gmane.comp.version-control.subversion.devel http://blog.gmane.org/gmane.comp.version-control.subversion.devel hourly 1 1901-01-01T00:00+00:00 Gmane http://gmane.org/img/gmane-25t.png http://gmane.org 1.6.23 up for signing/testing http://comments.gmane.org/gmane.comp.version-control.subversion.devel/142449 <pre>The 1.6.22 release artifacts are now available for testing/signing. Please get the tarballs from https://dist.apache.org/repos/dist/dev/subversion and add your signatures there. We plan to try and release on May 30th so please try and get your votes/signatures in place by May 28th EOD. Thanks! Please note the following known test failures (perhaps most easily triggered with APR-1.4.6): log_tests.py 30: log -g should ignore cyclic merges Can fail because 1.6.x is missing r1293229, which fixed occasionally missing merged-via notifications. diff_tests.py 32: repos-wc diff showing added entries with props Can fail because this test is missing various tweaks to account for random output order of 'svn diff' (this is independent of the APR 1.4.6 hash order problem). ruby swig test 'test_dump' Can fail because 1.6.x is missing r966458, which sorted property listings in dump files in alphabetical order, and this test does not parse dump files and sort properties for comparison. Side note: I did get a ruby failure I hadn't seen before but it looks like another case of sorting: [[[ 2) Failure: test_relocate(SvnWcTest) /home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/test/test_wc.rb:1007:in `test_relocate' /home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/svn/wc.rb:117:in `_open' /home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/svn/wc.rb:101:in `probe_open' /home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/test/test_wc.rb:990:in `test_relocate' /home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/svn/ra.rb:33:in `open' /home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/test/test_wc.rb:978:in `test_relocate' /home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/test/util.rb:181:in `make_context' /home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/test/test_wc.rb:975:in `test_relocate': &lt;[["0c50d5bb-1bd1-4596-8963-5d86d6b62908", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2", nil], ["0c50d5bb-1bd1-4596-8963-5d86d6b62908", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2"], ["0c50d5bb-1bd1-4596-8963-5d86d6b62908", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2/dir1", nil], ["0c50d5bb-1bd1-4596-8963-5d86d6b62908", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2/dir1", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2"], ["0c50d5bb-1bd1-4596-8963-5d86d6b62908", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2/dir1/hello.txt", nil], ["0c50d5bb-1bd1-4596-8963-5d86d6b62908", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2/dir1/hello.txt", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2"], ["0c50d5bb-1bd1-4596-8963-5d86d6b62908", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2/dir2", nil], ["0c50d5bb-1bd1-4596-8963-5d86d6b62908", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2/dir2", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2"], ["0c50d5bb-1bd1-4596-8963-5d86d6b62908", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2/dir2/hello2.txt", nil], ["0c50d5bb-1bd1-4596-8963-5d86d6b62908", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2/dir2/hello2.txt", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2"]]&gt; expected but was &lt;[["0c50d5bb-1bd1-4596-8963-5d86d6b62908", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2", nil], ["0c50d5bb-1bd1-4596-8963-5d86d6b62908", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2"], ["0c50d5bb-1bd1-4596-8963-5d86d6b62908", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2/dir2", nil], ["0c50d5bb-1bd1-4596-8963-5d86d6b62908", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2/dir2", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2"], ["0c50d5bb-1bd1-4596-8963-5d86d6b62908", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2/dir2/hello2.txt", nil], ["0c50d5bb-1bd1-4596-8963-5d86d6b62908", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2/dir2/hello2.txt", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2"], ["0c50d5bb-1bd1-4596-8963-5d86d6b62908", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2/dir1", nil], ["0c50d5bb-1bd1-4596-8963-5d86d6b62908", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2/dir1", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2"], ["0c50d5bb-1bd1-4596-8963-5d86d6b62908", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2/dir1/hello.txt", nil], ["0c50d5bb-1bd1-4596-8963-5d86d6b62908", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2/dir1/hello.txt", "file:///home/breser/subversion-1.6.23/subversion/bindings/swig/ruby/repos/dir2"]]&gt;. ]]] </pre> Ben Reser 2013-05-23T01:08:59 1.7.10 up for signing/testing http://comments.gmane.org/gmane.comp.version-control.subversion.devel/142438 <pre>The 1.7.10 release artifacts are now available for testing/signing. Please get the tarballs from https://dist.apache.org/repos/dist/dev/subversion and add your signatures there. I plan to try and release on May 30th so please try and get your votes/signatures in place by May 28th. Thanks! </pre> Stefan Sperling 2013-05-22T22:50:11 1.6.22 up for signing/testing http://comments.gmane.org/gmane.comp.version-control.subversion.devel/142436 <pre>The 1.6.22 release artifacts are now available for testing/signing. Please get the tarballs from https://dist.apache.org/repos/dist/dev/subversion and add your signatures there. I plan to try and release on May 30th so please try and get your votes/signatures in place by May 28th EOD (we need to wait for mirrors to sync for one day before release). Thanks! Please note the following known test failures (perhaps most easily triggered with APR-1.4.6): log_tests.py 30: log -g should ignore cyclic merges Can fail because 1.6.x is missing r1293229, which fixed occasionally missing merged-via notifications. diff_tests.py 32: repos-wc diff showing added entries with props Can fail because