You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by "Jens B. Jorgensen" <je...@tallan.com> on 2003/05/01 15:11:43 UTC
Working directory not locked (on commit) - 0.21.0
I just started using the 0.21.0 version of subversion. I created a repository on
my local drive with svnadmin and started importing source. I created some dirs
in the repository and then was adding directories. Then I was going back and
doing co's on the projects. I'd done a couple of property commits (svn:ignore)
in some directories and that worked fine. I got to one project though and I
started getting some errors saying "Working directory not locked". It is easy to
reproduce this on my system (Win XP, running svn withing cygwin but using the
win32 build provided by Brane Čibej. Lots of commands work (including update),
but not commit. I also tried doing a 'cleanup' and that didn't seem to help.
Here's the transcript:
jbj1@tallan2 /d/src/c++
$ svn co file:///svn/c++/pkcs11
A pkcs11/StdAfx.h
A pkcs11/pkcs11.cpp
A pkcs11/cryptoki
A pkcs11/cryptoki/pkcs11.h
A pkcs11/cryptoki/cryptoki.h
A pkcs11/cryptoki/pkcs11t.h
A pkcs11/cryptoki/pkcs11f.h
A pkcs11/pkcs11_session.cpp
A pkcs11/pkcs11.dsp
A pkcs11/setup.py
A pkcs11/pkcs11.h
A pkcs11/pkcs11test.py
A pkcs11/svnignore
A pkcs11/StdAfx.cpp
A pkcs11/pkcs11_structs.cpp
A pkcs11/readme.txt
A pkcs11/MANIFEST.in
A pkcs11/constants.cpp
A pkcs11/pkcs11.dsw
Checked out revision 10.
jbj1@tallan2 /d/src/c++
$ cd pkcs11
jbj1@tallan2 /d/src/c++/pkcs11
$ echo foo >> readme.txt
jbj1@tallan2 /d/src/c++/pkcs11
$ svn commit -m 'foo!'
svn: Working copy not locked
svn: Are all the targets part of the same working copy?
svn: directory not locked (d:/src/c++/pkcs11)
jbj1@tallan2 /d/src/c++/pkcs11
$ svn status
M readme.txt
jbj1@tallan2 /d/src/c++/pkcs11
$ svn info
Path:
Url: file:///svn/c++/pkcs11
Revision: 10
Node Kind: directory
Schedule: normal
Last Changed Author: jbj1
Last Changed Rev: 10
Last Changed Date: 2003-04-30 18:13:48 -0500 (Wed, 30 Apr 2003)
jbj1@tallan2 /d/src/c++/pkcs11
$ svn --version
svn, version 0.21.0 (r5639)
compiled Apr 17 2003, 00:21:13
Copyright (C) 2000-2003 CollabNet.
Subversion is open source software, see http://subversion.tigris.org/
The following repository access (RA) modules are available:
* ra_dav : Module for accessing a repository via WebDAV (DeltaV) protocol.
- handles 'http' schema
- handles 'https' schema
* ra_local : Module for accessing a repository on local disk.
- handles 'file' schema
* ra_svn : Module for accessing a repository using the svn network protocol.
- handles 'svn' schema
jbj1@tallan2 /d/src/c++/pkcs11
$ svn update
At revision 10.
Any ideas? Should I log a bug? I did scan through the bugs and couldn't find
anything that sounded like this one.
--
Jens B. Jorgensen
jens.jorgensen@tallan.com
"With a focused commitment to our clients and our people, we deliver value
through customized technology solutions"
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Working directory not locked (on commit) - 0.21.0
Posted by Ben Collins-Sussman <su...@collab.net>.
"Jens B. Jorgensen" <je...@tallan.com> writes:
> $ svn commit -m 'foo!'
> svn: Working copy not locked
> svn: Are all the targets part of the same working copy?
> svn: directory not locked (d:/src/c++/pkcs11)
I wonder if this is related to the win32 permissions-bug that gstein
just mentioned?
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
RE: Working directory not locked (on commit) - 0.21.0
Posted by Sander Striker <st...@apache.org>.
