Re: Restarting Apache during a commit through a proxy

Philip Martin wrote on Wed, Feb 16, 2011 at 23:13:27 +: > Branko Čibej writes: > > > In other words, use a proper crash-resistant transaction commit > > sequence, with automatic rollback as necessary. See the sqlite docs for > > a description of one way of doing this. :) > > Possibly. But t

Re: Updating revision numbers in log messages

Hyrum K Wright wrote on Wed, Feb 16, 2011 at 23:32:37 +: > We could probably work with infra to disable the post-revprop-change > hook for a bit while this operation runs. Just run it locally? (svnadmin setrevprop)

Re: Bindings use is *painful*

Subvertpy might be a useful alternative as well. Jelmer wrote them for the same reason... the swig ones just weren't very Pythonic: -John On Wed, Feb 16, 2011 at 10:08 AM, Hyrum K Wright wrote: > I'm writing a script in which I simply want to retrieve a set

Re: Updating revision numbers in log messages

On Wed, Feb 16, 2011 at 11:25 PM, Peter Samuelson wrote: > > [Hyrum K Wright] >> Through I series of scripts, I've updated the various log messages >> which reference pre-ASF revisions.  I have not yet made the changes to >> the ASF repo, for a couple of reasons: I'm not sure that we have >> conse

Re: Restarting Apache during a commit through a proxy

Branko Čibej writes: > On 17.02.2011 00:13, Philip Martin wrote: >> Branko Čibej writes: >> >>> In other words, use a proper crash-resistant transaction commit >>> sequence, with automatic rollback as necessary. See the sqlite docs for >>> a description of one way of doing this. :) >> Possibly.

Re: Updating revision numbers in log messages

[Hyrum K Wright] > Through I series of scripts, I've updated the various log messages > which reference pre-ASF revisions. I have not yet made the changes to > the ASF repo, for a couple of reasons: I'm not sure that we have > consensus on doing so, and there are a large number of changes to make

Re: Restarting Apache during a commit through a proxy

On 17.02.2011 00:13, Philip Martin wrote: > Branko Čibej writes: > >> In other words, use a proper crash-resistant transaction commit >> sequence, with automatic rollback as necessary. See the sqlite docs for >> a description of one way of doing this. :) > Possibly. But that probably introduces a

Re: Restarting Apache during a commit through a proxy

Branko Čibej writes: > In other words, use a proper crash-resistant transaction commit > sequence, with automatic rollback as necessary. See the sqlite docs for > a description of one way of doing this. :) Possibly. But that probably introduces an overhead that is entirely pointless in most cas

Re: Restarting Apache during a commit through a proxy

On 16.02.2011 23:56, Blair Zajac wrote: > On 02/16/2011 02:15 PM, Philip Martin wrote: >> Blair Zajac writes: >> >>> On 02/16/2011 08:44 AM, Philip Martin wrote: So if the timing is just right it's possible for one Apache process to start writing the transaction, for that process to stop

Re: Restarting Apache during a commit through a proxy

On 02/16/2011 02:15 PM, Philip Martin wrote: Blair Zajac writes: On 02/16/2011 08:44 AM, Philip Martin wrote: So if the timing is just right it's possible for one Apache process to start writing the transaction, for that process to stop, and for another process to take over the commit. WANdi

Re: Restarting Apache during a commit through a proxy

On 16.02.2011 23:15, Philip Martin wrote: > Blair Zajac writes: > >> On 02/16/2011 08:44 AM, Philip Martin wrote: >>> So if the timing is just right it's possible for one Apache process to >>> start writing the transaction, for that process to stop, and for another >>> process to take over the com

Re: Restarting Apache during a commit through a proxy

Blair Zajac writes: > On 02/16/2011 08:44 AM, Philip Martin wrote: >> So if the timing is just right it's possible for one Apache process to >> start writing the transaction, for that process to stop, and for another >> process to take over the commit. WANdisco observed problems on FSFS >> where

Re: Restarting Apache during a commit through a proxy

On 02/16/2011 08:44 AM, Philip Martin wrote: So if the timing is just right it's possible for one Apache process to start writing the transaction, for that process to stop, and for another process to take over the commit. WANdisco observed problems on FSFS where the transaction is synced at the

Re: Bindings use is *painful*

* Hyrum K Wright wrote: > I'm writing a script in which I simply want to retrieve a set of > arbitrary log message and then do stuff with them. Using the Python > bindings seemed like the logical choice (whether is was or not is now > debatable, but that isn't the point of this mail). Turns out

