RE: How to authenticate Subversion with SASL2 + LDAP
[sending to users@subversion.apache.org, iso us...@subversion.tigris.org (which is closed)] Hi Mark, I'm sorry, I have no experience getting SASL to work with pam. But I remember having read some things about pam in the Cyrus SASL documentation pages (included with the distribution I think). So I suppose it should work. Maybe someone else on the (correct) subversion users list can help you? You could also have a look at the cyrus-sasl mailing list and/or archive: http://cyrusimap.web.cmu.edu/lists.html Regards, Johan -Oorspronkelijk bericht- Van: grim...@gmail.com [mailto:grim...@gmail.com] Namens Mark Keisler Verzonden: zondag 24 januari 2010 20:05 Aan: Johan Corveleyn CC: i...@pablorizzo.com; us...@subversion.tigris.org Onderwerp: Re: How to authenticate Subversion with SASL2 + LDAP On Mon, Jun 15, 2009 at 9:36 AM, Johan Corveleyn johan.corvel...@uz.kuleuven.ac.be wrote: -Oorspronkelijk bericht- Van: rizzopa...@gmail.com [mailto:rizzopa...@gmail.com] Namens Pablo Manuel Rizzo Verzonden: donderdag 11 juni 2009 16:15 Aan: us...@subversion.tigris.org Onderwerp: Re: How to authenticate Subversion with SASL2 + LDAP On Thu, Jun 11, 2009 at 11:03, Stefan Sperling s...@elego.de wrote: On Thu, Jun 11, 2009 at 10:43:13AM -0300, Pablo Manuel Rizzo wrote: Hey Andrey, first you say it's not possible, now you say it's all there. Is something wrong with you? I'd say he just didn't know that SASL auth was added to svnserve. It's a new feature. Does anybody know how to use sasl to authenticate subversion with ldap? The book suggests that you read the SASL documentation to find out how to configure a given authentication scheme. Have you read this part of the svnbook? I quote http://svnbook.red- bean.com/nightly/en/svn.serverconfig.svnserve.html#svn.serverconfig.svn serve.sasl Because SASL provides so many different kinds of authentication mechanisms, it would be foolish (and far beyond the scope of this book) to try to describe every possible server-side configuration. Instead, we recommend that you read the documentation supplied in the doc/ subdirectory of the SASL source code. It goes into great detail about every mechanism and how to configure the server appropriately for each. Have you tried that? Yes, sasl documentation is quite poor, there are no quick tips or how tos. I see many people in this list and many others and forums is looking for a quick how-to too. If someone in this list was able to make it work and is willing to share how, it would be very nice and usefull to publish this information somewhere. SASL documentation is indeed very poor, and certainly when you're trying to focus on use of LDAP with svnserve. You have to learn all about the different password exchange mechanisms (plain, md5-digest, cram-digest, ...) figuring out all about saslauthd and other stuff, while all you want to know is: where do I configure the LDAP server, bind dn and bind password and stuff like that (like with Apache + LDAP). Short answer: it's not possible, unless you apply a specific patch and build svn server from source. See e.g. http://svn.haxx.se/dev/archive-2008- 01/0719.shtml. I haven't tried it myself though (can't compile from source). The reason why it's not possible, despite what the documentation of SVN says? Read the Known Issues in http://svn.collab.net/repos/svn/trunk/notes/sasl.txt: - ... As a consequence, you won't be able to use the saslauthd daemon to authenticate users, because that method only works with plain text passwords. - And it just so happens that LDAP authentication only works with saslauthd (and plain text passwords). This was one of my major frustrations when trying to get SVN up and running for the first time. We wanted to go for svnserve to get the last drop of performance out of it (and seemingly simple to set up). I lost days and days looking for a solution, why it wouldn't work despite following carefully the SASL instructions, ... until I found that known issue in the sasl.txt file. In the end we switched to Apache just for this (I have not regretted it, but just the time I lost with that ... I would have liked to have known this beforehand). The SVN devs could do current and future SVN users, administrators, ... a big service by clearly documenting this! Please put this in big bold letters in the book: svnserve + SASL + LDAP is not supported. Don't let people work on this assumption until they find the small print in that sasl.txt file. Please don't act in the book like all the SASL mechanisms are supported, because they clearly are not. And if it ever would be supported: please provide a simple example in the book for getting this up and running. Kind regards, Johan
Re: How to authenticate Subversion with SASL2 + LDAP