> From: Michael Wood [mailto:mwood@its.uct.ac.za]
> Sent: Friday, May 02, 2003 9:06 AM
[...]
>> Hey, dev-list: are there any outstanding patches that need to be
>> applied so this stuff isn't so painful on Windows?
>
> I have never used Subversion on Windows, let alone compiled it on
> Windows, but AFAIK you need to put a copy of APR into the top level
> directory of the Subversion source tree and the same for APR-UTIL before
> running gen-make.py
And don't forget about APR-ICONV...
Sander
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Working directory not locked (on commit) - 0.21.0
Posted by Michael Wood <mw...@its.uct.ac.za>.
On Thu, May 01, 2003 at 03:41:17PM -0500, cmpilato@collab.net wrote:
> "Jens B. Jorgensen" <je...@tallan.com> writes:
>
> > >Incorrect. It just involves running gen-make.py in the top of the
> > >Subversion source tree. Specifically, gen-make.py -t TYPE (see
> > >gen-make.py --help for a list of types).
> > >
> > Very cool. I'm into python. We're almost there now. I'm getting:
> >
> > $ HTTPD=..\\httpd-2.0.43 python gen-make.py -t dsp
> > Found libdb40.lib in db4-win32\lib
> > Traceback (most recent call last):
> > File "gen-make.py", line 89, in ?
> > main(conf, gentype, skip_depends=skip)
> > File "gen-make.py", line 49, in main
> > generator.write(oname)
> > File "build\generator\gen_msvc_dsp.py", line 117, in write
> > fname = self.find_win_project(project_path, ['.dsp'])
> > File "build\generator\gen_win.py", line 348, in find_win_project
> > raise gen_base.GenError("Unable to find project file for external
> > build rule
> > '%s'" % base)
> > gen_base.GenError: Unable to find project file for external build rule
> > 'apr\apr'
> >
> > It looked like I just needed to set HTTPD but that didn't seem to do
> > the trick. Do I need to create a subdir under svn like I did for
> > Berkeley_DB?
>
> Oh, whoops, now you've surpassed my level of knowledge about how this
> all works (or doesn't) in Windows these days.
>
> Hey, dev-list: are there any outstanding patches that need to be
> applied so this stuff isn't so painful on Windows?
I have never used Subversion on Windows, let alone compiled it on
Windows, but AFAIK you need to put a copy of APR into the top level
directory of the Subversion source tree and the same for APR-UTIL before
running gen-make.py
--
Michael Wood <mw...@its.uct.ac.za>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Working directory not locked (on commit) - 0.21.0
Posted by cm...@collab.net.
"Jens B. Jorgensen" <je...@tallan.com> writes:
> >Incorrect. It just involves running gen-make.py in the top of the
> >Subversion source tree. Specifically, gen-make.py -t TYPE (see
> >gen-make.py --help for a list of types).
> >
> Very cool. I'm into python. We're almost there now. I'm getting:
>
> $ HTTPD=..\\httpd-2.0.43 python gen-make.py -t dsp
> Found libdb40.lib in db4-win32\lib
> Traceback (most recent call last):
> File "gen-make.py", line 89, in ?
> main(conf, gentype, skip_depends=skip)
> File "gen-make.py", line 49, in main
> generator.write(oname)
> File "build\generator\gen_msvc_dsp.py", line 117, in write
> fname = self.find_win_project(project_path, ['.dsp'])
> File "build\generator\gen_win.py", line 348, in find_win_project
> raise gen_base.GenError("Unable to find project file for external
> build rule
> '%s'" % base)
> gen_base.GenError: Unable to find project file for external build rule
> 'apr\apr'
>
> It looked like I just needed to set HTTPD but that didn't seem to do
> the trick. Do I need to create a subdir under svn like I did for
> Berkeley_DB?
Oh, whoops, now you've surpassed my level of knowledge about how this
all works (or doesn't) in Windows these days.
Hey, dev-list: are there any outstanding patches that need to be
applied so this stuff isn't so painful on Windows?
