Community
Participate
Working Groups
Starting yesterday (Sunday, 10/16, approximately 5 PM, I am not able to "get" or "put" a critical file in Orbit's CVS ... if I try to "read" it, from CVS shell command, the error message says [cvs] cvs [checkout aborted]: unexpected '\x65' reading revision number in RCS file /cvsroot/tools/org.eclipse.orbit/org.eclipse.orbit.releng/.project,v I tried "touching" the file (removing a blank) to see if that'd fix it up, but, from Eclipse, get the following message when trying to commit from Eclipse The server reported an error while performing the "cvs commit" command. org.eclipse.orbit.releng: cvs [commit aborted]: unexpected '\x65' reading revision number in RCS file /cvsroot/tools/org.eclipse.orbit/org.eclipse.orbit.releng/.project,v I've tried searching the internet for an explanation, but what I've found so far are not very helpful ... they say something vague like "something is corrupt in your cvs" or "there is a network problem". Any ideas? We currently can not "build" Orbit at all, due to this problem. (So, I'll mark as 'blocking').
If you think it'd fix the issue, this is a very little file with not sensitive data in history ... if it'd work, we could just delete the version in cvs (from shell) and I could check in a new version? But, not sure if that's "risky" or not?
I think the issue was theses lines: desc @@ 773e3e94567; That desc followed by the ID seems to have caused CVS to return this error. I've pulled those lines, and can checkout the files locally. -M.
That seems to have fixed it. I builds and M build progressing as usual (and I can check in from Eclipse IDE). Guess we'll just write it off to "sunspots"? Or would this be reason to run fsck on whole disk? Thanks so much!
I think I'm late to the game on this one, but this now works for me as well. Thanks.
We are having this exact same problem again, exact same file. see error message in http://build.eclipse.org/orbit/committers/orbit-I/20111209192634/I20111209192634/buildFailedEarly.txt And get similar errors if I try to "fix" with Eclipse IDE: The server reported an error while performing the "cvs update" command. org.eclipse.orbit.releng: cvs [update aborted]: unexpected '\x65' reading revision number in RCS file /cvsroot/tools/org.eclipse.orbit/org.eclipse.orbit.releng/.project,v
Further, I see the same value, when looking directly on cvs server, desc @@ 773e3e94567; followed by desc @@ If I don't hear anything in ... oh, say ... 5 minutes? I may try editing the file myself (on server) to fix. I wonder ... what could be causing this problem? We never ever change it. Could it just be getting too many "tags" in it? Maybe I should change it in some trivial way ... just so we get a version with less "tags" in it?
(In reply to comment #6) > I wonder ... what could be causing this problem? We never ever change it. Could > it just be getting too many "tags" in it? Good question. The file is updated every time a new tag is added... But beyond that, I'm not sure why the file is being 'corrupted' -- and by corrupted, I mean at the RCS-level, not at the file system level.
The "server" did allow me to fix the file directly on server, in the same way, removing desc @@ 773e3e94567; and the file seems "ok" again. I could check it out of cvs and a build has progressed passed that point where it was failing. (And, I did make a new revision ... just in case :) ... but seems not have mattered to the number of tags in the file ... I guess cvs doesn't work that way? Suggestions welcome. [I will confess, I did find one place in our scripts where due to copy/paste error, we would sometimes try to tag the module files twice in a row (with same tag). I think that should normally just be ignored? Result in an error? Or just "do it again" with effectively no difference ... but ... maybe cvs has a bug when tagging twice in a row in quick succession?]
Not sure why this was opened in "architecture" bugzilla component? Changing to Orbit.