I'm using apache to authenticate svn against ldap. On Mon, Jan 25, 2010 at 05:50, Johan Corveleyn johan.corvel...@uz.kuleuven.ac.be wrote: [sending to users@subversion.apache.org, iso us...@subversion.tigris.org(which is closed)] Hi Mark, I'm sorry, I have no experience getting SASL to work with pam. But I remember having read some things about pam in the Cyrus SASL documentation pages (included with the distribution I think). So I suppose it should work. Maybe someone else on the (correct) subversion users list can help you? You could also have a look at the cyrus-sasl mailing list and/or archive: http://cyrusimap.web.cmu.edu/lists.html Regards, Johan -Oorspronkelijk bericht- Van: grim...@gmail.com [mailto:grim...@gmail.com] Namens Mark Keisler Verzonden: zondag 24 januari 2010 20:05 Aan: Johan Corveleyn CC: i...@pablorizzo.com; us...@subversion.tigris.org Onderwerp: Re: How to authenticate Subversion with SASL2 + LDAP On Mon, Jun 15, 2009 at 9:36 AM, Johan Corveleyn johan.corvel...@uz.kuleuven.ac.be wrote: -Oorspronkelijk bericht- Van: rizzopa...@gmail.com [mailto:rizzopa...@gmail.com] Namens Pablo Manuel Rizzo Verzonden: donderdag 11 juni 2009 16:15 Aan: us...@subversion.tigris.org Onderwerp: Re: How to authenticate Subversion with SASL2 + LDAP On Thu, Jun 11, 2009 at 11:03, Stefan Sperling s...@elego.de wrote: On Thu, Jun 11, 2009 at 10:43:13AM -0300, Pablo Manuel Rizzo wrote: Hey Andrey, first you say it's not possible, now you say it's all there. Is something wrong with you? I'd say he just didn't know that SASL auth was added to svnserve. It's a new feature. Does anybody know how to use sasl to authenticate subversion with ldap? The book suggests that you read the SASL documentation to find out how to configure a given authentication scheme. Have you read this part of the svnbook? I quote http://svnbook.red- bean.com/nightly/en/svn.serverconfig.svnserve.html#svn.serverconfig.svn serve.sasl Because SASL provides so many different kinds of authentication mechanisms, it would be foolish (and far beyond the scope of this book) to try to describe every possible server-side configuration. Instead, we recommend that you read the documentation supplied in the doc/ subdirectory of the SASL source code. It goes into great detail about every mechanism and how to configure the server appropriately for each. Have you tried that? Yes, sasl documentation is quite poor, there are no quick tips or how tos. I see many people in this list and many others and forums is looking for a quick how-to too. If someone in this list was able to make it work and is willing to share how, it would be very nice and usefull to publish this information somewhere. SASL documentation is indeed very poor, and certainly when you're trying to focus on use of LDAP with svnserve. You have to learn all about the different password exchange mechanisms (plain, md5-digest, cram-digest, ...) figuring out all about saslauthd and other stuff, while all you want to know is: where do I configure the LDAP server, bind dn and bind password and stuff like that (like with Apache + LDAP). Short answer: it's not possible, unless you apply a specific patch and build svn server from source. See e.g. http://svn.haxx.se/dev/archive-2008- 01/0719.shtml. I haven't tried it myself though (can't compile from source). The reason why it's not possible, despite what the documentation of SVN says? Read the Known Issues in http://svn.collab.net/repos/svn/trunk/notes/sasl.txt: - ... As a consequence, you won't be able to use the saslauthd daemon to authenticate users, because that method only works with plain text passwords. - And it just so happens that LDAP authentication only works with saslauthd (and plain text passwords). This was one of my major frustrations when trying to get SVN up and running for the first time. We wanted to go for svnserve to get the last drop of performance out of it (and seemingly simple to set up). I lost days and days looking for a solution, why it wouldn't work despite following carefully the SASL instructions, ... until I found that known issue in the sasl.txt file. In the end we switched to Apache just for this (I have not regretted it, but just the time I lost with that ... I would have liked to have known this beforehand). The SVN devs could do current and future SVN users, administrators, ... a big service by clearly documenting this! Please put this in big bold letters in the book: svnserve + SASL + LDAP is not supported. Don't let people work on this assumption until they find the small print in that
RE: File externals before subversion 1.6
From: gaff...@live.com To: users@subversion.apache.org Subject: File externals before subversion 1.6 Date: Mon, 25 Jan 2010 15:17:17 +0100 It seems, in one of our repositories, that Someone(TM) has been using svn:externals to include single files. This was done with version 1.4 of the SVN server, and it seems to have worked, even though I thought single file externals was first implemented for server version 1.6. How can this be? Or is the svn:exernals something that is handled by the client? (We did have version 1.6 of the client, though.) Asbjørn _ Windows Live: Keep your friends up to date with what you do online. http://www.microsoft.com/middleeast/windows/windowslive/see-it-in-action/social-network-basics.aspx?ocid=PID23461::T:WLMTAGL:ON:WL:en-xm:SI_SB_1:092010
Re: Externals within the same repository
On Mon, Jan 25, 2010 at 09:14, Asbjørn Sæbø gaff...@live.com wrote: Assume the following situation: There is a repository my_repo, structured like this: my_repo my_repo/trunk my_repo/trunk/a my_repo/trunk/a/file.h my_repo/trunk/b where my_repo/trunk/b has an svn:external to import the contents of my_repo/a. That is, b will have the contents of a external-ed in. Is this valid and sound SVN? Yes, just be careful to avoid any recursive externals - having an external in a pointing at b, while having an external in b pointing at a, would probably result in your WC filling all available disk space. Last I knew, Subversion did not check for recursion in externals.