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Working directory not locked (on commit) - 0.21.0
Posted by "Jens B. Jorgensen" <je...@tallan.com>.
cmpilato@collab.net wrote:
>"Jens B. Jorgensen" <je...@tallan.com> writes:
>
>
>
>>Ok, I'll bit. I have a working copy of svn plus I've built apache and
>>Berkeley_DB. The build instructions claim existence of a
>>subversion_msvc.dsp/w to build it. I can't find this there. Could
>>someone give me some guidance as to the best way to get my build to
>>happen with vc++. I'm getting a sneaking suspicion it involves
>>cygwin. Is this correct?
>>
>>
>
>Incorrect. It just involves running gen-make.py in the top of the
>Subversion source tree. Specifically, gen-make.py -t TYPE (see
>gen-make.py --help for a list of types).
>
>
Very cool. I'm into python. We're almost there now. I'm getting:
$ HTTPD=..\\httpd-2.0.43 python gen-make.py -t dsp
Found libdb40.lib in db4-win32\lib
Traceback (most recent call last):
File "gen-make.py", line 89, in ?
main(conf, gentype, skip_depends=skip)
File "gen-make.py", line 49, in main
generator.write(oname)
File "build\generator\gen_msvc_dsp.py", line 117, in write
fname = self.find_win_project(project_path, ['.dsp'])
File "build\generator\gen_win.py", line 348, in find_win_project
raise gen_base.GenError("Unable to find project file for external
build rule
'%s'" % base)
gen_base.GenError: Unable to find project file for external build rule
'apr\apr'
It looked like I just needed to set HTTPD but that didn't seem to do the
trick. Do I need to create a subdir under svn like I did for Berkeley_DB?
--
Jens B. Jorgensen
jens.jorgensen@tallan.com
"With a focused commitment to our clients and our people, we deliver value through customized technology solutions"
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Working directory not locked (on commit) - 0.21.0
Posted by cm...@collab.net.
"Jens B. Jorgensen" <je...@tallan.com> writes:
> Ok, I'll bit. I have a working copy of svn plus I've built apache and
> Berkeley_DB. The build instructions claim existence of a
> subversion_msvc.dsp/w to build it. I can't find this there. Could
> someone give me some guidance as to the best way to get my build to
> happen with vc++. I'm getting a sneaking suspicion it involves
> cygwin. Is this correct?
Incorrect. It just involves running gen-make.py in the top of the
Subversion source tree. Specifically, gen-make.py -t TYPE (see
gen-make.py --help for a list of types).
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Working directory not locked (on commit) - 0.21.0
Posted by "Jens B. Jorgensen" <je...@tallan.com>.
Philip Martin wrote:
>"Jens B. Jorgensen" <je...@tallan.com> writes:
>
>
>
>>Is there something I could take a look at?
>>
>>
>
>If you mean you want to debug the problem then you need to determine
>which path gets passed to do_open() in libsvn_wc/lock.c when opening
>an access baton for the directory. The access batons are opened when
>svn_client_commit in libsvn_client/commit.c calls svn_wc_adm_open
>(line 767). Several access batons may be opened, but one should be
>for the directory in question and that path becomes the hash key.
>When svn_client_commit attempts to retrieve the access baton using
>svn_wc_adm_probe_retrieve (line 780) it fails. (At least I am
>assuming that's what happens, but it's really just a guess, until
>someone tracks down the problem we won't know for sure.)
>
From the error message it appears that the retrieve call uses the "d:"
>path, which was returned by svn_path_get_absolute. I don't know why
>or how do_open would get the "/d/" form (if it does), but the path
>appears to come from svn_path_condense_targets which also uses
>svn_path_get_absolute.
>
>
>
Ok, I'll bit. I have a working copy of svn plus I've built apache and
Berkeley_DB. The build instructions claim existence of a
subversion_msvc.dsp/w to build it. I can't find this there. Could
someone give me some guidance as to the best way to get my build to
happen with vc++. I'm getting a sneaking suspicion it involves cygwin.
Is this correct?
--
Jens B. Jorgensen
jens.jorgensen@tallan.com
"With a focused commitment to our clients and our people, we deliver value through customized technology solutions"
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Working directory not locked (on commit) - 0.21.0
Posted by Philip Martin <ph...@codematters.co.uk>.