this test is missing various tweaks to account for random output order of 'svn diff' (this is independent of the APR 1.4.6 hash order problem). ruby swig test 'test_dump' Can fail because 1.6.x is missing r966458, which sorted property listings in dump files in alphabetical order, and this test does not parse dump files and sort properties for comparison. </pre> Stefan Sperling 2013-05-22T22:40:30 Tracing the client-side authn subsystem's work. http://comments.gmane.org/gmane.comp.version-control.subversion.devel/142433 <pre>Playing with my 'auth-notification' branch a few minutes ago, I enabled the notifications (via an environment variable) and ran an update command across all the items in my ~/projects directory. The resulting notification-filled trace is kinda cool. Nice to finally see where my authn subsystem interactions are happening for once! $ SVN_AUTH_NOTIFY=yes svn up * --config-dir=/tmp/foo Updating 'asf-committers': Credentials acquired (Plaintext (SSL server trust)) Credentials acquired (Gnome Keyring (simple)) At revision 41815. Updating 'cmpilato': Credentials acquired (Plaintext (SSL server trust)) Credentials acquired (Gnome Keyring (simple)) At revision 23. Updating 'dist-subversion': Credentials acquired (Plaintext (SSL server trust)) At revision 2033. Updating 'private-subversion': Credentials acquired (Auth Baton Cache) Credentials acquired (Auth Baton Cache) At revision 41815. Updating 'serf': Credentials acquired (Plaintext (SSL server trust)) At revision 1869. Skipped 'spec.subversion' Skipped 'spec.viewvc' Updating 'subversion': Credentials acquired (Auth Baton Cache) At revision 1485359. Updating 'svnbook': Credentials acquired (Plaintext (SSL server trust)) Credentials acquired (Auth Baton Cache) Credentials acquired (Auth Baton Cache) Credentials acquired (Auth Baton Cache) Credentials acquired (Auth Baton Cache) Fetching external item into 'svnbook/branches/1.4/ru/nb-bin': Credentials acquired (Auth Baton Cache) Credentials acquired (Auth Baton Cache) Credentials acquired (Auth Baton Cache) Credentials acquired (Auth Baton Cache) External at revision 2339. At revision 4500. Updating 'svnedge': Credentials acquired (Plaintext (SSL server trust)) Authentication realm: &lt;https://ctf.open.collab.net:443&gt; Authorization R ealm Password for 'cmpilato': Credentials acquired (Plaintext (prompt)) Credentials stored (Gnome Keyring (simple)) At revision 3592. Updating 'teamforge': Credentials acquired (Auth Baton Cache) Credentials acquired (Auth Baton Cache) Credentials acquired (Auth Baton Cache) Credentials acquired (Auth Baton Cache) Fetching external item into 'teamforge/trunk/core/mod_authnz_ctf/source': Credentials acquired (Auth Baton Cache) Credentials acquired (Auth Baton Cache) External at revision 518. At revision 59618. Updating 'viewvc': Authentication realm: &lt;http://viewvc.tigris.org:80&gt; CollabNet Subversion Repository Password for 'cmpilato': Credentials acquired (Plaintext (prompt)) Credentials stored (Gnome Keyring (simple)) Credentials acquired (Auth Baton Cache) Credentials stored (Gnome Keyring (simple)) Fetching external item into 'viewvc/tags/1.1.18/templates-contrib': Credentials acquired (Auth Baton Cache) Credentials stored (Gnome Keyring (simple)) External at revision 2847. Credentials acquired (Auth Baton Cache) Credentials stored (Gnome Keyring (simple)) Fetching external item into 'viewvc/tags/1.1.19/templates-contrib': Credentials acquired (Auth Baton Cache) Credentials stored (Gnome Keyring (simple)) External at revision 2883. Credentials acquired (Auth Baton Cache) Credentials stored (Gnome Keyring (simple)) Fetching external item into 'viewvc/branches/1.0.x/templates-contrib': Credentials acquired (Auth Baton Cache) Credentials stored (Gnome Keyring (simple)) External at revision 2904. Credentials acquired (Auth Baton Cache) Credentials stored (Gnome Keyring (simple)) Fetching external item into 'viewvc/branches/1.1.x/templates-contrib': Credentials acquired (Auth Baton Cache) Credentials stored (Gnome Keyring (simple)) External at revision 2904. Credentials acquired (Auth Baton Cache) Credentials stored (Gnome Keyring (simple)) Fetching external item into 'viewvc/tags/1.1.10/templates-contrib': Credentials acquired (Auth Baton Cache) Credentials stored (Gnome Keyring (simple)) External at revision 2534. Credentials acquired (Auth Baton Cache) Credentials stored (Gnome Keyring (simple)) Fetching external item into 'viewvc/trunk/templates-contrib': Credentials acquired (Auth Baton Cache) Credentials stored (Gnome Keyring (simple)) External at revision 2904. At revision 2904. Summary of updates: Updated 'asf-committers' to r41815. Updated 'cmpilato' to r23. Updated 'dist-subversion' to r2033. Updated 'private-subversion' to r41815. Updated 'serf' to r1869. Updated 'subversion' to r1485359. Updated 'svnbook' to r4500. Updated 'svnedge' to r3592. Updated 'teamforge' to r59618. Updated 'viewvc' to r2904. Summary of conflicts: Skipped paths: 2 Of particular interest to me, though, is the handling of externals. My viewvc sparse checkout has quite a few intra-repos externals defined in it. As such, all the authn credentials should be identical. But as you can see, Subversion is repeatedly fetching -- and re-storing -- these credentials from the Gnome Keyring. Weird (to me, at least). I suspect that what's happening is this: The authn code always tries to first store credentials in the provider that provided them and, failing that, iterates over providers of the same cred kind until one of them successfully stores the creds. In this case, it was the prompt provider that first fetched the creds (because I used --config-dir to point to an empty configuration directory) and the auth baton cache that did so thereafter. But neither of those *stores* creds, so time after time Subversion tries to store them in the first available provider that *can* -- the Gnome Keyring provider. That's my guess, at least. I wonder -- is there any interest in having this functionality in our trunk and productizing it a bit? It's all callback-based, so third-party clients could hook into the notification