Re: svn commit: r1071330 - in /subversion/branches/1.6.x: ./ subversion/tests/ subversion/tests/cmdline/ subversion/tests/cmdline/svntest/ subversion/tests/libsvn_fs/ subversion/tests/libsvn_fs_base/

On 02/16/2011 10:02 AM, Stefan Sperling wrote: On Wed, Feb 16, 2011 at 05:58:00PM -, hwri...@apache.org wrote: Author: hwright Date: Wed Feb 16 17:58:00 2011 New Revision: 1071330 URL: http://svn.apache.org/viewvc?rev=1071330&view=rev Log: Merge r1051744, r1051745, r1053185, r1053241 from t

Re: svn commit: r1071368 - in /subversion/trunk/subversion: libsvn_wc/status.c svn/status.c tests/cmdline/stat_tests.py

On Wed, Feb 16, 2011 at 02:40:17PM -0500, C. Michael Pilato wrote: > On 02/16/2011 02:27 PM, s...@apache.org wrote: > > Author: stsp > > Date: Wed Feb 16 19:27:40 2011 > > New Revision: 1071368 > > @@ -1793,7 +1793,6 @@ def status_with_tree_conflicts(sbox): > > > > #

Re: svn commit: r1071368 - in /subversion/trunk/subversion: libsvn_wc/status.c svn/status.c tests/cmdline/stat_tests.py

On 02/16/2011 02:27 PM, s...@apache.org wrote: > Author: stsp > Date: Wed Feb 16 19:27:40 2011 > New Revision: 1071368 > > URL: http://svn.apache.org/viewvc?rev=1071368&view=rev > Log: > Make 'svn status' report nested working copies as unversioned regardless > of the format of their working copie

Re: SQLite and callbacks

On 16.02.2011 20:26, Mark Phippard wrote: > 2011/2/16 Branko Čibej : > >> My not very humble opinion -- we can play silly buggers trying to >> optimize this bit of the query, but effort would be better spent in >> merging NODES and ACTUAL_NODE, which in turn would allow us to drop the >> second que

Re: SQLite and callbacks

On Wed, Feb 16, 2011 at 02:26:51PM -0500, Mark Phippard wrote: > 2011/2/16 Branko Čibej : > > > My not very humble opinion -- we can play silly buggers trying to > > optimize this bit of the query, but effort would be better spent in > > merging NODES and ACTUAL_NODE, which in turn would allow us

Re: [PATCH] Fix for issue 3742

On Wed, Feb 16, 2011 at 01:23:06PM +0530, Noorul Islam K M wrote: > > Issue description: > > 'svn status' should reported nested working copies as unversioned regardless > of > the format of their working copies. > > Issue link: http://subversion.tigris.org/issues/show_bug.cgi?id=3742 > > All

Re: Pristine store - spec

Hi Daniel. Thanks for this review. On Wed, 2011-02-16, Daniel Shahaf wrote: > Julian Foad wrote on Tue, Feb 15, 2011 at 15:06:43 +: > > Would anyone be able to review this spec please? I'm trying to get > > straight what locking / access control rules need to be. > > > > Sure, review below

Re: SQLite and callbacks

2011/2/16 Branko Čibej : > My not very humble opinion -- we can play silly buggers trying to > optimize this bit of the query, but effort would be better spent in > merging NODES and ACTUAL_NODE, which in turn would allow us to drop the > second query altogether and halve the total time needed to

Re: SQLite and callbacks

On 16.02.2011 19:25, Julian Foad wrote: > On Tue, 2011-02-15, Stefan Sperling wrote: >> [[[ >> Improve performance of svn proplist in a similar way as was done in r1039808. >> But, this time, avoid problems with callbacks invoked during sqlite >> transactions by storing results in a temporary table

Re: SQLite and callbacks

Julian Foad wrote on Wed, Feb 16, 2011 at 18:25:35 +: >AND NOT (SELECT 1 FROM actual_node > WHERE wc_id = ?1 AND local_relpath = ?2) > > (I'm not sure whether "NOT (SELECT 1 ...)" is the correct or best way to > say "this selection is empty", but you get the idea.) >

Re: SQLite and callbacks

On Tue, 2011-02-15, Stefan Sperling wrote: > [[[ > Improve performance of svn proplist in a similar way as was done in r1039808. > But, this time, avoid problems with callbacks invoked during sqlite > transactions by storing results in a temporary table and invoking > callbacks during a query on th

Updating revision numbers in log messages

Through I series of scripts, I've updated the various log messages which reference pre-ASF revisions. I have not yet made the changes to the ASF repo, for a couple of reasons: I'm not sure that we have consensus on doing so, and there are a large number of changes to make (~10 000, likely causing