"Jens B. Jorgensen" <je...@tallan.com> writes:
> Is there something I could take a look at?
If you mean you want to debug the problem then you need to determine
which path gets passed to do_open() in libsvn_wc/lock.c when opening
an access baton for the directory. The access batons are opened when
svn_client_commit in libsvn_client/commit.c calls svn_wc_adm_open
(line 767). Several access batons may be opened, but one should be
for the directory in question and that path becomes the hash key.
When svn_client_commit attempts to retrieve the access baton using
svn_wc_adm_probe_retrieve (line 780) it fails. (At least I am
assuming that's what happens, but it's really just a guess, until
someone tracks down the problem we won't know for sure.)
Re: Working directory not locked (on commit) - 0.21.0
Posted by rb...@rkbloom.net.
On Thu, 1 May 2003, Jens B. Jorgensen wrote:
> Philip Martin wrote:
>
> >"Jens B. Jorgensen" <je...@tallan.com> writes:
> >
> >>Philip Martin wrote:
> >>
> >>>"Jens B. Jorgensen" <je...@tallan.com> writes:
> >>>
> >>>>jbj1@tallan2 /d/src/c++/pkcs11
> >>>>$ svn commit -m 'foo!'
> >>>>svn: Working copy not locked
> >>>>svn: Are all the targets part of the same working copy?
> >>>>svn: directory not locked (d:/src/c++/pkcs11)
> >>>>
> >>>>
> >>>I suspect that the problem is the difference between the strings
> >>>"/d/src/c++/pkcs11" and "d:/src/c++/pkcs11".
> >>>
> >>>
> >>Well, I checked out the repository just as file:///svn/c++/pkcs11.
> >>
> >It's not the URL that matters, this is a working copy path problem.
> >
> >>The /d/src/c++/pkcs11 just shows up because I have d: "mounted" as
> >>/d for cygwin purposes.
> >
> >I don't know anything about cygwin, but Subversion uses paths to key
> >into a hash table. If the same path shows up in two different forms
> >it could produce your error.
> >
> Here's the command I used to create the repository:
>
> svnadmin create svn
>
> Here's the command I used to import the source:
>
> svn import -m 'initial version' file:///svn pkcs11new c++/pkcs11
>
> Basically I have d:/svn as my repository and I have my code under
> d:/src. All the commands have been run with a current working directory
> in d:. Is there somewhere I could check for pathname mismatches? I don't
> see anything suspicious in .svn/entries. I put the file up here:
>
> http://www.ultraemail.net/~jbj1/entries.zip
>
> With respect to rbb@rkbloom.net's comment about cygwin/windows
> difficulties I am aware of the potential problems using cygwin in
> windows. I don't think that I ever specified either d: or /d in any
> pathnames to any svn command through this whole thing (the repository
> was created and all subsequent commands run all in the span of
> yesterday). Not having specified either d: or /d anywhere I'm not sure
> where any other difficulties could have come in. Is there something I
> could take a look at?
If I had to hazard a guess, this is a conflict between APR and something
that SVN is doing natively. My guess (and it is a guess), is that SVN is
trying to get the current directory, and it is also using APR to convert a
directory to an absolute path. Because APR was compiled for native
Windows, it is returning d:\...., but svn is getting /d/.... from cygwin.
Of course, that is a complete guess, without looking at any of the
offending code. I am not even sure that svn is getting a path without
going through APR. All I can tell you, is that APR should _never_ be
returning a /d/... path on Windows, unless it was compiled in a cygwin
environment.
Ryan
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Working directory not locked (on commit) - 0.21.0
Posted by "Jens B. Jorgensen" <je...@tallan.com>.
Philip Martin wrote:
>"Jens B. Jorgensen" <je...@tallan.com> writes:
>
>
>
>>Philip Martin wrote:
>>
>>
>>
>>>"Jens B. Jorgensen" <je...@tallan.com> writes:
>>>
>>>
>>>>jbj1@tallan2 /d/src/c++/pkcs11
>>>>$ svn commit -m 'foo!'
>>>>svn: Working copy not locked
>>>>svn: Are all the targets part of the same working copy?
