I noticed some comments in sanity.sh,
around join-19 that made me think that
when adding a file to the trunk, perhaps a
dead revision should be added first, then
the new revision added as a change to that
dead revision, similar to what's done for
adds to a branch.
Is there a reason this
Hello!
The CVS version of CVS (CVS squared :-)) dumps the core on update if
compression is used.
In this case the protocol is pserver. It's the anonymous CVS repository of
CVS, head branch, all up-to-date.
$ gdb --quiet src/cvs
(gdb) set args -f -z3 -q up
(gdb) r
Starting program:
Pavel Roskin wrote:
Hello!
The CVS version of CVS (CVS squared :-)) dumps the core on update if
compression is used.
That was my fault. I'm suprised the error didn't get caught by the compiler,
though. Anyway, I wrote a fix. It should be checked in as soon as make check
finishes.
I
Derek R. Price writes:
I checked in a TODO item (#204) too. If anyone has time to write and submit a
sanity.sh test case for a compressed connection, 'twoult be much appreciated.
:) I'm only assuming there isn't one since this bug wasn't caught by my test
run either.
There's a test for
Thanks for the input. I'll try to build and install 1.11.1.
-Original Message-
From: bjacob [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, May 01, 2001 8:41 PM
To: Chuck.Irvine
Cc: bjacob; info-cvs; bug-cvs
Subject: Unexpected merge conflict - bug maybe?
[EMAIL PROTECTED] writes:
Hey all,
Just curious if anybody is still using some version of MSVC++ earlier
than 6.0 to compile CVS on Windows platforms? In other words, is it
still important to keep the project and make files generated by 4.0, as
is currently the case?
Derek
--
Derek Price CVS