Re: File externals before subversion 1.6
On Mon, Jan 25, 2010 at 09:30, Asbjørn Sæbø gaff...@live.com wrote: From: gaff...@live.com To: users@subversion.apache.org Subject: File externals before subversion 1.6 Date: Mon, 25 Jan 2010 15:17:17 +0100 It seems, in one of our repositories, that Someone(TM) has been using svn:externals to include single files. This was done with version 1.4 of the SVN server, and it seems to have worked, even though I thought single file externals was first implemented for server version 1.6. How can this be? Or is the svn:exernals something that is handled by the client? (We did have version 1.6 of the client, though.) Property expansion, including externals, is always done by the client. The server just sees it as another property to version.
Re: subversion-1.6.9 build error
On Jan 25, 2010, at 6:55 AM, Jiang Li wrote:
Hi Friends,
May I ask you a question?
I just started to use Subversion. I want to compile it from source code on my
Red Hat Linux ES5.4. But I face below error when running 'make'.
My steps are as below:
1. Get the source code subversion-1.6.9;
2. Get sqlite-amalgamation-3_6_22.zip and neon-0.29.3.tar.gz, extract them in
subversion folder and change the directory name to remove the version number;
2. Run 'autogensh' without error;
3. Run 'configure' without error;
4. Run 'make' with below error:
=
/bin/sh /tmp/subversion-1.6.9/libtool --tag=CC --silent --mode=compile gcc
-DLINUX=2 -D_REENTRANT -D_GNU_SOURCE -g -O2 -pthread
-I./subversion/include -I./subversion -I/usr/include/apr-1
-I/usr/include/apr-1 -I/tmp/subversion-1.6.9/neon/src
-I/usr/local/include/neon -I/tmp/subversion-1.6.9/sqlite-amalgamation -o
subversion/libsvn_fs_fs/rep-cache.lo -c subversion/libsvn_fs_fs/rep-cache.c
In file included from subversion/libsvn_fs_fs/rep-cache.c:28:
subversion/libsvn_fs_fs/rep-cache-db.h:8: error: expected identifier or '('
before string constant
subversion/libsvn_fs_fs/rep-cache-db.h:10: error: expected identifier or '('
before string constant
subversion/libsvn_fs_fs/rep-cache-db.h:24: error: expected identifier or '('
before string constant
subversion/libsvn_fs_fs/rep-cache-db.h:24: error: missing terminating
character
subversion/libsvn_fs_fs/rep-cache.c:34: error: missing terminating character
subversion/libsvn_fs_fs/rep-cache.c: In function 'svn_fs_fs__open_rep_cache':
subversion/libsvn_fs_fs/rep-cache.c:67: error: 'upgrade_sql' undeclared
(first use in this function)
subversion/libsvn_fs_fs/rep-cache.c:67: error: (Each undeclared identifier is
reported only once
subversion/libsvn_fs_fs/rep-cache.c:67: error: for each function it appears
in.)
make: *** [subversion/libsvn_fs_fs/rep-cache.lo] Error 1
==
I have no idea to resolve such issue even after googling the web. Could
anyone share some light on this?
It looks like there is a problem with subversion/libsvn_fs_fs/rep-cache-db.h,
which is a generated file. Could you show us the first 10 or 20 lines of that
file?
