice to be fool-proof but being clever-deliberate-saboteur-proof is much more difficult and not really required for my project at this time! I hope. > ADD_CUSTOM_COMMAND(TARGET signature. > I guess you missed that suggestion. Not true. I am merely not discussing it at this time because I have not yet really tried it out, I am discussing whether that is actually the way to 'DWIM' , you might still convince me otherwise! But thanks for the suggestion and if it is possible to get non-unique svnversion I'll be glad you pointed out the possiblity! _______________________________________________ CMake mailing list [email protected] http://www.cmake.org/mailman/listinfo/cmake
