Re: cvs abusing stderr?

2005-04-29 Thread Harald Dunkel
[EMAIL PROTECTED] wrote: Just to prevent an out-of-the-brain change: if they would we sent to stdout, this would break at least SmartCVS. I don't want to break your GUI, but I am sure that it can read from cvs' stdout instead of stderr, too. Wouldn't it be more consistent to reserve stderr

cvs abusing stderr?

2005-04-28 Thread Harald Dunkel
Hi folks, Any special reason why cvs prints a list of cvs server: Updating modulename on stderr instead of stdout? This is surely just a log message, which is supposed to be printed on stdout. On a huge project the _real_ error messages are hidden between all the garbage. It would be

Re: cvs abusing stderr?

2005-04-28 Thread Harald Dunkel
Jim.Hyslop wrote: I can't comment on why these messages go to stderr instead of stdout, but I will mention that the global -q flag suppresses them. I use '-q' so much it's in my .cvsrc file ;-) I have seen -q and -Q, but I don't want to suppress these messages completely. They should just

write proxies and pserver vs ext

2005-04-11 Thread Harald Dunkel
Hi folks, Is there any chance to get rid of the ext-only restriction for PrimaryServer? I surely would not like to give every developer and his mom an account on our central CVS server. Regards Harri ___ Bug-cvs mailing list Bug-cvs@gnu.org