Re: svn commit: r1071330 - in /subversion/branches/1.6.x: ./ subversion/tests/ subversion/tests/cmdline/ subversion/tests/cmdline/svntest/ subversion/tests/libsvn_fs/ subversion/tests/libsvn_fs_base/

On Wed, Feb 16, 2011 at 05:58:00PM -, hwri...@apache.org wrote: > Author: hwright > Date: Wed Feb 16 17:58:00 2011 > New Revision: 1071330 > > URL: http://svn.apache.org/viewvc?rev=1071330&view=rev > Log: > Merge r1051744, r1051745, r1053185, r1053241 from trunk, using the > instructions >

Restarting Apache during a commit through a proxy

WANdisco's MultiSite product is essentially a multiplexing http proxy that sits between the client and Apache. Testing has revealed an issue that probably affects other proxies, such as squid. The problem occurs during commit. When the client has a direct connection to Apache it holds a connecti

Re: Bindings use is *painful*

On Wed, Feb 16, 2011 at 4:01 PM, Ramkumar Ramachandra wrote: > Hi, > > Stefan Sperling writes: >> On Wed, Feb 16, 2011 at 03:08:39PM +, Hyrum K Wright wrote: >> This is because the bindings are generated with SWIG. >> We cannot really fix this without using a different approach. >> >> We could

Re: Bindings use is *painful*

Ramkumar Ramachandra wrote: > Stefan Sperling writes: > > On Wed, Feb 16, 2011 at 03:08:39PM +, Hyrum K Wright wrote: > > This is because the bindings are generated with SWIG. > > We cannot really fix this without using a different approach. > > > > We could improve the documentation. An intro

Re: Bindings use is *painful*

Hi, Stefan Sperling writes: > On Wed, Feb 16, 2011 at 03:08:39PM +, Hyrum K Wright wrote: > This is because the bindings are generated with SWIG. > We cannot really fix this without using a different approach. > > We could improve the documentation. An introductory document that > explains ho

Re: svn commit: r1070980 - /subversion/trunk/subversion/svnrdump/load_editor.c

Hi Daniel, Daniel Shahaf writes: > CC += artagnon > > Ramkumar, I think the lesson from Mike and I's recent fixes is: "Mark > unreachable code with a run-time assertion, not with a source comment". > > i.e., if those places were marked with SVN__NOT_IMPLEMENTED(), debugging > would have been eas

Re: Bindings use is *painful*

On 02/16/2011 10:47 AM, Stefan Sperling wrote: > Also, there's an API built on top of the bindings that is easier to use > and has better documentation: http://pysvn.tigris.org/ PySVN isn't built atop our bindings, IIRC. It's a ground-up rewrite of Python and CPP that calls directly into our C AP

Re: [PROPOSAL] commandline switch to override native eol style

Hi, Guenter Knauf writes: > I would like to suggest adding a commandline switch for svn to > override the native eol style. Why do I want this? On windows > platform the native eol style defaults to CRLF which is usual what > we want. Now if you want to work from MSYS bash shell and want to > chec

Re: svn commit: r1070980 - /subversion/trunk/subversion/svnrdump/load_editor.c

On 02/15/2011 10:40 PM, Daniel Shahaf wrote: > Daniel Shahaf wrote on Wed, Feb 16, 2011 at 05:29:58 +0200: >> (Tests pass if I add a SVN__NOT_IMPLEMENTED() to that function.) >> > > On the other hand, some tests fail if I add SVN__NOT_IMPLEMENTED() to > remove_node_props(). Another one for the to

Re: Bindings use is *painful*

On Wed, Feb 16, 2011 at 03:08:39PM +, Hyrum K Wright wrote: > I'm writing a script in which I simply want to retrieve a set of > arbitrary log message and then do stuff with them. Using the Python > bindings seemed like the logical choice (whether is was or not is now > debatable, but that isn

Bindings use is *painful*

I'm writing a script in which I simply want to retrieve a set of arbitrary log message and then do stuff with them. Using the Python bindings seemed like the logical choice (whether is was or not is now debatable, but that isn't the point of this mail). Turns out that using the Python bindings is

[PROPOSAL] commandline switch to override native eol style

Hi all, I would like to suggest adding a commandline switch for svn to override the native eol style. Why do I want this? On windows platform the native eol style defaults to CRLF which is usual what we want. Now if you want to work from MSYS bash shell and want to check out sources from a repo

Re: Pristine store - spec

No comments on the content, but reading the ensuing email thread, it may be useful to put the document in notes/wc-ng/pristines, and add questions / comments / corrections there. It would allow folks down the road to see all the critiques in one location, rather than reading N mails. -Hyrum On

Re: Is the svn:// protocol secure when encrypted via SASL?