>>>>svn: directory not locked (d:/src/c++/pkcs11)
>>>>
>>>>
>>>I suspect that the problem is the difference between the strings
>>>"/d/src/c++/pkcs11" and "d:/src/c++/pkcs11".
>>>
>>>
>>Well, I checked out the repository just as file:///svn/c++/pkcs11.
>>
>>
>
>It's not the URL that matters, this is a working copy path problem.
>
>
>
>>The /d/src/c++/pkcs11 just shows up because I have d: "mounted" as
>>/d for cygwin purposes.
>>
>>
>
>I don't know anything about cygwin, but Subversion uses paths to key
>into a hash table. If the same path shows up in two different forms
>it could produce your error.
>
>
>
Here's the command I used to create the repository:
svnadmin create svn
Here's the command I used to import the source:
svn import -m 'initial version' file:///svn pkcs11new c++/pkcs11
Basically I have d:/svn as my repository and I have my code under
d:/src. All the commands have been run with a current working directory
in d:. Is there somewhere I could check for pathname mismatches? I don't
see anything suspicious in .svn/entries. I put the file up here:
http://www.ultraemail.net/~jbj1/entries.zip
With respect to rbb@rkbloom.net's comment about cygwin/windows
difficulties I am aware of the potential problems using cygwin in
windows. I don't think that I ever specified either d: or /d in any
pathnames to any svn command through this whole thing (the repository
was created and all subsequent commands run all in the span of
yesterday). Not having specified either d: or /d anywhere I'm not sure
where any other difficulties could have come in. Is there something I
could take a look at?
--
Jens B. Jorgensen
jens.jorgensen@tallan.com
"With a focused commitment to our clients and our people, we deliver value through customized technology solutions"
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: (workaround) Working directory not locked (on commit) - 0.21.0
Posted by "Jens B. Jorgensen" <je...@tallan.com>.
Philip Martin wrote:
>"Jens B. Jorgensen" <je...@tallan.com> writes:
>
>
>
>>Philip Martin wrote:
>>
>>
>>
>>>"Jens B. Jorgensen" <je...@tallan.com> writes:
>>>
>>>
>>>>jbj1@tallan2 /d/src/c++/pkcs11
>>>>$ svn commit -m 'foo!'
>>>>svn: Working copy not locked
>>>>svn: Are all the targets part of the same working copy?
>>>>svn: directory not locked (d:/src/c++/pkcs11)
>>>>
>>>>
>>>I suspect that the problem is the difference between the strings
>>>"/d/src/c++/pkcs11" and "d:/src/c++/pkcs11".
>>>
>>>
>>Well, I checked out the repository just as file:///svn/c++/pkcs11.
>>
>>
>
>It's not the URL that matters, this is a working copy path problem.
>
>
>
>>The /d/src/c++/pkcs11 just shows up because I have d: "mounted" as
>>/d for cygwin purposes.
>>
>>
>
>I don't know anything about cygwin, but Subversion uses paths to key
>into a hash table. If the same path shows up in two different forms
>it could produce your error.
>
>
>
For the curious, while I was in the process of getting together a build
environment to debug this problem I noticed another post in which
someone used the URL form file:///<drive letter>:/blah/blah. I did not
know that there was a url form where the drive letter was supported (I'd
tried a couple of guesses at how to do this, not sure how I missed the
fairly obvious). I deleted the working copy and instead checked out as:
svn co file:///d:/svn/c++/pkcs11
instead of
svn co file:///svn/c++/pkcs11
This time the commit worked after I changed a file. I'm still interested
in figuring out why the latter didn't work so if I get back to getting
svn compiled (which I hope to do if for no other reason than to keep up
with HEAD) then I'll try to reproduce the error and submit a patch for
any bug I might find. And I promise to read the HACKING doc first! ;-)
--
Jens B. Jorgensen
jens.jorgensen@tallan.com
"With a focused commitment to our clients and our people, we deliver value through customized technology solutions"
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Working directory not locked (on commit) - 0.21.0
Posted by rb...@rkbloom.net.
