Re: [Groff] Re: valgrind again

[Alejandro Lopez-Valencia]

At 11:23 a.m. 25/02/2003 +0000, Ralph Corderoy wrote:

> Hi Werner,
>
> > Applied, thanks.  Please test.
>
> Just to say that I, like someone else recently on the list, have just
> experienced problems with updating from CVS.  This is on Red Hat 7.2
> with cvs RPM cvs-1.11.1p1-8.7 -- I updated to this recently following a
> security fix.
>
> From what I could see with strace the client was reading EOF from the
> server but then forgot about that and tried to read again, blocking
> (forever?).
>
> I got around it by disabling compression, but that might just be
> co-incidence if the problem comes and goes.
>
>     cvs -f -q up -dP
>
> I'm not saying it's a server problem, just letting people know of my
> experience so we can see if there's a pattern.

<OFF-TOPIC>
I would think it is a bug in the server. CVS 1.11.5 is the version with the 
security fix, and I've had similar problems accessing another CVS 
repository at SourceForge (which is always running all the latest stable 
versions and security fixes of everything).

The client code is not affected, and 1.11.1 is pretty much the same as 
using 1.11.5, except for a few changes on how CVS passwords are written to 
the local .cvspass file.
<OFF-TOPIC />


--
Alejandro Lopez-Valencia                         tora no shinden
python -c "print('ZHJhZHVsQDAwN211bmRvLmNvbQ=='.decode('base64'))"