If there's an easy way to decompress it, I certainly don't know it. Wireshark just gives me a pile of packets- I don't know what algorithm was used to compress the data, or even where the compressed data starts and ends. Is there any documentation anywhere as to which SASL modules encrypt the

Re: Is the svn:// protocol secure when encrypted via SASL?

No, I didn't mean it to sound harsh. I assumed that, having already captured the data in wireshark, it would be straightforward to get wireshark to decompress it or dump it to a file. Thanks for jumping in and correcting me. Daniel Julian Foad wrote on Wed, Feb 16, 2011 at 12:58:22 +: > On

Re: Is the svn:// protocol secure when encrypted via SASL?

On Wed, 2011-02-16, Daniel Shahaf wrote: > Keith Palmer Jr. wrote on Tue, Feb 15, 2011 at 16:26:19 -0500: > > > > I tried to capture the traffic with Wireshark, but it appears that > > everything is compressed over the wire anyway... so I can't tell if > > I'm looking at just compressed data, or c

Re: Pristine store - spec

Hi Greg. Thanks for reviewing. On Tue, 2011-02-15, Greg Stein wrote: > On Tue, Feb 15, 2011 at 10:06, Julian Foad wrote: > >... > > * Operating Procedures > > * > > * The steps should be carried out in the order specified. (See rationale.) > > * > > * * To add a pristine, do the

Re: Pristine store - spec

On 16.02.2011 13:06, Daniel Shahaf wrote: > Julian Foad wrote on Tue, Feb 15, 2011 at 15:06:43 +: >> * * This specification is conceptually simple, but requires completing disk >> * operations within SDB transactions, which may make it too inefficient >> * in practice. An alternative sp

Re: contribute for subversion project

On Wed, Feb 16, 2011 at 11:58:24AM +0530, Kapila Witharana wrote: > Hi, > > I'm Kapila, a student of Post Graduate Division of University of Colombo > School of Computing (UCSC), University of Colombo, Sri Lanka. I'm now > looking for a research level project (1 year) for my final year project. >

Re: Pristine store - spec

Julian Foad wrote on Tue, Feb 15, 2011 at 15:06:43 +: > Would anyone be able to review this spec please? I'm trying to get > straight what locking / access control rules need to be. > Sure, review below. Having finished the review, though, I wonder if I amn't missing some details of the big

Re: #3641 vetoed

Daniel Shahaf writes: >> My veto for r962377+r962378 was based on the test showing that the >> implementation was not correct. Are you proposing to merge those 3 >> revisions as well and make the test PASS? > > Yes. Good, you can remove the veto. -- Philip

Re: SQLite and callbacks

On 16.02.2011 11:52, Philip Martin wrote: > Stefan Sperling writes: > >> [[[ >> Improve performance of svn proplist in a similar way as was done in r1039808. >> But, this time, avoid problems with callbacks invoked during sqlite >> transactions by storing results in a temporary table and invoking

Re: #3641 vetoed

Philip Martin wrote on Wed, Feb 16, 2011 at 11:09:14 +: > Daniel Shahaf writes: > > > Philip, > > > > You vetoed the r962377 group since descend_into_replace() is XFail. > > > > I've un-xfailed that test in these commits: > > svn commit: r1063572 - in /subversion/trunk/subversion/tests/cmdlin

Re: #3641 vetoed

Daniel Shahaf writes: > Philip, > > You vetoed the r962377 group since descend_into_replace() is XFail. > > I've un-xfailed that test in these commits: > 3485 * 11Jan26,02:35 danielsh@apache.o (2.5K) svn commit: r1063572 - in > /subversion/trunk/subversion/tests/cmdline/ > 3486 * 11Jan26,02:

Re: SQLite and callbacks

Stefan Sperling writes: > [[[ > Improve performance of svn proplist in a similar way as was done in r1039808. > But, this time, avoid problems with callbacks invoked during sqlite > transactions by storing results in a temporary table and invoking > callbacks during a query on the temporary table

Re: [PATCH] - Fix for issue #3792

Noorul Islam K M writes: > Daniel Shahaf writes: > >> Noorul Islam K M wrote on Wed, Feb 09, 2011 at 14:28:49 +0530: >> >>> Daniel Shahaf writes: >>> > Noorul Islam K M wrote on Sat, Feb 05, 2011 at 12:55:55 +0530: >>> >> Daniel Shahaf writes: >>> >> > Noorul Islam K M wrote on Thu, Feb 03, 20