Package: openoffice.org Version: 2.0.4.dfsg.2-7etch2 Severity: important hi,
I've got a mixed environment, with nodes running Sarge/Woody/Etch and docs over Samba remote volumes. Suddenly, after some config tweaks some users on Etch could not open rw their docs anymore with ooffice. It took quite a bit of time to figure out what the problem is, since: 1. OOo2 error msg is quite misleading - it talks about permissions, but those are fine while locking is the issue. 2. in similar bugrep the issue was docs on NFS with no lockd running - but here I see cifsoplockd running, other apps including OOo1 run just fine. So WTH? turn out that OOo2 (checked up to 2.3.0) is one of those apps which (man mount.cifs) "break with cifs style mandatory byte range locks" and the fix is the nobrl option, so mount command should look like % mount -t cifs -o nobrl,... Bug appeared as I switched mount opts from smbfs to cifs on a Etch node. Don't know the internals of OOo2, but I think it shouldn't break grossly like that (couldn't even save new docs -> "general IO error"), and error msg should be more helpful. thanks -- paolo -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
