Re: issue 3486 should be reopened?

2011-01-12 Thread Gavin Beau Baumanis
Thanks Stefan,
I will remove it from my watch list.



On 13/01/2011, at 6:07 AM, Stefan Sperling wrote:

> On Mon, Nov 29, 2010 at 09:14:09AM +1100, Gavin Beau Baumanis wrote:
>> On 27/11/2010, at 12:46 AM, Stefan Sperling wrote:
>>> On Fri, Nov 26, 2010 at 02:45:07PM +0100, Stefan Sperling wrote:
 On Fri, Nov 26, 2010 at 02:40:07PM +0100, Peter Parker wrote:
> On 13.11.2010 16:07, Daniel Shahaf wrote:
>> Daniel Shahaf wrote on Sat, Nov 13, 2010 at 14:35:13 +0200:
>>> Attached the non-windows version of your script I used for testing.
>> [ SNIP ]
> So as Daniel has same error on trunk, is it possible to reopen issue 3486 
> ?
> 
> Best regards,
> Peter
 
 Yes please. That assertion should never be hit.
 Thanks for reporting this.
 
 Stefan
>>> 
>>> PS: Is your name really Peter Parker? That's awesome!
>> 
>> Hmmm.. My "spidey-senses" are telling me "There's a bug in there, somehwere."
> 
> The bug has been fixed and the fix has been nominated for backport to 1.6.16.
> See http://subversion.tigris.org/issues/show_bug.cgi?id=3486#desc7
> 
> Thanks for reporting this, Peter!
> 
> Stefan


Re: issue 3486 should be reopened?

2011-01-12 Thread Stefan Sperling
On Mon, Nov 29, 2010 at 09:14:09AM +1100, Gavin Beau Baumanis wrote:
> On 27/11/2010, at 12:46 AM, Stefan Sperling wrote:
> > On Fri, Nov 26, 2010 at 02:45:07PM +0100, Stefan Sperling wrote:
> >> On Fri, Nov 26, 2010 at 02:40:07PM +0100, Peter Parker wrote:
> >>> On 13.11.2010 16:07, Daniel Shahaf wrote:
>  Daniel Shahaf wrote on Sat, Nov 13, 2010 at 14:35:13 +0200:
> > Attached the non-windows version of your script I used for testing.
>  [ SNIP ]
> >>> So as Daniel has same error on trunk, is it possible to reopen issue 3486 
> >>> ?
> >>> 
> >>> Best regards,
> >>> Peter
> >> 
> >> Yes please. That assertion should never be hit.
> >> Thanks for reporting this.
> >> 
> >> Stefan
> > 
> > PS: Is your name really Peter Parker? That's awesome!
> 
> Hmmm.. My "spidey-senses" are telling me "There's a bug in there, somehwere."

The bug has been fixed and the fix has been nominated for backport to 1.6.16.
See http://subversion.tigris.org/issues/show_bug.cgi?id=3486#desc7

Thanks for reporting this, Peter!

Stefan


Re: issue 3486 should be reopened?

2010-11-28 Thread Gavin Beau Baumanis
On 27/11/2010, at 12:46 AM, Stefan Sperling wrote:
> On Fri, Nov 26, 2010 at 02:45:07PM +0100, Stefan Sperling wrote:
>> On Fri, Nov 26, 2010 at 02:40:07PM +0100, Peter Parker wrote:
>>> On 13.11.2010 16:07, Daniel Shahaf wrote:
 Daniel Shahaf wrote on Sat, Nov 13, 2010 at 14:35:13 +0200:
> Attached the non-windows version of your script I used for testing.
 [ SNIP ]
>>> So as Daniel has same error on trunk, is it possible to reopen issue 3486 ?
>>> 
>>> Best regards,
>>> Peter
>> 
>> Yes please. That assertion should never be hit.
>> Thanks for reporting this.
>> 
>> Stefan
> 
> PS: Is your name really Peter Parker? That's awesome!

Hmmm.. My "spidey-senses" are telling me "There's a bug in there, somehwere."


Gavin "Beau" Baumanis


Re: issue 3486 should be reopened?

2010-11-26 Thread Stefan Sperling
On Fri, Nov 26, 2010 at 02:45:07PM +0100, Stefan Sperling wrote:
> On Fri, Nov 26, 2010 at 02:40:07PM +0100, Peter Parker wrote:
> > On 13.11.2010 16:07, Daniel Shahaf wrote:
> > > Daniel Shahaf wrote on Sat, Nov 13, 2010 at 14:35:13 +0200:
> > >> Attached the non-windows version of your script I used for testing.
> > > [ SNIP ]
> > So as Daniel has same error on trunk, is it possible to reopen issue 3486 ?
> > 
> > Best regards,
> > Peter
> 
> Yes please. That assertion should never be hit.
> Thanks for reporting this.
> 
> Stefan

PS: Is your name really Peter Parker? That's awesome!


Re: issue 3486 should be reopened?

2010-11-26 Thread Stefan Sperling
On Fri, Nov 26, 2010 at 02:40:07PM +0100, Peter Parker wrote:
> On 13.11.2010 16:07, Daniel Shahaf wrote:
> > Daniel Shahaf wrote on Sat, Nov 13, 2010 at 14:35:13 +0200:
> >> Attached the non-windows version of your script I used for testing.
> > [ SNIP ]
> So as Daniel has same error on trunk, is it possible to reopen issue 3486 ?
> 
> Best regards,
> Peter

Yes please. That assertion should never be hit.
Thanks for reporting this.

Stefan


Re: issue 3486 should be reopened?