system, too. (Cc:ing Philip especially, as he's expressed irritation at the opacity of our authn subsystem before, too.) </pre> C. Michael Pilato 2013-05-22T19:44:18 Create/Apply Patch - UTF-16 and binary support http://comments.gmane.org/gmane.comp.version-control.subversion.devel/142426 <pre>Hi, I’m requesting that someone fix the Create/Apply patch functionality of Subversion. Having spoken to Stefan Küng of TortoiseSVN fame, he mentioned that there are a few problems with the create /apply patch that need fixing on the Subversion side before we can do anything on the TortoiseSVN side. Two specific things come to mind: -UTF-16 support. -Binary format file. The use case is as follows: I have some uncommitted changes in my working copy – an added .png binary file, and a modified UTF-16 encoded windows resource-script file (.rc). I want to send changes to a friend for review before committing them to the repo. My friend should just receive the .patch file and be able to apply the patch to get to the same Working copy as is on my system. Right now, I have to think about binary and UTF-16 encoded files, and put those in a zip file. Stefan told me that someone had implemented such functionality in the past - that it did not end up getting used because the diff clients needed changing. I’m kind of surprised/disappointed that a reliable patch mechanism does not seem to exist for svn. When I do a commit, the changes are bunched up in some binary format and sent to the server. Can we not capture this kind of data in some text-encoded format, and have the patch program understand the format? Below are some archived posts about Create/Apply Patch http://svn.haxx.se/tsvnusers/archive-2011-05/0012.shtml http://svn.haxx.se/tsvn/archive-2009-04/0197.shtml Best regards, Friedrich Brunzema </pre> Friedrich.Brunzema 2013-05-22T16:29:16 over-tolerant code in run_file_install? http://comments.gmane.org/gmane.comp.version-control.subversion.devel/142409 <pre>When trying to consolidate file copying into svn_io_copy_file, I ran into the code around workqueue.c:590 which derailed my plans somewhat: /* With a single db we might want to install files in a missing directory. Simply trying this scenario on error won't do any harm and at least one user reported this problem on IRC. */ Apparently added in r1142088, the exact motivation behind the code remains somewhat unclear. Does anyone remember precisely what prompted it, and what, if anything, it is needed for today? (The "won't do any harm" and "at least one user reported this problem on IRC" parts are red flags here; those phrases suggest that the code attempts to do more than promised, and that's almost as bad as not doing enough.) The reason why it is troublesome is that it makes it hard to use svn_io_copy_file for the "atomic copy" (copy to temporary + rename) pattern in run_file_install, since the "helpful" code attempts to create some missing directories if the final rename fails, and then re- try the rename. But we want svn_io_copy_file to take care of it, because it already does an atomic copy, and it appears fairly clearly to be the right place for any file copy improvements. Of course we could use svn_io_copy_file for an atomic copy to a temporary, and then a second rename to the final destination, but that would be wasteful and silly. Why improve file copying? Because it consumes a fair bit of time and I/ O bandwidth of many operations, such as checkouts and updates. In addition, exact file copies account for almost half the disk space used by a working copy, and being able to use fast and space- conserving COW copies where available would be nice. There are also possibilities of server-side file copies (NFSv4.2, SMB2) as well as new client-side file copying APIs (win32 has CopyFileEx; copy_range or something similar may be coming to Linux). </pre> Mattias Engdegård 2013-05-20T20:23:24 Feature list for 1.9 http://comments.gmane.org/gmane.comp.version-control.subversion.devel/142391 <pre>Now that we've released a 1.8 candidate, I thought it would be a good time to share with the rest of dev&lt; at &gt; what we five subversives at WANdisco have been discussing for 1.9. This is only a very high-level list as I don't want to make this mail too long ... I'm sure we'll have enough time in Berlin to argue about specifics. A. Server-side rename tracking * probably involves implementing EV2 (with shims) which includes rename operation * change the way copy-id is generated to avoid possible unique-key collisions: o curently rename is copy+delete: (node-id,copy-id,txn-id) -&gt; (node-id,new-copy-id,new-txn-id) o a proper rename does: (node-id,copy-id,txn-id) -&gt; (node-id,copy-id,new-txn-id) o the latter conflicts with the current way that copy-ids are "inherited" in the tree hierarchy during lazy copying * server (and possibly client?) should support simple heuristics for converting copy+delete records from older wire protocol into move records (possibly part of EV2 shim layer?) * metadata structures for rename support added by "svnadmin upgrade", no dump+reload required B. Merge enhancements * general merge algorithm improvements (common algo for currently different cherrypick vs. whole-hawg merges, etc.) * merge tracks renames * possibly: redefinition of mergeinfo architecture o mergeinfo no longer based on path+revision but on node IDs o implies node IDs are exposed to the client and become part of the repos API o merginfo storage decoupled from versioned properties o don't talk to me about backward-compatibility shims ... C. Read-only support for older working copies * one of the drawbacks of WC-NG is that it does not have a defined architecture for backward compatibility, so, for starters, define that * next, define what "read-only" actually means, as it does not necessarily mean "don't change the wc.db" * implement backward-compat layers for read-only operations on 1.7 and 1.8 working copies, with intent to maintain compatibility in future releases as well D. Miscellaneous * Solution for Unicode normalization issues (client and server) o make both client and server metadata indexes normalization-agnostic o do