Commits on Feb 28, 2009
  1. Fix segfault(II) invalid tagged version may not have a pre or post P…

     Occurs when git cvsimport is run again with the same file (and the previous cvsps output is cached)
    committed Feb 28, 2009
Commits on Feb 26, 2009
  1. Fix seg fault when computing 'funky' patchets.

    If the tagged version of the file doesn't exist on any branches badness ensues.
    These missing versions are treated as 'after'
    e.g. The following (deleted) file has a 1.5 revision tagged with v20040316 but this doesn't exist in the repository: cvs rlog org.eclipse.ui.ide/icons/full/etool16/new_wiz.gif
    Warning looks like:
    WARNING: Branch == NULL for: icons/full/etool16/new_wiz.gif v20040316 1.4 1.5 0
    committed Feb 26, 2009
Commits on Jun 17, 2006
  1. Get rid of strsep in cache.c, avoid useless copy

    This destructive parsing is bad, as shown by the fact that its mere
    use in the code makes tuning the parser a hell.  So let's drop it
    before going further.
    This also removes the need to memcpy the buffer, since we won't be
    modifying it.
    Note: there are also loads of strsep() calls in cvs_direct.c to get rid of.
    ydirson committed Jun 17, 2006
Commits on Jun 16, 2006
  1. Separate CPPFLAGS from CFLAGS

    This allows to set CGLAGS on command-line without interfering with
    preprocessor options
    ydirson committed Jun 16, 2006
  2. Diff opts typo fix.

    Roberto C. Sanchez committed with ydirson Jun 2, 2005
  3. Dynamically allocate the log buffer to prevent warning messages

    On (Apple's anonymous CVS server for
    WebKit), some very long log entries were included in CVS.  I got tired
    of cvsps-2.1 truncating them, so I made the 'logbuff' buffer be
    dynamically allocated.
    ddkilzer committed with ydirson Jun 19, 2005
  4. cvsps: should ignore TRUNK branch if it exists in log

    On (where access to the now-open-source
    WebKit is provided), some files have a branch named "TRUNK" with a
    revision of "1".  Yes, I know that's not supposed to be possible, but
    they guys at Apple found a way to do it.  Apparently, using "TRUNK"
    works in places that "HEAD" does not for some cvs commands.
    The current version (2.1) will die when it hits the "TRUNK: 1"
    branch/revision.  The attached patch will ignore any branch or tag named
    ddkilzer committed with ydirson Jun 19, 2005
Commits on Jun 15, 2006
  1. Dependency handling

    (cherry picked from 2e3fed14f81ff18fc115f123c331f986908f1c3f commit)
    ydirson committed Apr 11, 2006
  2. Call cvs with -q flag when fetching the log

    This gets rid of the neverending output polution cvs sends to stderr,
    which is mostly useless and has a strong potential to hide real problem
    ydirson committed Jun 15, 2006
  3. Handle cvs repo with modules

    Parse 'Working file' lines from cvs log output. This alow to work with
    cvs repos
    with modules. To enable this you need to add --no-rlog to cvsps command
    line args.
    This patch was made to import such repo into git. But git-cvsimport
    can't load such data.
    Lanozavr committed with ydirson Jun 15, 2006
Commits on Jun 13, 2006
  1. Use INADDR_NONE instead of -1 to check inet_addr() result

    INADDR_NONE is not equal to -1 on 64-bit systems.
    Signed-off-by: Pavel Roskin <>
    proski committed with ydirson Jun 13, 2006
  2. Use __linux__ conditional, not LINUX.

    __linux__ is defined automatically, LINUX is not.
    Signed-off-by: Pavel Roskin <>
    proski committed with ydirson Jun 13, 2006
  3. FreeBSD isn't evil - just misguided

    cvsps fails against a FreeBSD CVS pserver. By fails I mean it fails back
    to using the 'log' command rather the 'rlog' which would give me stable
    cvsps requires the remote pserver to be 1.11.1 or later for rlog to
    work and, indeed, the other side is a later version. Unfortunately
    FreeBSD have, misguidedly, modified the version string such that cvsps
    misdetects things.
    Anand Kumria committed with ydirson Jun 10, 2006
Commits on Jun 10, 2006
  1. Fix branch ancestor calculation

    Not having any ancestor at all means that any valid ancestor (even of
    "depth 0") is fine.
    Linus Torvalds committed with ydirson Jun 10, 2006
  2. Improve handling of file collisions in the same patchset

    Take the file revision into account.
    Linus Torvalds committed with ydirson Jun 10, 2006
  3. Increase log-length limit to 64kB

    Yeah, it should be dynamic. I'm lazy.
    Linus Torvalds committed with ydirson Jun 10, 2006
Commits on Apr 11, 2006
  1. cvsps-2.1 tarball import

    Yann DIRSON committed Apr 11, 2006