"Proposal to improve CVS binary file implementation" would probably have been a better title. The larger issue of devising a true "fix" for binary file implementation is a good topic for discussion,
Sandeep, Note: CVSNT (yes it works on linux too, LGPL) has a 'commitid' property - all files with the same 'commitid' were committed together. CVSNT also has 'bugid' which is a user defined change se
But is it really normal behavior for an update -r HEAD to bring in files that have been removed? I've found 3 resurrected files so far. No, it shouldn't do that. Here's the sequence of events in case
Thanks, that did help. I was able to re-remove the file after removing the HEAD sticky tag. My update script was blindly specifying the tag for each project, which seems to be what got me into this m
That may be, or possibly the original question wasn't clear. :-) Are you sure ? The CVS subdirectories should be ignored by default. Still, it's probably a good idea to undo the import. First, back u
Hi all My requirement is to parse the the username, comment and last revision from a specified branch from the log info Can any help me with the syntax of the command. I tried may of the options, not
Hello all, I am new to CVS, but have been involved in version control systems for over a decade. My company has been using MKS RCS for nine years. It has served us well, but due to shortcomings we ne
Derek Robert Price <address@hidden> wrote on 11/05/2003 12:43:14 PM: Yes. And if the example was a child merging to the great grandparent, the GCA would be where the parent begins. The GCA is the mos
No, you can't update an existing revision. When you fix bugs, you need to commit those as new revisions. The normal procedure is to change your second sentence to read as follows: "For each deploy we
[ On , November 7, 2001 at 18:24:35 (GMT), address@hidden wrote: ] That suggests there's something lacking in how your users make use of branches. The best way to work with branches is to ensure the
I thought the attic files accounted for it, but I've just run a full "cvs log -rBRANCHNAME" on our tree, with a perl script to parse out the 'dead' files - and we still get something like 30-40 files
I come again, implement `$DateZone: (ZONENAME)$'. In this implementation, a writer specifies the timezone. Then, the behavior is a place-independent. A problem of this implementation is the literal Z
Can we try and standardize across the various commands. There are several ways to specify various options. It appears to be mostly consistent except for "cvs log". "cvs update, checkout, rtag, tag" a
Hi Larry, Either cvs status -v call71 or more /export/cvs/build/emailor/call71,v I can find TEST_BRANCH. One problem I found is the working machine build(free BSD) cannot directly read file on cvs se
It seems i'm not allowed to post to the CVS developers list. Someone please forward or pick it up otherwise. Let me know if you need something. - Hubert -- Hubert Feyrer <address@hidden> Attached bel
--Original Message-- [mailto:address@hidden On Behalf Of Vikram Sachdeva --Original Message-- This is how I used a tool like CVS Web / ViewCVS to help me do exactly what you are asking for. I used GN
This replaces local f1.c with cvs revision 1.1 of f1.c. Until you commit. When you commit, the history will be updated on the branch you commit to (the branch from which the orrigional f1.c was check
With CVS it is not possible to change history. Also, it is not a good idea to think about file revisions in terms of version numbers. However, it is possible to 'work around' the issue. There are 2 p