Re: svn commit: r1500343 - /subversion/branches/invoke-diff-cmd-feature/subversion/svnlook/svnlook.c

2013-07-06 Thread Daniel Shahaf
On Sat, Jul 06, 2013 at 11:36:42PM -, g...@apache.org wrote: > @@ -135,6 +136,9 @@ static const apr_getopt_option_t options >{"diff-cmd", svnlook__diff_cmd, 1, > N_("use ARG as diff command")}, > > + {"invoke-diff-cmd", svnlook__invoke_diff_cmd, 1, > + N_("use ARG as dif

Re: svn commit: r1500074 - /subversion/trunk/subversion/libsvn_client/switch.c

2013-07-06 Thread Greg Stein
On Jul 6, 2013 10:11 AM, "Konstantin Kolinko" wrote: > 2013/7/6 Greg Stein : >... > > Did you intend to keep the filename in the message, too? The above > > isn't just a style change, but it (now) retains the filename in the > > path. >... > The last command produces > [[[ > D:\workspace\test>svn

Re: svn commit: r1500074 - /subversion/trunk/subversion/libsvn_client/switch.c

2013-07-06 Thread Konstantin Kolinko
2013/7/6 Greg Stein : > On Fri, Jul 5, 2013 at 1:21 PM, wrote: >>... >> +++ subversion/trunk/subversion/libsvn_client/switch.c Fri Jul 5 17:21:09 >> 2013 >> @@ -239,7 +239,7 @@ switch_internal(svn_revnum_t *result_rev >> return svn_error_createf(SVN_ERR_CLIENT_UNRELATED_RESOURCES, NULL

Re: svn commit: r1500226 - /subversion/trunk/subversion/libsvn_delta/compat.c

2013-07-06 Thread Branko Čibej
On 06.07.2013 11:03, Greg Stein wrote: > Ah! I see. Ignore previous messages :-) > > (in the true branch, you exit with a URL, so prepend sucks; the > prepend was properly needed for the false branch, which you just > fixed) Indeed. The diff lacks a bit of contest from which that would be obvious.

Re: svn commit: r1500091 - /subversion/site/publish/docs/community-guide/general.part.html

2013-07-06 Thread Greg Stein
On Fri, Jul 5, 2013 at 2:20 PM, wrote: >... > +++ subversion/site/publish/docs/community-guide/general.part.html Fri Jul 5 > 18:20:06 2013 > @@ -47,6 +47,15 @@ Subversion repository. To participate:< > various release branches. > > > +The Subversion Community also meets in Berlin for

Re: svn commit: r1500226 - /subversion/trunk/subversion/libsvn_delta/compat.c

2013-07-06 Thread Greg Stein
Ah! I see. Ignore previous messages :-) (in the true branch, you exit with a URL, so prepend sucks; the prepend was properly needed for the false branch, which you just fixed) Thx! -g On Sat, Jul 6, 2013 at 4:57 AM, Greg Stein wrote: > How the heck did you get a URL into >copyfrom_relpath ?? >

Re: svn commit: r1500226 - /subversion/trunk/subversion/libsvn_delta/compat.c

2013-07-06 Thread Greg Stein
How the heck did you get a URL into >copyfrom_relpath ?? Oh. I think I see. You based a URL to BASE_RELPATH in the shim construction, didn't you? On Sat, Jul 6, 2013 at 4:45 AM, Greg Stein wrote: > Wait a sec. I seem to recall that passing a URL into Ev2 is defined as > "incorrect". I remember g

Re: svn commit: r1500226 - /subversion/trunk/subversion/libsvn_delta/compat.c

2013-07-06 Thread Greg Stein
Wait a sec. I seem to recall that passing a URL into Ev2 is defined as "incorrect". I remember going through some of this with Hyrum. For now, I'm going to say -0 to this change. I need to dig deeper thru email and changelists, and reserve the right to fight harder :-P Cheers, -g On Sat, Jul 6,

Re: git-svn "Temp file with moniker 'svn_delta' already in use" and skelta mode

2013-07-06 Thread Bert Huijben
Note that the commit logic in libsvn_client uses exactly the same driver pattern, but even in a more extreme way: it opens all nodes before closing the first file that will receive content changes. Serf was only the first driver to do it this way in the other direction. Bert Sent from W