not change the wire protocol or API constraints (e.g., neither will require paths to be normalized) * Repository cache server Stefan2 has been working on a faster multi-process shared cache as part of the FSFS enhancements, and would like to get that part into 1.9. * svnpubsub enhancements o generate events for revprop and other unversioned metadata changes o proper Pythonic packaging </pre> Branko Čibej 2013-05-20T11:32:22 [svnbench] Revision: 1484368 compiled May 20 2013, 00:21:58 on x86_64-unknown-linux-gnu http://comments.gmane.org/gmane.comp.version-control.subversion.devel/142387 <pre>1.7.0&lt; at &gt;1181106 vs. trunk&lt; at &gt;1484356 Started at Mon May 20 00:26:14 UTC 2013 *DISCLAIMER* - This tests only file://-URL access on a GNU/Linux VM. This is intended to measure changes in performance of the local working copy layer, *only*. These results are *not* generally true for everyone. Charts of this data are available at http://svn-qavm.apache.org/charts/ Averaged-total results across all runs: --------------------------------------- Compare trunk&lt; at &gt;1484356 to 1.7.0 N avg operation 21/9 0.56|-32.898 TOTAL RUN 1K/530 0.88| -0.002 add 42/18 0.79| -0.180 checkout 168/72 0.64| -0.722 commit 21/9 0.69| -0.008 copy 21/9 0.79| -0.061 delete 105/45 0.18| -3.572 info 42/18 0.52| -1.035 merge 1K/516 0.85| -0.002 mkdir 56/21 0.81| -0.002 propdel 15K/6K 0.78| -0.002 proplist 16K/6K 0.80| -0.002 propset 1K/591 0.78| -0.002 ps 42/18 1.88| +0.009 resolve 42/18 0.83| -0.033 resolved 294/126 0.69| -0.055 status 21/9 0.71| -0.313 switch 294/126 0.77| -0.157 update (legend: "1.23|+0.45" means: slower by factor 1.23 and by 0.45 seconds; factor &lt; 1 and seconds &lt; 0 means 'trunk&lt; at &gt;1484356' is faster. "2/3" means: '1.7.0' has 2 timings on record, the other has 3.) Above totals split into separate &lt;dir-levels&gt;x&lt;dir-spread&gt; runs: ---------------------------------------------------------------- Compare trunk&lt; at &gt;1484356,5x5 to 1.7.0,5x5 N avg operation 7/3 0.57|-87.761 TOTAL RUN 1K/456 0.89| -0.002 add 14/6 0.83| -0.374 checkout 56/24 0.66| -1.776 commit 7/3 0.58| -0.014 copy 7/3 0.82| -0.124 delete 35/15 0.17|-10.513 info 14/6 0.55| -2.593 merge 994/470 0.86| -0.001 mkdir 56/20 0.79| -0.002 propdel 14K/6K 0.80| -0.002 proplist 15K/6K 0.82| -0.002 propset 1K/552 0.79| -0.002 ps 14/6 3.81| +0.024 resolve 14/6 0.83| -0.083 resolved 98/42 0.72| -0.124 status 7/3 0.76| -0.674 switch 98/42 0.84| -0.268 update (legend: "1.23|+0.45" means: slower by factor 1.23 and by 0.45 seconds; factor &lt; 1 and seconds &lt; 0 means 'trunk&lt; at &gt;1484356,5x5' is faster. "2/3" means: '1.7.0,5x5' has 2 timings on record, the other has 3.) Compare trunk&lt; at &gt;1484356,100x1 to 1.7.0,100x1 N avg operation 7/3 0.51| -8.642 TOTAL RUN 196/71 0.79| -0.004 add 14/6 0.49| -0.118 checkout 56/24 0.46| -0.299 commit 7/3 0.74| -0.006 copy 7/3 0.62| -0.048 delete 35/15 0.42| -0.174 info 14/6 0.35| -0.391 merge 98/46 0.83| -0.003 mkdir 763/337 0.60| -0.005 proplist 616/273 0.62| -0.005 propset 49/33 0.64| -0.005 ps 14/6 1.40| +0.004 resolve 14/6 0.91| -0.007 resolved 98/42 0.62| -0.030 status 7/3 0.46| -0.217 switch 98/42 0.45| -0.179 update (legend: "1.23|+0.45" means: slower by factor 1.23 and by 0.45 seconds; factor &lt; 1 and seconds &lt; 0 means 'trunk&lt; at &gt;1484356,100x1' is faster. "2/3" means: '1.7.0,100x1' has 2 timings on record, the other has 3.) Compare trunk&lt; at &gt;1484356,1x100 to 1.7.0,1x100 N avg operation 7/3 0.53| -2.291 TOTAL RUN 7/3 0.58| -0.025 add 14/6 0.54| -0.047 checkout 56/24 0.49| -0.091 commit 7/3 0.80| -0.004 copy 7/3 0.60| -0.013 delete 35/15 0.63| -0.029 info 14/6 0.37| -0.122 merge 259/111 0.53| -0.006 proplist 294/126 0.55| -0.006 propset 14/6 0.53| -0.006 ps 14/6 0.86| -0.002 resolve 14/6 0.63| -0.008 resolved 98/42 0.54| -0.010 status 7/3 0.50| -0.048 switch 98/42 0.59| -0.025 update (legend: "1.23|+0.45" means: slower by factor 1.23 and by 0.45 seconds; factor &lt; 1 and seconds &lt; 0 means 'trunk&lt; at &gt;1484356,1x100' is faster. "2/3" means: '1.7.0,1x100' has 2 timings on record, the other has 3.) More detail: ------------ Timings for 1.7.0,5x5 N min max avg operation (unit is seconds) 7 192.14 212.78 202.03 TOTAL RUN 1K 0.01 1.76 0.02 add 14 0.02 4.88 2.27 checkout 56 1.07 17.29 5.26 commit 7 0.01 0.13 0.03 copy 7 0.61 0.80 0.70 delete 35 6.32 31.69 12.65 info 14 5.33 8.49 5.72 merge 994 0.01 0.04 0.01 mkdir 56 0.01 0.02 0.01 propdel 14K 0.01 0.03 0.01 proplist 15K 0.01 0.06 0.01 propset 1K 0.01 0.03 0.01 ps 14 0.01 0.01 0.01 resolve 14 0.40 0.59 0.49 resolved 98 0.17 0.88 0.44 status 7 2.71 2.88 2.78 switch 98 0.21 4.85 1.64 update -- Timings for trunk&lt; at &gt;1484356,5x5 N min max avg operation (unit is seconds) 3 110.04 119.77 114.27 TOTAL RUN 456 0.01 1.41 0.02 add 6 0.02 4.04 1.89 checkout 24 0.57 8.36 3.48 commit 3 0.02 0.02 0.02 copy 3 0.53 0.61 0.57 delete 15 0.61 4.66 2.13 info 6 2.34 4.23 3.12 merge 470 0.01 0.02 0.01 mkdir 20 0.01 0.01 0.01 propdel 6K 0.01 0.02 0.01 proplist 6K 0.01 0.03 0.01 propset 552 0.01 0.03 0.01 ps 6 0.03 0.04 0.03 resolve 6 0.30 0.55 0.41 resolved 42 0.12 0.54 0.31 status 3 2.08 2.13 2.11 switch 42 0.20 3.90 1.38 update -- Compare trunk&lt; at &gt;1484356,5x5 to 1.7.0,5x5 N min max avg operation 7/3 0.57|-82.093 0.56|-93.007 0.57|-87.761 TOTAL RUN 1K/456 0.91| -0.001 0.80| -0.352 0.89| -0.002 add 14/6 0.93| -0.001 0.83| -0.839 0.83| -0.374 checkout 56/24 0.53| -0.503 0.48| -8.934 0.66| -1.776 commit 7/3 1.22| +0.003 0.15| -0.114 0.58| -0.014 copy 7/3 0.86| -0.088 0.76| -0.195 0.82| -0.124 delete 35/15 0.10| -5.719 0.15|-27.030 0.17|-10.513 info 14/6 0.44| -2.989 0.50| -4.264 0.55| -2.593 merge 994/470 0.91| -0.001 0.50| -0.020 0.86| -0.001 mkdir 56/20 0.88| -0.001 0.38| -0.012 0.79| -0.002 propdel 14K/6K 0.83| -0.001 0.49| -0.017 0.80| -0.002 proplist 15K/6K 0.84| -0.001 0.41| -0.038 0.82| -0.002 propset 1K/552 0.86| -0.001 0.90| -0.003 0.79| -0.002 ps 14/6 3.03| +0.017 4.41| +0.031 3.81| +0.024 resolve 14/6 0.74| -0.102 0.93| -0.043 0.83| -0.083 resolved 98/42 0.69| -0.053 0.61| -0.342 0.72| -0.124 status 7/3 0.77| -0.629 0.74| -0.751 0.76| -0.674 switch 