Re: Permission to translate your page at http://svnbook.red-bean.com/

[Lorenz]

Anja Skrba wrote:
>I am writing to inquire regarding your web page about version control 
>with subversion where I have found a lot of useful information. My name 
>is Anja and I'm currently studying at the Faculty of Computer Science in 
>Belgrade.
>Here is the URL of your article: 
>http://svnbook.red-bean.com/en/1.7/svn-book.html
>
>I would like to share it with the people from Former Yugoslav Republics: 
>Serbia, Montenegro, Croatia, Slovenia, Macedonia, Bosnia and Herzegovina.
>
>I would be grateful if you could allow me to translate your writing into 
>Serbo-Croatian language, that is used in all Former Yugoslav Republics 
>and to post it on my website. Hopefully, it will help our people to 
>gather some additional knowledge about computing.

you'd better ask this question on the book mailing list 

svnbook-...@red-bean.com
-- 

Lorenz

Re: Problems with permissions (changed vith svn commit)

[Daniel Shahaf]

Try
# umask 077
(in the same shell, prior to testing)

Sebastijan V wrote on Thu, Mar 21, 2013 at 02:50:52 +0100:
> Hello
>
> I'm using Fedora. Until subversion-1.6.18-1 everything was OK.
> From version 1.7.8-1 I have following problem:
>
> # touch testfile
> # ls -la testfile
> -rw-r--r--. 1 root root 0 Mar 21 02:49 testfile
>
> # chmod 600 testfile
> # svn add testfile
> # ls -la testfile
> -rw---. 1 root root 0 Mar 21 02:49 testfile
>
> # svn -m '' ci
> # ls -la testfile
> -rw-r--r--. 1 root root 0 Mar 21 02:49 testfile
>
> I don't think, that's OK.
> I was searching on the net, how to preserve permissions, but didn't find  
> anything.
> Is any option, to leave permissions as they are while svn commit?
>
> with */strace svn -m '' ci/* I can see 'chmod 644'
>
> best regards, sebastijan

Re: windows search index flags on checked out files - is a bug?

[Burt]

I have the same issue.  Did you ever find a fix?


Michał Araszkiewicz  gmail.com> writes:

> 
> 
> Hello. 
> I look to confirm whether what i describe below is a bug.
> 
> 
> 
> Issue concerns windows search flags of files checked out on Windows 7.
> 
> 
> 
> Checked out folders have the flag "allow files in this folder to have contents
indexed in addition to file properties" enabled, but files inside them have the
flag "allow this file to have contents indexed in addition to file properties"
cleared. Therefore contents of the files is not indexed. I expect it to be
indexed in order to be able to use windows search (native OS search) on local
working copy (I use it to search documentation - Word, text, excel files).
> 
> 
> 
> I first encoundered it with TortoiseSVN, but subsequently tested it doing a
fresh check-out using command-line svn (as suggested by otik in thread linked
below).
> 
> 
> 
> This behavior still exists.
> 
> 
> 
> 
> Version 1.6.16 is not affected. This older version however had a nuisannce
that it also set the flags allowing for indexing svn metadata. It was possible
to workaround this issue by custom search rules.
> 
> 
> relevant thread
> 
> 
>
http://tortoisesvn.tigris.org/ds/viewMessage.do?dsForumId=4061&viewType=browseAll&dsMessageId=3007689#messagefocus
 
> 
> google groups copy of the
threadhttps://groups.google.com/forum/?fromgroups=#!topic/tortoisesvn/gtCPsFDQk80
> 
> 
> 
> 
> version of TSVN:TortoiseSVN 1.7.9, Build 23248 - 64 Bit , 2012/08/30
18:25:37Subversion 1.7.6, apr 1.4.6apr-utils 1.3.12neon 0.29.6OpenSSL 1.0.1c 10
May 2012zlib 1.2.7 
> 
> 
> 
> thanks for concern. best,
> Michal
> 

windows binaries with xampp

[Andrew Peterson]

Has anyone successfully used xampp with the binaries from:
Win32Svn  (32-bit client, server
and bindings, MSI and ZIPs; maintained by David Darj )


I've tried a bunch of different configurations, but every time I tried to
load the mod_dav_svn.so and the mod_authz_svn.so, apache just crashes and
tells me it can't load the module.

Thanks,

Andrew

RE: windows search index flags on checked out files - is a bug?

[Bert Huijben]

> -Original Message-
> From: Burt [mailto:burtcampb...@yahoo.com]
> Sent: donderdag 21 maart 2013 12:23
> To: users@subversion.apache.org
> Subject: Re: windows search index flags on checked out files - is a bug?
> 
> I have the same issue.  Did you ever find a fix?
> 
> 
> Michał Araszkiewicz  gmail.com> writes:
> 
> >
> >
> > Hello.
> > I look to confirm whether what i describe below is a bug.

I think this is caused by subversion creating files in .svn/tmp/ and then 
moving them into the working copy.

The .svn directory is marked as not for indexing to avoid the indexer 
continuously opening files that we are about to move (which slows us down 
before moving every file). And on Windows 7 without SP1 these sometimes 
triggered a bug in NTFS which made Windows report that the drive was corrupt.

I'm not really sure what the right way of fixing would be. Maybe Subversion 
should use two separate temp directories in .svn/ One for real temp files that 
mustn't be indexed and one for files that are about to be moved into the 
working copy.

Bert

Re: windows search index flags on checked out files - is a bug?