Thanks,
-Hyrum
Invalid large, negative Content-Length
When trying to svn update a working copy on Linux (svn 1.6.6) I'm receiving the error 413 Request Entity Too Large. Indeed, the Content-Length I'm seeing the client send when I set neon-debug-mask = 138 in [global] of the servers file is: Content-Length: -5221747158880878327 This seems similar to bug #3558 [1] but I'm just trying to update a working repository, I don't have any 2+GB files, and I'm on Linux and not Windows. Any ideas on what I can do to debug this? -Doug [1] http://subversion.tigris.org/issues/show_bug.cgi?id=3558 signature.asc Description: OpenPGP digital signature
Re: Externals within the same repository
Greetings, Asbjørn Sæbø! Assume the following situation: There is a repository my_repo, structured like this: my_repo my_repo/trunk my_repo/trunk/a my_repo/trunk/a/file.h my_repo/trunk/b where my_repo/trunk/b has an svn:external to import the contents of my_repo/a. That is, b will have the contents of a external-ed in. Is this valid and sound SVN? I'm pretty sure file duplication is not valid in any development process, it's not a question of versioning. Ask yourself what REALLY you're trying to achieve. Other than creating a mess that not reflect your workspace. -- WBR, Andrey Repin (anrdae...@freemail.ru) 25.01.2010, 17:48 Sorry for my terrible english...
Re: Keeping tags/ and branches/ in the trunk
Greetings, trans! Is it possible to keep tags/ and branches/ in the trunk/ so as not to have to differentiate these three at the top of a repository. Everything is possible, but ask yourself, do you really want to checkout whole repository every time you request to checkout trunk? In other words, instead of: branches/ BRANCH_xyz/ tags/ RELEASE_1_0_0/ trunk/ README bin/ lib/ One could have: README bin/ branches/ BRANCH_xyz/ lib/ tags/ RELEASE_1_0_0/ I think I might prefer this approach, if possible. The problem of course is in creating new tags and branches, i.e. making a copy, I would need to prevent the tags/ and branches/ directories from being copied in the process. Is there a way to do that? And this one, too. Read svnbook... SVN is a filesystem, all basic rules of file handling apply. -- WBR, Andrey Repin (anrdae...@freemail.ru) 25.01.2010, 17:45 Sorry for my terrible english...
Re: subversion-1.6.9 build error
Can you try just removing the file and then running 'make' again?
On Jan 25, 2010, at 8:48 AM, Jiang Li wrote:
Hi Hyrum,
Thank you so much for your quick reply!
Here is all the content of file subversion/libsvn_fs_fs/rep-cache-db.h.
Just for your information, I just downloaded subversion 1.6.6 and I was able
to compile this version successfully. I am fine with v1.6.6, but I will try
to compile 1.6.9 tomorrow for a last try. This is late in my evening. I
appreciate your any suggestion.
$ cat rep-cache-db.h
/* This file is automatically generated from
* subversion/libsvn_fs_fs/rep-cache-db.sql
* Do not edit it directly, but edit the source file and rerun 'make'
*/
#define REP_CACHE_DB_SQL \
\
\pragma auto_vacuum = 1;
\
\
\create table rep_cache (hash text not null primary key,
\revision integer not null,
\offset integer not null,
\size integer not null,
\expanded_size integer not null);
Jiang Li
2010/1/25 Hyrum K. Wright hyrum_wri...@mail.utexas.edu
On Jan 25, 2010, at 6:55 AM, Jiang Li wrote:
Hi Friends,
May I ask you a question?
I just started to use Subversion. I want to compile it from source code on
my Red Hat Linux ES5.4. But I face below error when running 'make'.
My steps are as below:
1. Get the source code subversion-1.6.9;
2. Get sqlite-amalgamation-3_6_22.zip and neon-0.29.3.tar.gz, extract them
in subversion folder and change the directory name to remove the version
number;
2. Run 'autogensh' without error;
3. Run 'configure' without error;
4. Run 'make' with below error:
=
/bin/sh /tmp/subversion-1.6.9/libtool --tag=CC --silent --mode=compile gcc
-DLINUX=2 -D_REENTRANT -D_GNU_SOURCE -g -O2 -pthread
-I./subversion/include -I./subversion -I/usr/include/apr-1
-I/usr/include/apr-1 -I/tmp/subversion-1.6.9/neon/src
-I/usr/local/include/neon -I/tmp/subversion-1.6.9/sqlite-amalgamation -o
subversion/libsvn_fs_fs/rep-cache.lo -c subversion/libsvn_fs_fs/rep-cache.c
In file included from subversion/libsvn_fs_fs/rep-cache.c:28:
subversion/libsvn_fs_fs/rep-cache-db.h:8: error: expected identifier or '('
before string constant
subversion/libsvn_fs_fs/rep-cache-db.h:10: error: expected identifier or
'(' before string constant
subversion/libsvn_fs_fs/rep-cache-db.h:24: error: expected identifier or
'(' before string constant
subversion/libsvn_fs_fs/rep-cache-db.h:24: error: missing terminating
character
subversion/libsvn_fs_fs/rep-cache.c:34: error: missing terminating
character
subversion/libsvn_fs_fs/rep-cache.c: In function
'svn_fs_fs__open_rep_cache':
subversion/libsvn_fs_fs/rep-cache.c:67: error: 'upgrade_sql' undeclared
(first use in this function)
subversion/libsvn_fs_fs/rep-cache.c:67: error: (Each undeclared identifier
is reported only once
subversion/libsvn_fs_fs/rep-cache.c:67: error: for each function it appears
in.)
make: *** [subversion/libsvn_fs_fs/rep-cache.lo] Error 1
==
I have no idea to resolve such issue even after googling the web. Could
anyone share some light on this?