98/42 0.94| -0.012 0.80| -0.951 0.84| -0.268 update (legend: "1.23|+0.45" means: slower by factor 1.23 and by 0.45 seconds; factor &lt; 1 and seconds &lt; 0 means 'trunk&lt; at &gt;1484356,5x5' is faster. "2/3" means: '1.7.0,5x5' has 2 timings on record, the other has 3.) Timings for 1.7.0,100x1 N min max avg operation (unit is seconds) 7 13.04 22.49 17.55 TOTAL RUN 196 0.01 0.13 0.02 add 14 0.02 0.74 0.23 checkout 56 0.12 1.41 0.56 commit 7 0.02 0.03 0.02 copy 7 0.09 0.22 0.13 delete 35 0.15 0.80 0.30 info 14 0.46 0.91 0.60 merge 98 0.01 0.03 0.01 mkdir 763 0.01 0.03 0.01 proplist 616 0.01 0.03 0.01 propset 49 0.01 0.03 0.01 ps 14 0.01 0.02 0.01 resolve 14 0.05 0.11 0.07 resolved 98 0.04 0.20 0.08 status 7 0.33 0.65 0.40 switch 98 0.04 0.86 0.33 update -- Timings for trunk&lt; at &gt;1484356,100x1 N min max avg operation (unit is seconds) 3 8.66 9.17 8.91 TOTAL RUN 71 0.01 0.10 0.01 add 6 0.02 0.22 0.11 checkout 24 0.09 0.52 0.26 commit 3 0.02 0.02 0.02 copy 3 0.08 0.08 0.08 delete 15 0.10 0.19 0.13 info 6 0.19 0.24 0.21 merge 46 0.01 0.02 0.01 mkdir 1 0.01 0.01 0.01 propdel 337 0.01 0.01 0.01 proplist 273 0.01 0.01 0.01 propset 33 0.01 0.01 0.01 ps 6 0.01 0.01 0.01 resolve 6 0.06 0.07 0.06 resolved 42 0.03 0.07 0.05 status 3 0.18 0.19 0.18 switch 42 0.04 0.28 0.15 update -- Compare trunk&lt; at &gt;1484356,100x1 to 1.7.0,100x1 N min max avg operation 7/3 0.66| -4.381 0.41|-13.322 0.51| -8.642 TOTAL RUN 196/71 0.92| -0.001 0.78| -0.028 0.79| -0.004 add 14/6 0.91| -0.002 0.29| -0.520 0.49| -0.118 checkout 56/24 0.79| -0.025 0.37| -0.897 0.46| -0.299 commit 7/3 0.96| -0.001 0.54| -0.015 0.74| -0.006 copy 7/3 0.83| -0.016 0.37| -0.137 0.62| -0.048 delete 35/15 0.65| -0.052 0.24| -0.611 0.42| -0.174 info 14/6 0.40| -0.275 0.26| -0.674 0.35| -0.391 merge 98/46 0.91| -0.001 0.57| -0.014 0.83| -0.003 mkdir 763/337 0.82| -0.001 0.36| -0.017 0.60| -0.005 proplist 616/273 0.84| -0.001 0.36| -0.020 0.62| -0.005 propset 49/33 0.82| -0.002 0.36| -0.020 0.64| -0.005 ps 14/6 1.47| +0.004 0.72| -0.005 1.40| +0.004 resolve 14/6 1.28| +0.014 0.60| -0.044 0.91| -0.007 resolved 98/42 0.72| -0.011 0.35| -0.131 0.62| -0.030 status 7/3 0.55| -0.151 0.29| -0.463 0.46| -0.217 switch 98/42 0.90| -0.004 0.33| -0.575 0.45| -0.179 update (legend: "1.23|+0.45" means: slower by factor 1.23 and by 0.45 seconds; factor &lt; 1 and seconds &lt; 0 means 'trunk&lt; at &gt;1484356,100x1' is faster. "2/3" means: '1.7.0,100x1' has 2 timings on record, the other has 3.) Timings for 1.7.0,1x100 N min max avg operation (unit is seconds) 7 3.09 7.20 4.88 TOTAL RUN 7 0.04 0.09 0.06 add 14 0.02 0.27 0.10 checkout 56 0.03 0.56 0.18 commit 7 0.01 0.03 0.02 copy 7 0.02 0.05 0.03 delete 35 0.04 0.18 0.08 info 14 0.12 0.30 0.19 merge 259 0.01 0.02 0.01 proplist 294 0.01 0.02 0.01 propset 14 0.01 0.02 0.01 ps 14 0.01 0.02 0.01 resolve 14 0.01 0.04 0.02 resolved 98 0.01 0.05 0.02 status 7 0.06 0.15 0.10 switch 98 0.02 0.25 0.06 update -- Timings for trunk&lt; at &gt;1484356,1x100 N min max avg operation (unit is seconds) 3 2.57 2.62 2.59 TOTAL RUN 3 0.03 0.03 0.03 add 6 0.02 0.09 0.06 checkout 24 0.02 0.20 0.09 commit 3 0.02 0.02 0.02 copy 3 0.02 0.02 0.02 delete 15 0.04 0.08 0.05 info 6 0.06 0.09 0.07 merge 111 0.01 0.01 0.01 proplist 126 0.01 0.01 0.01 propset 6 0.01 0.01 0.01 ps 6 0.01 0.01 0.01 resolve 6 0.01 0.02 0.01 resolved 42 0.01 0.02 0.01 status 3 0.05 0.05 0.05 switch 42 0.02 0.09 0.03 update -- Compare trunk&lt; at &gt;1484356,1x100 to 1.7.0,1x100 N min max avg operation 7/3 0.83| -0.522 0.36| -4.582 0.53| -2.291 TOTAL RUN 7/3 0.91| -0.003 0.39| -0.055 0.58| -0.025 add 14/6 0.92| -0.001 0.35| -0.174 0.54| -0.047 checkout 56/24 0.68| -0.011 0.35| -0.359 0.49| -0.091 commit 7/3 1.12| +0.002 0.61| -0.012 0.80| -0.004 copy 7/3 0.92| -0.002 0.41| -0.028 0.60| -0.013 delete 35/15 1.00| -0.000 0.42| -0.106 0.63| -0.029 info 14/6 0.49| -0.060 0.29| -0.211 0.37| -0.122 merge 259/111 0.83| -0.001 0.43| -0.010 0.53| -0.006 proplist 294/126 0.86| -0.001 0.70| -0.006 0.55| -0.006 propset 14/6 0.85| -0.001 0.36| -0.013 0.53| -0.006 ps 14/6 1.28| +0.002 0.63| -0.006 0.86| -0.002 resolve 14/6 0.84| -0.002 0.48| -0.018 0.63| -0.008 resolved 98/42 0.83| -0.002 0.36| -0.029 0.54| -0.010 status 7/3 0.79| -0.013 0.33| -0.098 0.50| -0.048 switch 98/42 0.99| -0.000 0.36| -0.164 0.59| -0.025 update (legend: "1.23|+0.45" means: slower by factor 1.23 and by 0.45 seconds; factor &lt; 1 and seconds &lt; 0 means 'trunk&lt; at &gt;1484356,1x100' is faster. "2/3" means: '1.7.0,1x100' has 2 timings on record, the other has 3.) Timings for 1.7.0 N min max avg operation (unit is seconds) 21 3.09 212.78 74.82 TOTAL RUN 1K 0.01 1.76 0.02 add 42 0.02 4.88 0.87 checkout 168 0.03 17.29 2.00 commit 21 0.01 0.13 0.03 copy 21 0.02 0.80 0.29 delete 105 0.04 31.69 4.34 info 42 0.12 8.49 2.17 merge 1K 0.01 0.04 0.01 mkdir 56 0.01 0.02 0.01 propdel 15K 0.01 0.03 0.01 proplist 16K 0.01 0.06 0.01 propset 1K 0.01 0.03 0.01 ps 42 0.01 0.02 0.01 resolve 42 0.01 0.59 0.20 resolved 294 0.01 0.88 0.18 status 21 0.06 2.88 1.09 switch 294 0.02 4.85 0.68 update -- Timings for trunk&lt; at &gt;1484356 N min max avg operation (unit is seconds) 9 2.57 119.77 41.92 TOTAL RUN 530 0.01 1.41 0.02 add 18 0.02 4.04 0.69 checkout 72 0.02 8.36 1.28 commit 9 0.02 0.02 0.02 copy 9 0.02 0.61 0.22 delete 45 0.04 4.66 0.77 info 18 0.06 4.23 1.13 merge 516 0.01 0.02 0.01 mkdir 21 0.01 0.01 0.01 propdel 6K 0.01 0.02 0.01 proplist 6K 0.01 0.03 0.01 propset 591 0.01 0.03 0.01 ps 18 0.01 0.04 0.02 resolve 18 0.01 0.55 0.16 resolved 126 0.01 0.54 0.12 status 9 0.05 2.13 0.78 switch 126 0.02 3.90 0.52 update -- Compare trunk&lt; at &gt;1484356 to 1.7.0 N min max avg operation 21/9 0.83| -0.522 0.56|-93.007 0.56|-32.898 TOTAL RUN 1K/530 0.91| -0.001 0.80| -0.352 0.88| -0.002 add 42/18 0.91| -0.002 0.83| -0.839 0.79| -0.180 checkout 168/72 0.68| -0.011 0.48| -8.934 0.64| -0.722 commit 21/9 1.11| +0.001 0.15| -0.114 0.69| -0.008 copy 21/9 0.92| -0.002 0.76| -0.195 0.79| -0.061 delete 105/45 1.00| -0.000 0.15|-27.030 0.18| -3.572 info 42/18 0.49| -0.060 0.50| -4.264 0.52| -1.035 merge 1K/516 0.91| -0.001 0.50| -0.020 0.85| -0.002 mkdir 56/21 0.88| -0.001 