Anyone else seeing this (cvs wierdness)?

[Peter Wemm]

On Fri, Jun 6, 2008 at 8:55 AM, David O'Brien <obrien at freebsd.org> wrote:
> On Fri, May 16, 2008 at 04:12:39PM +0200, Ian FREISLICH wrote:
>> Recently - I guess in the last month or two - successive cvs updates
>> always "updates" files in the following directories, this with no
>> update to the CVS repo:
>>
>> cddl/contrib/opensolaris
>> contrib/ntp
>> contrib/ipfilter
>> contrib/expat
>> contrib/tcsh
>>
>> I sync a local CVS repo using cvsup and I update my source using
>> 'cvs -q update -PdA'
>
> Why are you always using "update -A"?  Basically all the reports of
> weirdness are due to folks not fully understanding what -A does and is
> for.
>
> If -A removes stickly dates, tags, and (what you're seeing here) stickly
> options.  Options can be set locally, in the ,v file on the server.  Some
> keywords (such as $Name$) may need to be updated due to "update -A".

We use -A because we've often messed with sticky dates and tags in our
checked out copy and want to reset anything we've forgotten and start
from a known state.  For the last 14 years, -A has done exactly that.
If a stray sticky date or sticky tag had been set, -A would reset the
tag and fetch a new copy and life was good.

This is not what is happening now.  If the server has a nonstandard
rcs keyword expansion mode, cvs fetches a fresh copy, each and every
time.  Even if the checked out copy has the correct expansion mode.
Over and over and over again.

Many places in the source tree were affected by clowns using 'cvs
import -ko' in spite of it being removed from the instructions in 1997
and no longer being part of standard procedure.  Ports is badly
affected by -ko at this point.

-- 
Peter Wemm - peter at wemm.org; peter at FreeBSD.org; peter at yahoo-inc.com
"All of this is for nothing if we don't go to the stars" - JMS/B5
"If Java had true garbage collection, most programs would delete
themselves upon execution." -- Robert Sewell