It looks like there is a problem with subversion/libsvn_fs_fs/rep-cache-db.h,
which is a generated file. Could you show us the first 10 or 20 lines of
that file?
Thanks,
-Hyrum
RE: Externals within the same repository
Date: Mon, 25 Jan 2010 09:32:49 -0500 Subject: Re: Externals within the same repository From: andy.l...@gmail.com To: gaff...@live.com CC: users@subversion.apache.org On Mon, Jan 25, 2010 at 09:14, Asbjørn Sæbø wrote: Assume the following situation: There is a repository my_repo, structured like this: my_repo my_repo/trunk my_repo/trunk/a my_repo/trunk/a/file.h my_repo/trunk/b where my_repo/trunk/b has an svn:external to import the contents of my_repo/a. That is, b will have the contents of a external-ed in. Is this valid and sound SVN? Yes, just be careful to avoid any recursive externals - having an external in a pointing at b, while having an external in b pointing at a, would probably result in your WC filling all available disk space. Last I knew, Subversion did not check for recursion in externals. OK. But it does seem a bit weird, though. When doing a fresh check out of this, all these internal externals do show up as merged in the check out log. As TortoiseSVN presents it: - External C:\asa\my_repotrunk\a\file.h Merged C:\asa\my_repotrunk\a\file.h Since this was a clean check out, there should not really be any merging going on, so this message seems a bit confusing. Any comments? Asbjørn _ Keep your friends updated—even when you’re not signed in. http://www.microsoft.com/middleeast/windows/windowslive/see-it-in-action/social-network-basics.aspx?ocid=PID23461::T:WLMTAGL:ON:WL:en-xm:SI_SB_5:092010
Re: subversion-1.6.9 build error
On Jan 25, 2010, at 8:55 AM, Jiang Li wrote:
Hyrum,
No luck. The file will be re-generated again after running 'make'.
As it should be.
2010/1/25 Hyrum K. Wright hyrum_wri...@mail.utexas.edu
Can you try just removing the file and then running 'make' again?
On Jan 25, 2010, at 8:48 AM, Jiang Li wrote:
Hi Hyrum,
Thank you so much for your quick reply!
Here is all the content of file subversion/libsvn_fs_fs/rep-cache-db.h.
Just for your information, I just downloaded subversion 1.6.6 and I was
able to compile this version successfully. I am fine with v1.6.6, but I
will try to compile 1.6.9 tomorrow for a last try. This is late in my
evening. I appreciate your any suggestion.
$ cat rep-cache-db.h
/* This file is automatically generated from
* subversion/libsvn_fs_fs/rep-cache-db.sql
* Do not edit it directly, but edit the source file and rerun 'make'
*/
#define REP_CACHE_DB_SQL \
\
\pragma auto_vacuum = 1;
\
\
\create table rep_cache (hash text not null primary key,
\revision integer not null,
\offset integer not null,
\size integer not null,
\expanded_size integer not null);
This does not look like the correct contents of the file (there aren't any
terminating characters on each line, for instance).
The longer term solution is for us to generate that file before creating the
tarball (much as we pre-generate the swig bindings), rather than requiring
people to have a locally-installed Python to generate that file locally.
Here are the contents of rep-cache-db.h which work for me:
=
/* This file is automatically generated from
* subversion/libsvn_fs_fs/rep-cache-db.sql
* Do not edit it directly, but edit the source file and rerun 'make'
*/
#define REP_CACHE_DB_SQL \
pragma auto_vacuum = 1; \
create table rep_cache (hash text not null primary key, \
revision integer not null, \
offset integer not null, \
size integer not null, \
expanded_size integer not null); \
=
Jiang Li
2010/1/25 Hyrum K. Wright hyrum_wri...@mail.utexas.edu
On Jan 25, 2010, at 6:55 AM, Jiang Li wrote:
Hi Friends,
May I ask you a question?
I just started to use Subversion. I want to compile it from source code
on my Red Hat Linux ES5.4. But I face below error when running 'make'.