0.52| -0.010 0.81| -0.002 propdel 15K/6K 0.83| -0.001 0.49| -0.017 0.78| -0.002 proplist 16K/6K 0.84| -0.001 0.41| -0.038 0.80| -0.002 propset 1K/591 0.86| -0.001 0.89| -0.003 0.78| -0.002 ps 42/18 1.28| +0.002 2.21| +0.022 1.88| +0.009 resolve 42/18 0.84| -0.002 0.93| -0.043 0.83| -0.033 resolved 294/126 0.83| -0.002 0.61| -0.342 0.69| -0.055 status 21/9 0.79| -0.013 0.74| -0.751 0.71| -0.313 switch 294/126 0.99| -0.000 0.80| -0.951 0.77| -0.157 update (legend: "1.23|+0.45" means: slower by factor 1.23 and by 0.45 seconds; factor &lt; 1 and seconds &lt; 0 means 'trunk&lt; at &gt;1484356' is faster. "2/3" means: '1.7.0' has 2 timings on record, the other has 3.) Had started at Mon May 20 00:26:14 UTC 2013, done at Mon May 20 00:36:36 UTC 2013 wrote chart file: /home/neels/svnbench/charts/.busy/compare_1.7.0_trunk&lt; at &gt;last12.svg wrote chart file: /home/neels/svnbench/charts/.busy/compare_1.7.0_trunk&lt; at &gt;last12,5x5.svg wrote chart file: /home/neels/svnbench/charts/.busy/compare_1.7.0_trunk&lt; at &gt;last12,1x100.svg wrote chart file: /home/neels/svnbench/charts/.busy/compare_1.7.0_trunk&lt; at &gt;last12,100x1.svg </pre> neels< at >apache.org 2013-05-20T00:37:07 Review of invoke-diff-cmd-feature branch http://comments.gmane.org/gmane.comp.version-control.subversion.devel/142383 <pre>Review in diff form: Index: subversion/include/svn_client.h =================================================================== --- subversion/include/svn_client.h(revision 1484305) +++ subversion/include/svn_client.h(working copy) &lt; at &gt;&lt; at &gt; -2988,8 +2988,8 &lt; at &gt;&lt; at &gt; svn_client_blame(const char *path_or_url, * * &lt; at &gt;a invoke_diff_cmd is used to call an external diff program but may * not be &lt; at &gt;c NULL. The command line invocation will override the - * invoke-diff-cmd invocation entry(if any) in the Subversion - * configuration file. + * invoke-diff-cmd invocation entry(if any) in &lt; at &gt;c ctx-&gt;config. + * ### "May not be NULL" !? log-cmd.c and deprecated.c pass NULL for it. * * Generated headers are encoded using &lt; at &gt;a header_encoding. * &lt; at &gt;&lt; at &gt; -3047,7 +3047,8 &lt; at &gt;&lt; at &gt; svn_client_diff7(const apr_array_header_t *options svn_client_ctx_t *ctx, apr_pool_t *pool); -/** Similar to svn_client_diff7(), but with &lt; at &gt;a invoke_diff_cmd. +/** Similar to svn_client_diff7(), but with &lt; at &gt;a invoke_diff_cmd + * set to &lt; at &gt;c NULL. * * &lt; at &gt;deprecated Provided for backward compatibility with the 1.8 API. * &lt; at &gt;since New in 1.8. &lt; at &gt;&lt; at &gt; -3245,6 +3246,7 &lt; at &gt;&lt; at &gt; svn_client_diff_peg7(const apr_array_header_t *dif /** Similar to svn_client_peg5(), but with &lt; at &gt;a no_diff_added set to * FALSE, &lt; at &gt;a ignore_properties set to FALSE and &lt; at &gt;a properties_only * set to false. + * ### copy-pasto? ^^ * * &lt; at &gt;deprecated Provided for backward compatibility with the 1.7 API. * &lt; at &gt;since New in 1.9. Index: subversion/include/svn_io.h =================================================================== --- subversion/include/svn_io.h(revision 1484305) +++ subversion/include/svn_io.h(working copy) &lt; at &gt;&lt; at &gt; -2281,6 +2281,8 &lt; at &gt;&lt; at &gt; svn_io_file_readline(apr_file_t *file, /** Parse a user defined command to contain dynamically created labels * and filenames. * + * &lt; at &gt;todo Document the parameters and return value. + * * &lt; at &gt;since New in 1.9. */ const char ** &lt; at &gt;&lt; at &gt; -2295,6 +2297,8 &lt; at &gt;&lt; at &gt; svn_io_create_custom_diff_cmd(const char *label1, /** Run the external diff command defined by the invoke-diff-cmd * option. + * + * &lt; at &gt;todo Document the parameters and return value. * * &lt; at &gt;since New in 1.9. */ Index: subversion/libsvn_client/diff.c =================================================================== --- subversion/libsvn_client/diff.c(revision 1484305) +++ subversion/libsvn_client/diff.c(working copy) &lt; at &gt;&lt; at &gt; -790,6 +790,7 &lt; at &gt;&lt; at &gt; diff_content_changed(svn_boolean_t *wrote_header, } + /* ### Should this code error if both are set? */ if (diff_cmd_baton-&gt;diff_cmd || diff_cmd_baton-&gt;invoke_diff_cmd) { apr_file_t *outfile; &lt; at &gt;&lt; at &gt; -821,6 +822,10 &lt; at &gt;&lt; at &gt; diff_content_changed(svn_boolean_t *wrote_header, scratch_pool, scratch_pool)); if (diff_cmd_baton-&gt;diff_cmd) + /* ### "." is not canonical, and the docstring doesn't bless passing + a non-canonical dirent. (See svn_dirent_canonicalize()) + Should this pass "" (canonicalized "."), or should svn_io.h + bless passing non-canonical dirents (at least in this case)? */ SVN_ERR(svn_io_run_diff2(".", diff_cmd_baton-&gt;options.for_external.argv, diff_cmd_baton-&gt;options.for_external.argc, &lt; at &gt;&lt; at &gt; -2522,6 +2527,7 &lt; at &gt;&lt; at &gt; set_up_diff_cmd_and_options(struct diff_cmd_baton argv = apr_palloc(pool, argc * sizeof(char *)); for (i = 0; i &lt; argc; i++) SVN_ERR(svn_utf_cstring_to_utf8(&amp;argv[i], + /* next line shouldn't have been un-indented. */ APR_ARRAY_IDX(options, i, const char *), pool)); } diff_cmd_baton-&gt;options.for_external.argv = argv; Index: subversion/libsvn_subr/config_file.c =================================================================== --- subversion/libsvn_subr/config_file.c(revision 1484305) +++ subversion/libsvn_subr/config_file.c(working copy) &lt; at &gt;&lt; at &gt; -1086,6 +1086,9 &lt; at &gt;&lt; at &gt; svn_config_ensure(const char *config_dir, apr_pool "# diff3-has-program-arg = [yes | no]" NL "### Set invoke-diff-cmd to the absolute path of your 'diff'" NL "### program." NL + /* ### Document the replaceables */ + /* ### Document how the setting may contain argv too, + not only argv[0] */ "### This will override the compile-time default, which is to use" NL "### Subversion's internal diff implementation." NL "# invoke-diff-cmd = \"diff -y --label %l1% %f1% --label %l2% %f2%\""NL Index: subversion/libsvn_subr/io.c =================================================================== --- subversion/libsvn_subr/io.c(revision 