[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

RE: svn commit: r1476563 - /subversion/trunk/subversion/svnserve/serve.c

From: Bert Huijben <bert_at_qqmail.nl>
Date: Sat, 27 Apr 2013 16:43:29 +0200

> -----Original Message-----
> From: Daniel Shahaf [mailto:danielsh_at_elego.de]
> Sent: zaterdag 27 april 2013 16:36
> To: dev_at_subversion.apache.org
> Cc: commits_at_subversion.apache.org
> Subject: Re: svn commit: r1476563 -
> /subversion/trunk/subversion/svnserve/serve.c
>
> That's wrong since 'B' may return SVN_RA_SVN__UNSPECIFIED_NUMBER.
> In
> fact, you reverted r1466659 which was a bugfix.
>
> brane_at_apache.org wrote on Sat, Apr 27, 2013 at 12:28:44 -0000:
> > Author: brane
> > Date: Sat Apr 27 12:28:44 2013
> > New Revision: 1476563
> >
> > URL: http://svn.apache.org/r1476563
> > Log:
> > * subversion/svnserve/serve.c (update): Silence a warning about
> conversion
> > of apr_uint64_t to svn_boolean_t by explicitly making the argument to
> > svn_log__update a boolean expression.
> > (update, switch_cmd): Incidentally, replace the cases where we had
> > checks like this: "variable == TRUE" to "variable != FALSE" instead,
> > since the latter is more robust given C's truth-value rules.

Maybe we should add some file local macro to clearly document this special
behavior..

I think Julian tried adding a comment for this special behavior, but this
commit showed that just this comment doesn't work well enough.

(Another option would be to start using our tristate enum for this case...
But we made this svnserve<->libsvn_ra_svn api public following our old
inter-library rules)

        Bert
Received on 2013-04-27 16:44:29 CEST

This is an archived mail posted to the Subversion Dev mailing list.