This Bugzilla instance is a read-only archive of historic NetBeans bug reports. To report a bug in NetBeans please follow the project's instructions for reporting issues.
Summary: | Merge doesn't work | ||
---|---|---|---|
Product: | obsolete | Reporter: | Michal Zlamal <mzlamal> |
Component: | collabnet | Assignee: | support <support> |
Status: | RESOLVED INVALID | ||
Severity: | blocker | CC: | jf4jbug, mboyer, nbre |
Priority: | P1 | ||
Version: | 3.x | ||
Hardware: | Other | ||
OS: | Linux | ||
Issue Type: | DEFECT | Exception Reporter: | |
Attachments: |
Log from seesion
History log with steps to reproduce |
Description
Michal Zlamal
2002-06-10 17:55:30 UTC
Created attachment 6183 [details]
Log from seesion
Tested on Solaris 8 with cvs client 1.11 with same result. Merging between two branches (trunk is not involved) seems to behave as allways non-destructive, which makes merging according to cvs manual http://www.cvshome.org/docs/manual/cvs_5.html#SEC61 impossible. I did the test on official NetBeans build machine. I don't use ~/.cvsrc on that machine. Created attachment 6184 [details]
History log with steps to reproduce
Accepting for investigation. This issue has been entered as pcn10040 for investigation. A developer has been assigned to the issue and has requested output of your commands without the -q switch, as this could be hiding some useful information. This has been closed internally as invalid because "cvs diff -r orion_fcs -r orion_fcs_ml" shows no differences. The log shows a cross-branch merge has taken place. If this is incorrect, please reopen. Just to answer an earlier question: Yes, CVS prints output that seems to claim it's merging changes into the working copy :-). But that's just because the revision numbers of the files differ. CVS doesn't know that there is no difference, until it actually goes to take the diff, at which point, it sends an empty diff with the result that you think you're merging changes in, when in reality nothing is happening because there are no changes to merge. So the end result was correct, just the output was a bit misleading, that's all. We recently moved out from Collabnet's infrastructure |