1484305) +++ subversion/libsvn_subr/io.c(working copy) &lt; at &gt;&lt; at &gt; -2936,16 +2936,22 &lt; at &gt;&lt; at &gt; svn_io_create_custom_diff_cmd(const char *label1, const char ** ret; int argv; + /* Just take two pool arguments instead? (result+scratch) */ subpool = svn_pool_create(pool); + /* ### space after comma */ + /* This reimplements "split a string into argv". Is there an existing + svn/apr function that does that can be reused here? (We might gain + an "escape spaces" functionality this way.) */ tmp = svn_cstring_split(cmd," ",TRUE, subpool); + /* ### s/calloc/alloc/; you initialize everything explicitly. */ ret = apr_pcalloc(pool, tmp-&gt;nelts * tmp-&gt;elt_size*sizeof(char *)); for (argv = 0; argv &lt; tmp-&gt;nelts ; argv++) - { + { /* needs to be indented two spaces, not one */ svn_stringbuf_t *com; int i; &lt; at &gt;&lt; at &gt; -2969,6 +2975,14 &lt; at &gt;&lt; at &gt; svn_io_create_custom_diff_cmd(const char *label1, { len = strlen(found); + /* ### Looks dubious. Won't this access com-&gt;data[-1] when + !strcmp(com-&gt;data, token_list[0]) ? + How does this parse "%%%f1%"? Is "%%f1%%" an error? + (The answers to both of these questions should have been + decided prior to coding, and I recall a list thread but I + don't recall that thread reached consensus on a specific + escaping UI.) Has consensus on the UI implemented here + been reached? */ /* if we find a % in front of this, consume it */ if (com-&gt;data[com-&gt;len - strlen(found)-1] == '%') svn_stringbuf_remove(com, strlen(found)-1, 1); &lt; at &gt;&lt; at &gt; -2993,8 +3007,8 &lt; at &gt;&lt; at &gt; svn_io_create_custom_diff_cmd(const char *label1, } } ret[argv] = com-&gt;data; - } - ret[argv] = NULL; + } + ret[argv] = NULL; /* segfault. need (nelts + 1) in the allocation */ svn_pool_destroy(subpool); return ret; &lt; at &gt;&lt; at &gt; -3018,6 +3032,7 &lt; at &gt;&lt; at &gt; svn_io_run_external_diff(const char *dir, if (0 == strlen(external_diff_cmd)) return svn_error_createf(SVN_ERR_CL_INSUFFICIENT_ARGS, NULL, _("The --invoke-diff-cmd string was empty.\n")); + /* ### No newline in error messages */ cmd = svn_io_create_custom_diff_cmd(label1, label2, NULL, tmpfile1, tmpfile2, NULL, &lt; at &gt;&lt; at &gt; -3025,6 +3040,7 &lt; at &gt;&lt; at &gt; svn_io_run_external_diff(const char *dir, if (pexitcode == NULL) pexitcode = &amp;exitcode; + /* if invoke_diff_cmd is "", cmd[0] would be NULL? */ SVN_ERR(svn_io_run_cmd(dir, cmd[0], cmd, pexitcode, NULL, TRUE, NULL, outfile, errfile, pool)); &lt; at &gt;&lt; at &gt; -3036,10 +3052,10 &lt; at &gt;&lt; at &gt; svn_io_run_external_diff(const char *dir, for (i = 0, size = 0; cmd[i]; i++) size += strlen(cmd[i]) + 1; - failed_command = apr_pcalloc(pool, size * sizeof(char *)); + failed_command = apr_palloc(pool, size * sizeof(char)); for (i = 0; cmd[i]; i++) - { + { /* wrong indent */ strcat(failed_command, cmd[i]); strcat(failed_command, " "); } &lt; at &gt;&lt; at &gt; -3048,6 +3064,9 &lt; at &gt;&lt; at &gt; svn_io_run_external_diff(const char *dir, _("'%s' was expanded to '%s' and returned %d\n"), external_diff_cmd, svn_dirent_local_style(failed_command, pool), + /* ### Huh? failed_comand is dirent, it's + a string those first " "-separated word + is a dirent. */ *pexitcode); } return SVN_NO_ERROR; Index: subversion/svn/log-cmd.c =================================================================== --- subversion/svn/log-cmd.c(revision 1484305) +++ subversion/svn/log-cmd.c(working copy) &lt; at &gt;&lt; at &gt; -140,7 +140,7 &lt; at &gt;&lt; at &gt; display_diff(const svn_log_entry_t *log_entry, outstream, errstream, NULL, - NULL, + NULL, /* ### why not pass invoke_diff_cmd? */ - ctx, pool)); + ctx, pool)); /* s/tabs/spaces/ */ SVN_ERR(svn_stream_puts(outstream, _("\n"))); return SVN_NO_ERROR; Index: subversion/svn/svn.c =================================================================== --- subversion/svn/svn.c(revision 1484305) +++ subversion/svn/svn.c(working copy) &lt; at &gt;&lt; at &gt; -2751,8 +2751,12 &lt; at &gt;&lt; at &gt; sub_main(int argc, const char *argv[], apr_pool_t svn_config_set(cfg_config, SVN_CONFIG_SECTION_HELPERS, SVN_CONFIG_OPTION_DIFF_CMD, opt_state.diff.diff_cmd); } - else + else { + /* ### Why not set this unconditionally and let the library figure out + which option takes precedence? Or, alternatively, error out + if --diff-cmd and --invoke-diff-cmd both specified (to avoid + silently ignoring the latter)? */ if (opt_state.diff.invoke_diff_cmd) svn_config_set(cfg_config, SVN_CONFIG_SECTION_HELPERS, SVN_CONFIG_OPTION_INVOKE_DIFF_CMD, opt_state.diff.invoke_diff_cmd); Index: subversion/tests/cmdline/diff_tests.py =================================================================== --- subversion/tests/cmdline/diff_tests.py(revision 1484305) +++ subversion/tests/cmdline/diff_tests.py(working copy) &lt; at &gt;&lt; at &gt; -3266,6 +3266,8 &lt; at &gt;&lt; at &gt; def diff_invoke_external_diffcmd(sbox): diff_script_path = os.path.abspath(".")+"/diff" + # wrap to 80 columns, align second argument, + # consider printing repr() in the hook for easier parsing svntest.main.create_python_hook_script(diff_script_path, 'import sys\n' 'for arg in sys.argv[1:]:\n print(arg)\n') if sys.platform == 'win32': </pre> Daniel Shahaf 2013-05-19T17:23:13 All the fuzz about properties and hashes http://comments.gmane.org/gmane.comp.version-control.subversion.devel/142344 <pre>Hi there, I just want to give you a bit of background for what my latest commits and others following shortly are all about: log and mergeinfo performance. This is linked to my fsfs-format7 work but parts that apply to /trunk as well have been committed there. The current state is that an 'svn log -v' is now network limited for some repos (i.e. &gt; 1Gb/s data transfer) with 'svn log -v -g' slowly catching up. All-caches-cold performance shall ultimately be close to hot caches performance. Details will be presented in Berlin. Current measurements on a fsfs-format7 mirror of our ASF repo (1,457,326 revs, 12,840,090 changes) on an HDD RAID with a 10Gb Ethernet connection between client and server: $time svn-bench null-log