Community technical support mailing list was retired 2010 and replaced with a professional technical support team. For assistance please contact: Pre-sales Technical support via email to sales@march-hare.com.
On Tue, 17 Aug 2004 02:31:27 -0700 (PDT), abhishek jaiswal <abhi658606 at yahoo.com> wrote: >However what I think and you would also affirm to is that it would be >better if the revision number started from 0.1 in the CVS and hence >after every release the revision of all the files could be pulled to >say 1, 2, 3 or the version no. No, no, no, NO! The revision numbers will increment *independently* for every file in the project as this file is modified and committed! Normally this means that some files have small changes and stay at low revision numbers for the duration of many releases of the software whereas other files are changed a lot and the revision of these files increment at a faster rate. At the time you release a version of your software the source files will each have a revision that is not in synchronizm with any other file except by pure chance. So you cannot depend on revisions for anything really. But whenevr you do a release you put a release tag on the project and now the exect revision number for each file is connected to the release tag. Then at a later time you can update your project to this tag value and all files will be retrieved from CVS exactly as they were when the tag was put in place. Files that have later been added will be deleted from your project folder by CVS etc. All-in-all you will now be ready to check what went wrong with the released version because you can load it back into your development system. This is what the CVS system is used for. >For right now, I will satisfy by thinking that since revision of the >file is 1.1 and hence the file is being used for development of first >version of the software and whenever 1st version is released after >that all the files will be pulled to revision no 2 and hence now the >files are used to develop version 2 of the software. Also before >pulling every file to revision 2, a branch will be created for bug >fixing of revision 1. No, this is also not true. No "pull" of any revision of any filetakes place and no branch is ever created unless you do so yourself! I suggest you try to read up on CVS fundamentals before you continue. Here is a link to a good on-line book, which deals with the unix version of CVS but is pretty complete nevertheless. Of course the many enhancements that has been added to CVSNT are not discussed here, but that is not what you need anyway. You need to get a hang on what CVS is used for... Here is the link: http://cvsbook.red-bean.com/cvsbook.html /Bo (Bo Berglund, developer in Sweden)