Subject: Re: pkg/36930: net/csup can't update from CVS repository
To: None <pkg-manager@netbsd.org, gnats-admin@netbsd.org,>
From: YONETANI Tomokazu <qhwt+nbsd@les.ath.cx>
List: pkgsrc-bugs
Date: 09/06/2007 15:30:04
The following reply was made to PR pkg/36930; it has been noted by GNATS.

From: YONETANI Tomokazu <qhwt+nbsd@les.ath.cx>
To: "Jeremy C. Reed" <reed@reedmedia.net>
Cc: gnats-bugs@NetBSD.org, imil@gcu.info
Subject: Re: pkg/36930: net/csup can't update from CVS repository
	containing commitid keywords
Date: Thu, 6 Sep 2007 23:04:09 +0900

 On Thu, Sep 06, 2007 at 06:50:47AM -0500, Jeremy C. Reed wrote:
 > On Thu, 6 Sep 2007, qhwt+nbsd@les.ath.cx wrote:
 > 
 > > >Fix:
 > > $NetBSD$
 > > 
 > > --- updater.c.orig	14 Mar 2006 03:51:10 -0000
 > > +++ updater.c
 > > @@ -856,6 +856,15 @@ updater_diff_batch(struct updater *up, s
 > >  			if (error)
 > >  				goto bad;
 > >  			break;
 > > +
 > > +		case 'N':
 > > +			fprintf(stderr, "newphrase %s ignored\n", line);
 > > +			/* ignore appended phrases(like commitid) */
 > > +			while ((line = stream_getln(rd, NULL)) != NULL) {
 > > +				if (strcmp(line, ".") == 0)
 > > +					break;
 > > +			}
 > > +			break;
 > >  		default:
 > >  			error = UPDATER_ERR_PROTO;
 > >  			goto bad;
 > 
 > 
 > Has this problem been reported upstream?
 > 
 > FreeBSD project is official maintainer. I think this can be reported to 
 > "Maxime Henrion" <mux AT FreeBSD dot ORG>.
 
 Yes, that was why I wrote
   I talked to the author of Csup last year but failed to convince him to
   change the behavior.  I also asked the author of CVSup about this issue
   but received no replies yet.
 
 Most *BSD users are not affected by this behavior, as long as they use
 the cvs command in the base installation, since it's most likely
 cvs-1.11.x with some local modifications and/or backported fixes.
 
 > I had a quick look at FreeBSD's latest CVS for src/contrib/csup and I 
 > don't see any updates since March 14, 2006 for it. I do see 
 > src/contrib/csup/updater.c,v 1.1.1.2 2006/03/14  but I didn't check 
 > to see if that is what we are using.
 
 Yes, that's HEAD and there's no difference between RELENG_6 either.
 
 Cheers.