My steps are as below:
1. Get the source code subversion-1.6.9;
2. Get sqlite-amalgamation-3_6_22.zip and neon-0.29.3.tar.gz, extract
them in subversion folder and change the directory name to remove the
version number;
2. Run 'autogensh' without error;
3. Run 'configure' without error;
4. Run 'make' with below error:
=
/bin/sh /tmp/subversion-1.6.9/libtool --tag=CC --silent --mode=compile
gcc -DLINUX=2 -D_REENTRANT -D_GNU_SOURCE -g -O2 -pthread
-I./subversion/include -I./subversion -I/usr/include/apr-1
-I/usr/include/apr-1 -I/tmp/subversion-1.6.9/neon/src
-I/usr/local/include/neon -I/tmp/subversion-1.6.9/sqlite-amalgamation -o
subversion/libsvn_fs_fs/rep-cache.lo -c
subversion/libsvn_fs_fs/rep-cache.c
In file included from subversion/libsvn_fs_fs/rep-cache.c:28:
subversion/libsvn_fs_fs/rep-cache-db.h:8: error: expected identifier or
'(' before string constant
subversion/libsvn_fs_fs/rep-cache-db.h:10: error: expected identifier or
'(' before string constant
subversion/libsvn_fs_fs/rep-cache-db.h:24: error: expected identifier or
'(' before string constant
subversion/libsvn_fs_fs/rep-cache-db.h:24: error: missing terminating
character
subversion/libsvn_fs_fs/rep-cache.c:34: error: missing terminating
character
subversion/libsvn_fs_fs/rep-cache.c: In function
'svn_fs_fs__open_rep_cache':
subversion/libsvn_fs_fs/rep-cache.c:67: error: 'upgrade_sql' undeclared
(first use in this function)
subversion/libsvn_fs_fs/rep-cache.c:67: error: (Each undeclared
identifier is reported only once
subversion/libsvn_fs_fs/rep-cache.c:67: error: for each function it
appears in.)
make: *** [subversion/libsvn_fs_fs/rep-cache.lo] Error 1
==
I have no idea to resolve such issue even after googling the web. Could
anyone share some light on this?
It looks like there is a problem with
subversion/libsvn_fs_fs/rep-cache-db.h, which is a generated file. Could
you show us the first 10 or 20 lines of that file?
Thanks,
-Hyrum
Re: subversion-1.6.9 build error
Hyrum,
Thanks! Tomorrow I will download Subversion 1.6.9 dependency package for a
try.
I suspect the problem was caused by not using that package. I will update
you the result.
I will go offline now. Thanks again for your help!
Jiang Li
2010/1/25 Hyrum K. Wright hyrum_wri...@mail.utexas.edu
On Jan 25, 2010, at 8:55 AM, Jiang Li wrote:
Hyrum,
No luck. The file will be re-generated again after running 'make'.
As it should be.
2010/1/25 Hyrum K. Wright hyrum_wri...@mail.utexas.edu
Can you try just removing the file and then running 'make' again?
On Jan 25, 2010, at 8:48 AM, Jiang Li wrote:
Hi Hyrum,
Thank you so much for your quick reply!
Here is all the content of file subversion/libsvn_fs_fs/rep-cache-db.h.
Just for your information, I just downloaded subversion 1.6.6 and I was
able to compile this version successfully. I am fine with v1.6.6, but I will
try to compile 1.6.9 tomorrow for a last try. This is late in my evening. I
appreciate your any suggestion.
$ cat rep-cache-db.h
/* This file is automatically generated from
* subversion/libsvn_fs_fs/rep-cache-db.sql
* Do not edit it directly, but edit the source file and rerun 'make'
*/
#define REP_CACHE_DB_SQL \
\
\pragma auto_vacuum = 1;
\
\
\create table rep_cache (hash text not null primary key,
\revision integer not null,
\offset integer not null,
\size integer not null,
\expanded_size integer not null);
This does not look like the correct contents of the file (there aren't any
terminating characters on each line, for instance).
The longer term solution is for us to generate that file before creating
the tarball (much as we pre-generate the swig bindings), rather than
requiring people to have a locally-installed Python to generate that file
locally.
Here are the contents of rep-cache-db.h which work for me:
=
/* This file is automatically generated from
* subversion/libsvn_fs_fs/rep-cache-db.sql
* Do not edit it directly, but edit the source file and rerun 'make'
*/
#define REP_CACHE_DB_SQL \
pragma auto_vacuum = 1; \
create table rep_cache (hash text not null primary key, \
revision integer not null, \
offset integer not null, \
size integer not null, \
expanded_size integer not null); \
=
Jiang Li
2010/1/25 Hyrum K. Wright hyrum_wri...@mail.utexas.edu
On Jan 25, 2010, at 6:55 AM, Jiang Li wrote:
Hi Friends,
May I ask you a question?