[Thorsten Schöning]

Guten Tag Bert Huijben,
am Donnerstag, 21. März 2013 um 21:28 schrieben Sie:

> I'm not really sure what the right way of fixing would be. Maybe
> Subversion should use two separate temp directories in .svn/ One for
> real temp files that mustn't be indexed and one for files that are
> about to be moved into the working copy.

What about just setting the flag again if necessary? The interesting
thing is only how to know if it is necessary, since it could be
cleared by the user by purpose and the file get reverted afterwards or
else. SVN could track this flag before any operation to the file and
set/clear it again after the operation as needed or by default just
set the flag if the parent folder has set it. This surely fits 99% of
use cases as only a small number of users will clear the flag on
single files.

Mit freundlichen Grüßen,

Thorsten Schöning

-- 
Thorsten Schöning   E-Mail:thorsten.schoen...@am-soft.de
AM-SoFT IT-Systeme  http://www.AM-SoFT.de/

Telefon...05151-  9468- 55
Fax...05151-  9468- 88
Mobil..0178-8 9468- 04

AM-SoFT GmbH IT-Systeme, Brandenburger Str. 7c, 31789 Hameln
AG Hannover HRB 207 694 - Geschäftsführer: Andreas Muchow

Re: SVN access denied when comparing revisions

[Bill Cebula]

Phillip,

Thanks for the reply.  If the line anon-access = read is commented out,
can't that be interpreted as anon-access = none on startup of the server?

It seems like a 1 line conditional statement is all that is needed to fix
this particular issue.  I realize it is related to a larger bug.

Thanks,

Bill


On Wed, Mar 20, 2013 at 6:41 AM, Philip Martin
wrote:

> Bill Cebula  writes:
>
> > I would expect the same behavior whether the* anon-access* line is
> > commented or whether it is specified as *none*.  However, if the
> > anon-access line is commented, you get the error even though you the*
> > authz* specifies read/write privileges for
> > the entire repository.
>
> This is a known bug.  The svn:// protocol chooses either anonymous or
> authenticated access right at the start of the connection and has no way
> to upgrade later.
>
> http://subversion.tigris.org/issues/show_bug.cgi?id=2712
>
> --
> Certified & Supported Apache Subversion Downloads:
> http://www.wandisco.com/subversion/download
>

Re: SVN access denied when comparing revisions

[Daniel Shahaf]

Bill Cebula wrote on Thu, Mar 21, 2013 at 18:30:23 -0500:
> Phillip,
> 
> Thanks for the reply.  If the line anon-access = read is commented out,
> can't that be interpreted as anon-access = none on startup of the server?
> 

No, lack of explicit anon-access is interpreted as anon-access=read.
(and that can't be changed for compatibility reasons)

Daniel

> It seems like a 1 line conditional statement is all that is needed to fix
> this particular issue.  I realize it is related to a larger bug.
> 
> Thanks,
> 
> Bill
> 
> 
> On Wed, Mar 20, 2013 at 6:41 AM, Philip Martin
> wrote:
> 
> > Bill Cebula  writes:
> >
> > > I would expect the same behavior whether the* anon-access* line is
> > > commented or whether it is specified as *none*.  However, if the
> > > anon-access line is commented, you get the error even though you the*
> > > authz* specifies read/write privileges for
> > > the entire repository.
> >
> > This is a known bug.  The svn:// protocol chooses either anonymous or
> > authenticated access right at the start of the connection and has no way
> > to upgrade later.
> >
> > http://subversion.tigris.org/issues/show_bug.cgi?id=2712
> >
> > --
> > Certified & Supported Apache Subversion Downloads:
> > http://www.wandisco.com/subversion/download
> >

Re: SVN access denied when comparing revisions

[Bill Cebula]

If that information ( Lack of explicit anon-access is interpreted as
anon-access=read) was in the svnserv.conf file, it would have saved me a few
hours of troubleshooting.

Thanks,

Bill

On Thu, Mar 21, 2013 at 7:31 PM, Daniel Shahaf wrote:

> Bill Cebula wrote on Thu, Mar 21, 2013 at 18:30:23 -0500:
> > Phillip,
> >
> > Thanks for the reply.  If the line anon-access = read is commented out,
> > can't that be interpreted as anon-access = none on startup of the server?
> >
>
> No, lack of explicit anon-access is interpreted as anon-access=read.
> (and that can't be changed for compatibility reasons)
>
> Daniel
>
> > It seems like a 1 line conditional statement is all that is needed to fix
> > this particular issue.  I realize it is related to a larger bug.
> >
> > Thanks,
> >
> > Bill
> >
> >
> > On Wed, Mar 20, 2013 at 6:41 AM, Philip Martin
> > wrote:
> >
> > > Bill Cebula  writes:
> > >
> > > > I would expect the same behavior whether the* anon-access* line is
> > > > commented or whether it is specified as *none*.  However, if the
> > > > anon-access line is commented, you get the error even though you the*
> > > > authz* specifies read/write privileges for
> > > > the entire repository.
> > >
> > > This is a known bug.  The svn:// protocol chooses either anonymous or
> > > authenticated access right at the start of the connection and has no
> way
> > > to upgrade later.
> > >
> > > http://subversion.tigris.org/issues/show_bug.cgi?id=2712
> > >
> > > --
> > > Certified & Supported Apache Subversion Downloads:
> > > http://www.wandisco.com/subversion/download
> > >
>

Base revision on a svn: external from a pre-commit hook script

[luis montes]

Is it possible to know the base revision on a svn: external from a
pre-commit hook script? If the external changes with the current
transaction I figured I could possibly parse the svnlook diff output, but
what if it does not change with the current transaction? I understand that
information is kept on the client, so does the client send it to the
server? For each external?
Thanks,
Luis