On 1 May 2003, Philip Martin wrote:
> "Jens B. Jorgensen" <je...@tallan.com> writes:
>
> > Philip Martin wrote:
> >
> > >"Jens B. Jorgensen" <je...@tallan.com> writes:
> >
> > >>jbj1@tallan2 /d/src/c++/pkcs11
> > >>$ svn commit -m 'foo!'
> > >>svn: Working copy not locked
> > >>svn: Are all the targets part of the same working copy?
> > >>svn: directory not locked (d:/src/c++/pkcs11)
> > >
> > >I suspect that the problem is the difference between the strings
> > >"/d/src/c++/pkcs11" and "d:/src/c++/pkcs11".
> >
> > Well, I checked out the repository just as file:///svn/c++/pkcs11.
>
> It's not the URL that matters, this is a working copy path problem.
>
> > The /d/src/c++/pkcs11 just shows up because I have d: "mounted" as
> > /d for cygwin purposes.
>
> I don't know anything about cygwin, but Subversion uses paths to key
> into a hash table. If the same path shows up in two different forms
> it could produce your error.
I don't want to speak out of turn, but this is a common problem with
cygwin. You are using a native Windows program in an emulated Unix
environment, things are almost guaranteed to go wrong. You either need
to use a cygwin compiled version of SVN, or get out of cygwin. All MHO,
but this is the type of reason that I stopped using cygwin, and started
using native Windows apps when I work on Windows.
Ryan
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Working directory not locked (on commit) - 0.21.0
Posted by Philip Martin <ph...@codematters.co.uk>.
"Jens B. Jorgensen" <je...@tallan.com> writes:
> Philip Martin wrote:
>
> >"Jens B. Jorgensen" <je...@tallan.com> writes:
>
> >>jbj1@tallan2 /d/src/c++/pkcs11
> >>$ svn commit -m 'foo!'
> >>svn: Working copy not locked
> >>svn: Are all the targets part of the same working copy?
> >>svn: directory not locked (d:/src/c++/pkcs11)
> >
> >I suspect that the problem is the difference between the strings
> >"/d/src/c++/pkcs11" and "d:/src/c++/pkcs11".
>
> Well, I checked out the repository just as file:///svn/c++/pkcs11.
It's not the URL that matters, this is a working copy path problem.
> The /d/src/c++/pkcs11 just shows up because I have d: "mounted" as
> /d for cygwin purposes.
I don't know anything about cygwin, but Subversion uses paths to key
into a hash table. If the same path shows up in two different forms
it could produce your error.
--
Philip Martin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Working directory not locked (on commit) - 0.21.0
Posted by "Jens B. Jorgensen" <je...@tallan.com>.
Philip Martin wrote:
>"Jens B. Jorgensen" <je...@tallan.com> writes:
>
>
>
>>jbj1@tallan2 /d/src/c++/pkcs11
>>$ svn commit -m 'foo!'
>>svn: Working copy not locked
>>svn: Are all the targets part of the same working copy?
>>svn: directory not locked (d:/src/c++/pkcs11)
>>
>>
>
>I suspect that the problem is the difference between the strings
>"/d/src/c++/pkcs11" and "d:/src/c++/pkcs11".
>
>
>
Well, I checked out the repository just as file:///svn/c++/pkcs11. The
/d/src/c++/pkcs11 just shows up because I have d: "mounted" as /d for
cygwin purposes.
Also, as far as permissions go this particular drive where both the
repository and working copy life is FAT32, not NTFS if that makes a
difference.
--
Jens B. Jorgensen
jens.jorgensen@tallan.com
"With a focused commitment to our clients and our people, we deliver value through customized technology solutions"
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Working directory not locked (on commit) - 0.21.0
Posted by Philip Martin <ph...@codematters.co.uk>.
"Jens B. Jorgensen" <je...@tallan.com> writes:
> jbj1@tallan2 /d/src/c++/pkcs11
> $ svn commit -m 'foo!'
> svn: Working copy not locked
> svn: Are all the targets part of the same working copy?
> svn: directory not locked (d:/src/c++/pkcs11)
I suspect that the problem is the difference between the strings
"/d/src/c++/pkcs11" and "d:/src/c++/pkcs11".
--
Philip Martin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org