I just started to use Subversion. I want to compile it from source
code on my Red Hat Linux ES5.4. But I face below error when running 'make'.
My steps are as below:
1. Get the source code subversion-1.6.9;
2. Get sqlite-amalgamation-3_6_22.zip and neon-0.29.3.tar.gz, extract
them in subversion folder and change the directory name to remove the
version number;
2. Run 'autogensh' without error;
3. Run 'configure' without error;
4. Run 'make' with below error:
=
/bin/sh /tmp/subversion-1.6.9/libtool --tag=CC --silent
--mode=compile gcc -DLINUX=2 -D_REENTRANT -D_GNU_SOURCE -g -O2 -pthread
-I./subversion/include -I./subversion -I/usr/include/apr-1
-I/usr/include/apr-1 -I/tmp/subversion-1.6.9/neon/src
-I/usr/local/include/neon -I/tmp/subversion-1.6.9/sqlite-amalgamation -o
subversion/libsvn_fs_fs/rep-cache.lo -c subversion/libsvn_fs_fs/rep-cache.c
In file included from subversion/libsvn_fs_fs/rep-cache.c:28:
subversion/libsvn_fs_fs/rep-cache-db.h:8: error: expected identifier
or '(' before string constant
subversion/libsvn_fs_fs/rep-cache-db.h:10: error: expected identifier
or '(' before string constant
subversion/libsvn_fs_fs/rep-cache-db.h:24: error: expected identifier
or '(' before string constant
subversion/libsvn_fs_fs/rep-cache-db.h:24: error: missing terminating
character
subversion/libsvn_fs_fs/rep-cache.c:34: error: missing terminating
character
subversion/libsvn_fs_fs/rep-cache.c: In function
'svn_fs_fs__open_rep_cache':
subversion/libsvn_fs_fs/rep-cache.c:67: error: 'upgrade_sql'
undeclared (first use in this function)
subversion/libsvn_fs_fs/rep-cache.c:67: error: (Each undeclared
identifier is reported only once
subversion/libsvn_fs_fs/rep-cache.c:67: error: for each function it
appears in.)
make: *** [subversion/libsvn_fs_fs/rep-cache.lo] Error 1
==
I have no idea to resolve such issue even after googling the web.
Could anyone share some light on this?
It looks like there is a problem with
subversion/libsvn_fs_fs/rep-cache-db.h, which is a generated file. Could
you show us the first 10 or 20 lines of that
Re: Invalid large, negative Content-Length
On Mon, Jan 25, 2010 at 09:45:19AM -0500, Doug Warner wrote: When trying to svn update a working copy on Linux (svn 1.6.6) I'm receiving the error 413 Request Entity Too Large. Indeed, the Content-Length I'm seeing the client send when I set neon-debug-mask = 138 in [global] of the servers file is: Content-Length: -5221747158880878327 This seems similar to bug #3558 [1] but I'm just trying to update a working repository, I don't have any 2+GB files, and I'm on Linux and not Windows. Any ideas on what I can do to debug this? You could rebuild svn with the --enable-maintainer-mode configure flag. Then you'll see the file and line number where the error is thrown. You could then have the client break at the spot in gdb, and post a backtrace in case it contains anything interesting (like funky-looking function arguments). This may tell us e.g. if the problem is within neon or within svn. In addition to the complete neon debug output, a tcpdump -s9 -X of the running update would also be nice, to see what client and server are sending to each other over the wire. You could post this information here, or attach it to issue #3558, your call. TIA, Stefan
Re: Keeping tags/ and branches/ in the trunk
On Jan 25, 9:30 am, Andy Levy andy.l...@gmail.com wrote: I don't understand what advantage this would give you. I do see the potential for a lot of confusion and copying/moving of items which shouldn't be operated on, though. Subversion itself applies no special meaning to tags branches so there is nothing *technical* from stopping you from doing this; it seems like you're going against convention for no discernible reason though. Merely convenience --perhaps a minor reason, but a reason nonetheless. At least for me, I find it nice to have the source code closer to the surface. Having now used git for some projects its one of the things I find I like about it. However I prefer Subversion's approach to tags and branches. So I simply wondered, would it be possible to have both? But there are problems with the approach, e.g. checking out the trunk would pull down the tags and branches as well. So I am inquiring if there is anyway to configure Subversion to exclude paths from a checkout, as well as tagging, branching, etc. that would make this doable. trans
Re: Keeping tags/ and branches/ in the trunk
On Mon, Jan 25, 2010 at 11:05, trans transf...@gmail.com wrote: On Jan 25, 9:30 am, Andy Levy andy.l...@gmail.com wrote: I don't understand what advantage this would give you. I do see the potential for a lot of confusion and copying/moving of items which shouldn't be operated on, though. Subversion itself applies no special meaning to tags branches so there is nothing *technical* from stopping you from doing this; it seems like you're going against convention for no discernible reason though. Merely convenience --perhaps a minor reason, but a reason nonetheless. At least for me, I find it nice to have the source code closer to the surface. Having now used git for some projects its one of the things I find I like about it. However I prefer Subversion's approach to tags and branches. So I simply wondered, would it be possible to have both? But there are problems with the approach, e.g. checking out the trunk would pull down the tags and branches as well. So I am inquiring if there is anyway to configure Subversion to exclude paths from a checkout, as well as tagging, branching, etc. that would make this doable. Sparse Checkouts will handle the former, but the latter you'll have to step very carefully or script your branching tagging (and may have to do it in steps, instead of all at once).
