Hi Valery, "Valery Fine (Faine)" wrote: > > Did I miss some points. The whole ROOT can be re-built via CMZ > utility in 10 minutes. This is true, but unless you know how to use CMZ, then all you get is a hardwired build of ROOT, which is whatever the ROOT developers choose as the "default build". The problem is that the user has no control over how ROOT will be built. In Jonathan's case, he wanted to build a special debugged version of ROOT, and unless he is willing to learn how to use CMZ, or gets help from someone else who knows how to use CMZ, then he is stuck. > For sure the installation of that CMZ is much simpler (just copy one > file) and the size of that utility (1.5 Mg) is much smaller that > proposed CYGNUS way. If one is an expert with CMZ, then I agree that CMZ is simpler. However, I suspect that there are many more ROOT users who know how to read a Makefile (and know how to use CVS) than know how to use CMZ. In my own case, I have been struggling with CMZ for over a year, which is what prompted me to add the GNU Autoconf, Automake and Libtool support to ROOT in the first place. Granted, I may not be a whiz-bang developer, but I find CMZ just plain hard and difficult to use. I understand that all of the ROOT developers are CMZ experts, and that CMZ has served you well. However, I think it is unreasonable to assume that developers who want to modify ROOT for some special need they have (e.g. like Jonathan's case) are going to learn CMZ. > Anyway all "know-how" information how ROOT is built by platforms are > kept in there and one needs to pick it up from that CMZ. This is actually the problem. I agree with you that all of the know-how is in the CMZ file, but I don't have the CMZ know-how to get at that information. I would like to add the Win9x/WinNT build rules to the new ROOT Automake files, but I need someone (perhaps a CMZ wizard who can cull this information from `root.cmz') to tell me what the logical structure of the build is on Win9x/WinNT, i.e. what libraries and executables are built, and what sources go into these libraries and executables. -- Matthew D. Langston SLD, Stanford Linear Accelerator Center langston@SLAC.Stanford.EDU 541.346.4722
This archive was generated by hypermail 2b29 : Tue Jan 04 2000 - 00:43:31 MET