2010-11-26 Thread Peter Parker
On 13.11.2010 16:07, Daniel Shahaf wrote:
> Daniel Shahaf wrote on Sat, Nov 13, 2010 at 14:35:13 +0200:
>> Attached the non-windows version of your script I used for testing.
> [ SNIP ]
So as Daniel has same error on trunk, is it possible to reopen issue 3486 ?

Best regards,
Peter



Re: issue 3486 should be reopened?

2010-11-13 Thread Daniel Shahaf
Daniel Shahaf wrote on Sat, Nov 13, 2010 at 14:35:13 +0200:
> Attached the non-windows version of your script I used for testing.

[[[
#!/bin/sh -ex

rm -rf repo wc
REPO_PATH=$(pwd)/repo

$svnadmin create repo
$svn co file://$REPO_PATH wc
pushd wc

mkdir -p Server/trunk/dir
echo x >> Server/trunk/dir/foo.c
echo x >> Server/trunk/dir/bar.c
$svn add Server
$svn commit -m "1"

$svn move Server Source
$svn commit -m "2"

mkdir -p Source/branches
$svn add Source/branches
$svn commit -m "3"

$svn copy Source/trunk Source/branches/3.0
$svn commit -m "4"

$svn delete Source/trunk/dir/foo.c
$svn commit -m "5"


# few checkins
echo xxx >> Source/branches/3.0/dir/foo.c
$svn commit -m "6"

echo xxx >> Source/branches/3.0/dir/bar.c
$svn commit -m "7"

# merge down bar.c only
pushd Source/trunk
# wtf - 6:7 is exclusive on the left side? -r 7 or -r 7:7 does nothing
$svn merge "^"/Source/branches/3.0 -r 6:7
$svn up
$svn commit -m "8"

popd
echo xxx >> Source/branches/3.0/dir/foo.c
$svn commit -m "9"

pushd Source/trunk
test -n "$1" && $svn up
$svn merge "^"/Source/branches/3.0 -r5:9

popd
popd
]]]


Re: issue 3486 should be reopened?

2010-11-13 Thread Daniel Shahaf
Peter Parker wrote on Sat, Nov 13, 2010 at 05:55:06 +0100:
> Hello,
> I still have still issues with fixed issue:
> #3486 ("svn: Attempt to add tree conflict that already exists")
> 
> See attached script for provoking this error.
> 

With trunk, I get:

[[[
+ /home/daniel/src/svn/trunk/subversion/svn/svn merge '^/Source/branches/3.0' 
-r5:9
subversion/libsvn_client/merge.c:8923: (apr_err=195016)
svn: Cannot merge into mixed-revision working copy [7:8]; try updating first
]]]

and when I add an '$svn up' before the '$svn merge', I get:

[[[
+ /home/daniel/src/svn/trunk/subversion/svn/svn merge '^/Source/branches/3.0' 
-r5:9
--- Merging r6 into '.':
   C dir/foo.c
Summary of conflicts:
  Tree conflicts: 1
subversion/libsvn_wc/tree_conflicts.c:403: (apr_err=155016)
svn: Attempt to add tree conflict that already exists at 
'/tmp/sandbox-scripts/wc/Source/trunk/dir/foo.c'
]]]

Attached the non-windows version of your script I used for testing.

Daniel


> Same results with webdav except you get additional error:
> svn: Attempt to add tree conflict that already exists at 'dir\foo.c'
> svn: Error reading spooled REPORT request response
> 
> I used subversion 1.6.13 on windows (script should be easily adapted to
> unix(change line 2 and 8))
> 
> Biggest problem is the bail out on merge leaving all other merges
> unprocessed.
> 
> Best regards,
> Peter
> 
> 



issue 3486 should be reopened?

2010-11-12 Thread Peter Parker
Hello,
I still have still issues with fixed issue:
#3486 ("svn: Attempt to add tree conflict that already exists")

See attached script for provoking this error.

Same results with webdav except you get additional error:
svn: Attempt to add tree conflict that already exists at 'dir\foo.c'
svn: Error reading spooled REPORT request response

I used subversion 1.6.13 on windows (script should be easily adapted to
unix(change line 2 and 8))

Biggest problem is the bail out on merge leaving all other merges
unprocessed.

Best regards,
Peter


@rem config
SET REPO_PATH=file:///c:/tmp/treeconflict/repo
@rem end config

@REM rmdir /S /Q repo wc

svnadmin create repo
svn co %REPO_PATH% wc
pushd wc

mkdir Server\trunk\dir
echo x >> Server\trunk\dir\foo.c
echo x >> Server\trunk\dir\bar.c
svn add Server
svn commit -m "1"

svn move Server Source
svn commit -m "2"

mkdir Source\branches
svn add Source\branches
svn commit -m "3"

svn copy Source\trunk Source\branches\3.0
svn commit -m "4"

svn delete Source\trunk\dir\foo.c
svn commit -m "5"


@rem few checkins
echo xxx >> Source\branches\3.0\dir\foo.c
svn commit -m "6"

echo xxx >> Source\branches\3.0\dir\bar.c
svn commit -m "7"

@rem merge down bar.c only
pushd Source\trunk
@rem wtf - 6:7 is exclusive on the left side? -r 7 or -r 7:7 does nothing
svn merge ^^/Source/branches/3.0 -r 6:7
svn up
svn commit -m "8"

popd
echo xxx >> Source\branches\3.0\dir\foo.c
svn commit -m "9"

pushd Source\trunk
svn merge ^^/Source/branches/3.0 -r5:9

popd
popd