Subversion on Linux, Authentication on Windows
Hi all! I have a Linux machine that uses Winbind to perform domain authentication using a Windows Server 2003 server. On this Linux PC there is also a Subersion server (snvserve). I know how to use Subversion integrated security to manage users' access to repositories, but I would like to configure SVN so it can grant or deny access to repository using Windows' user information. Is it possible? Could you explain me how to perform this goal? Thank you in advance. -- Marco Minerva [MCPD], marco.mine...@gmail.com http://blogs.ugidotnet.org/marcom
Re: Probable bug with svn copy
Paul Burba wrote: On Mon, Jan 25, 2010 at 2:17 AM, Noorul Islam K M noo...@collab.net wrote: + # Move new added file to another one and commit. + second_path = os.path.join(new_path, 'second') + rav_svn(None, None, [], 'move', first_path, second_path) + rav_svn(None, None, [Committed revision 4.\n], 'ci', '-m', ^ The expected_stderr is , not Committed revision 4.\n. The expected_stdout is something like: [[[ Adding svn-test-work\working_copies\copy_tests-77\A\New Adding svn-test-work\working_copies\copy_tests-77\A\New\second Committed revision 4. ]]] So you can either pass None as expected_stdout (in which case it isn't checked at all), pass a single string regex describing the above output, or use svntest.actions.run_and_verify_commit(). Oops, sorry, I already committed it in r902841. I'll change that or at least add a comment. Also, would you mind taking a stab at expanding the test to cover Alan's original problem? Specifically that a second working copy, when updated, gets both the move destination *and* the source added. Or a separate test - either way would be really helpful. - Julian
Re: Invalid large, negative Content-Length
On 01/25/2010 10:12 AM, Stefan Sperling wrote: You could rebuild svn with the --enable-maintainer-mode configure flag. Then you'll see the file and line number where the error is thrown. You could then have the client break at the spot in gdb, and post a backtrace in case it contains anything interesting (like funky-looking function arguments). This may tell us e.g. if the problem is within neon or within svn. I'll see where I get with this. In addition to the complete neon debug output, a tcpdump -s9 -X of the running update would also be nice, to see what client and server are sending to each other over the wire. This won't be too useful for my situation since everything is done over SSL. Thanks for the tips; I'll post back with what I find. -Doug signature.asc Description: OpenPGP digital signature
Re: Invalid large, negative Content-Length
On 01/25/2010 01:55 PM, Doug Warner wrote: On 01/25/2010 10:12 AM, Stefan Sperling wrote: You could rebuild svn with the --enable-maintainer-mode configure flag. Then you'll see the file and line number where the error is thrown. You could then have the client break at the spot in gdb, and post a backtrace in case it contains anything interesting (like funky-looking function arguments). This may tell us e.g. if the problem is within neon or within svn. I'll see where I get with this. Unfortunately, after rebuilding with the specified option the error went away. After rebuilding again without it the error was still gone. I'll post back if I wind up getting a reproducible error. -Doug signature.asc Description: OpenPGP digital signature
Problem with svn after Ubuntu upgrade
I upgraded to Ubunto v.9.04 (jaunty), and now I get: svn: error while loading shared libraries: libldap_r.so.2: cannot open shared object file: No such file or directory According to ldd, there are 44 (!) missing shared libraries. I have tried uninstalling and reinstalling subversion. I tried searching Ubuntu forums and have posted this same question there, but no responses so far. Advice would be greatly appreciated. Best regards, Robert Edgar.
