Merge from vendor branch GCC:
[dragonfly.git] / contrib / cvs-1.12 / BUGS
1See the Cederqvist manual (cvs.texinfo) for information on how to
2report bugs (and what will happen to your bug reports if you do).
4The following is a list of some of the known bugs. It may or may not
5be comprehensive. We would dearly love for people to volunteer to
6help us keep it up to date (for starters, if you notice any
7inaccuracies, please let bug-cvs know as described in the Cederqvist
8manual). There are some other reported bugs in MINOR-BUGS; the
9difference, at least in theory, is that those bugs are less serious.
12* For platform-specific information (in some cases including known
13bugs), see README.VMS, windows-NT/README, or os2/README. There is no
14similar file for the unix-like operating systems (not yet, at least).
15This file also might contain some platform-specific bugs.
18* If your login name contains a space or various other characters
19(particularly an issue on Windows), CVS will have trouble (it will
20write invalid RCS files, probably). The fix would be to have CVS
21change such characters to underscores before writing them to the RCS
22file. Furthermore, the LOGNAME or USER environment variables usually
23won't override the system login name, so this can be hard to work
27* If you specify the -w global option to client/server CVS, it only
28overrides a CVSREAD environment variable set on the client, not a
29CVSREAD variable which was set on the server (for example, in .bashrc
30when the server was run via rsh). The fix of course will be to
31provide a "Option-read-write" request which sends -w, in addition to
32"Global_option -r" which sends -r.
35* Symbolic links to files will not work with or without LockDir. In the
36repository, you should avoid using symbolic links to files since this issue
37can cause data loss. Symlinks are only a problem when writing files. If your
38repository does not allow any write access, symlinks are not a problem.
41* Symbolic links to directories will not work with LockDir. In the
42repository, you should avoid using symbolic links to directories if
43you intend to use LockDir as the correct directory will NOT be locked
44by CVS during write. Directory symlinks are not recommended, but should work
45as long as LockDir is not being used. Symlinks are only a problem when
46writing files. If your repository does not allow any write access, symlinks
47are never a problem, whether or not LockDir is in use.
50* The -m option to "cvs add" does not work with client/server CVS.
51CVS will accept the option, but it won't actually set the
52file's description.
55* cvs update walks into a user's work directory if there's a directory
56 of the same name in the repository even if the user's directory
57 doesn't yet have a CVS admin sub-directory. This can greatly confuse
58 users who try to add the same directory at nearly the same time.
61* From: "Charles M. Hannum" <>
62 To:
63 Subject: Still one more bug
64 Date: Sat, 25 Feb 1995 17:01:15 -0500
66 mycroft@duality [1]; cd /usr/src/lib/libc
67 mycroft@duality [1]; cvs diff -C2 '-D1 day ago' -Dnow
68 cvs server: Diffing .
69 cvs server: Diffing DB
70 cvs [server aborted]: could not chdir to DB: No such file or directory
71 mycroft@duality [1];
73 `DB' is an old directory, which no longer has files in it, and is
74 removed automatically when I use the `-P' option to checkout.
76 This error doesn't occur when run locally.
78 P.S. Is anyone working on fixing these bugs?
81* CVS does not always seem to be waiting to the next filesystem timestamp
82quanta after commits. So far this has only shown up in testing under the BSDI
83OS. The symptoms are that ocassionally CVS will not notice that modified files
84are modified, though the file must be modified within a short time after the
85commit, probably milliseconds or seconds, for this symptom to be noticed. One
86suspected cause is that one of the calls to sleep_past() is being called with
87an incorrect value, though this does not explain why symptoms have only been
88noticed under BSDI.
91* The CVS server is leaving a temp directory (/tmp/cvs-serv*) on AIX 4.3 under
92very rare circumstances (one out of c. 10,500 test cases). This appears to be
93dependent on some sort of race condition as it disappears with tracing enabled
94and under the debugger. Informative feedback is welcome.
97* UNICOS 9.0 on Cray currently fails testing in both client/server and
98 writeproxy modes.
101* Status
103 This experimental version of CVS contains new features which may not have
104 been tested as thoroughly as the stable release. It is classified as:
106 /*-------------.
107 | Experimental |
108 `-------------*/
110 /*-------------------------.
111 | Sane for full scale use. |
112 `-------------------------*/