svn://server1/apache-f7 -v real 1m5.063s user 0m19.464s sys 0m1.972s $time svn-bench null-log svn://server1/apache-f7 -v real 0m23.276s user 0m17.424s sys 0m1.780s $(clear all caches etc.) $time svn-bench null-log svn://server1/apache-f7/subversion/trunk -v -g real 1m14.043s user 0m0.288s sys 0m0.048s $time svn-bench null-log svn://server1/apache-f7/subversion/trunk -v -g real 0m1.575s user 0m0.308s sys 0m0.040s Please note that this is with HEAD + local patches still in my commit queue. </pre> Stefan Fuhrmann 2013-05-17T09:24:13 Apache Subversion 1.8.0-rc2 Released http://comments.gmane.org/gmane.comp.version-control.subversion.devel/142338 <pre>I'm happy to announce the release of Apache Subversion 1.8.0-rc2, the first public release candidate of Subversion 1.8.0. Please choose the mirror closest to you by visiting: http://subversion.apache.org/download/#pre-releases The SHA1 checksums are: 272a32237cf297f9cca5f511e600e46ac19ab052 subversion-1.8.0-rc2.tar.bz2 9b8d319f6b556f856348a72563fd2d5005a6fa31 subversion-1.8.0-rc2.tar.gz bab03d90dc8cc27877372ba875009c5b878c9785 subversion-1.8.0-rc2.zip PGP Signatures are available at: http://www.apache.org/dist/subversion/subversion-1.8.0-rc2.tar.bz2.asc http://www.apache.org/dist/subversion/subversion-1.8.0-rc2.tar.gz.asc http://www.apache.org/dist/subversion/subversion-1.8.0-rc2.zip.asc For this release, the following people have provided PGP signatures: Ben Reser [4096R/16A0DE01] with fingerprint: 19BB CAEF 7B19 B280 A0E2 175E 62D4 8FAD 16A0 DE01 Branko Čibej [2048R/C8628501] with fingerprint: 8769 28CD 4954 EA74 87B6 B96C 29B8 92D0 C862 8501 C. Michael Pilato [4096R/FE681333] with fingerprint: 753B 2F9D F717 FA23 A43E E7C3 F5E0 F001 FE68 1333 Ivan Zhakov [4096R/F6AD8147] with fingerprint: 4829 8F0F E47F 4B8A 43FD 6525 919F 6F61 F6AD 8147 Mark Phippard [1024D/035A96A9] with fingerprint: D315 89DB E1C1 E9BA D218 39FD 265D F8A0 035A 96A9 Paul T. Burba [4096R/56F3D7BC] with fingerprint: 1A0F E7C6 B3C5 F8D4 D0C4 A20B 64DD C071 56F3 D7BC Philip Martin [2048R/ED1A599C] with fingerprint: A844 790F B574 3606 EE95 9207 76D7 88E1 ED1A 599C Stefan Sperling [2048R/9A59B973] with fingerprint: 8BC4 DAE0 C5A4 D65F 4044 0107 4F7D BAA9 9A59 B973 This is a release candidate for what will eventually become Apache Subversion 1.8.0. It is thought to be free of blocking issues, and if none are found will become the final release. For this reason, we encourage thorough testing in as many environments as possible. This release candidate begins the four-week "soak" period to allow for further testing, and barring show-stopping bugs, the final 1.8.0 release can be expected on or near June 14th. A pre-release means the Subversion developers feel that this release is ready for widespread testing by the community. There are known issues (and unknown ones!), so please use it at your own risk, though we do encourage people to test this release thoroughly. Of particular note, please remember than persistent data, such as the working copy or repository formats may change before the final release, and there may not be an upgrade path from the pre-releases to the final. As a note to operating system distro packagers: while we wish to have this release candidate widely tested, we do not feel that it is ready for packaging and providing to end-users through a distro package system. Packaging a release candidate poses many problems, the biggest being that our policy lets us break compatibility between the release candidate and the final release, if we find something serious enough. Having many users depending on a release candidate through their distro would cause no end of pain and frustration that we do not want to have to deal with. However, if your distro has a branch that is clearly labeled as containing experimental and often broken software, and explicitly destined to consenting developers and integrators only, then we're okay with packaging the release candidate there. Just don't let it near the end users please. Release notes for the 1.8.x release series may be found at: http://subversion.apache.org/docs/release-notes/1.8.html You can find the list of changes between 1.8.0-rc2 and earlier versions at: http://svn.apache.org/repos/asf/subversion/tags/1.8.0-rc2/CHANGES Questions, comments, and bug reports to users&lt; at &gt;subversion.apache.org. Thanks, - The Subversion Team </pre> Ben Reser 2013-05-16T23:01:25 Uncaught tree conflicts http://comments.gmane.org/gmane.comp.version-control.subversion.devel/142337 <pre>Hi dev, what output do you expect for these two almost similar cases? [[[ # make a simple dir with file mkdir srcdir echo x &gt; srcdir/x svn add srcdir svn ci -m1 # make a locally added dir mkdir branchdir echo yy &gt; branchdir/x svn add branchdir # create a branch behind the locally added dir svn cp -m2 ^/srcdir ^/branchdir # case (2) adds only this line: #rm -rf branchdir svn up --accept=postpone svn st ]]] The output you expect -- a tree conflict, right? -- you're not getting it. Case (1) gives a *text* conflict: [[[ + svn up --accept=postpone Updating '.': C branchdir/x E branchdir Updated to revision 2. Summary of conflicts: Text conflicts: 1 ]]] Case (2) goes completely nuts: [[[ + svn up --accept=postpone Updating '.': subversion/svn/update-cmd.c:172, subversion/libsvn_client/update.c:706, subversion/libsvn_client/update.c:614, subversion/libsvn_client/update.c:466, subversion/libsvn_wc/adm_crawler.c:845, subversion/libsvn_repos/reporter.c:1542, subversion/libsvn_repos/reporter.c:1453, subversion/libsvn_repos/reporter.c:1445, subversion/libsvn_repos/reporter.c:1383, subversion/libsvn_repos/reporter.c:1321, subversion/libsvn_repos/reporter.c:1040, subversion/libsvn_repos/reporter.c:1321, subversion/libsvn_repos/reporter.c:1074, subversion/libsvn_wc/update_editor.c:4374, subversion/libsvn_wc/update_editor.c:4071, subversion/libsvn_wc/update_editor.c:3911, subversion/libsvn_wc/merge.c:1176, subversion/libsvn_wc/merge.c:869, subversion/libsvn_wc/merge.c:414, subversion/libsvn_diff/diff_file.c:1332, subversion/libsvn_diff/diff3.c:277, subversion/libsvn_diff/diff_file.c:788, subversion/libsvn_subr/io.c:3271: (apr_err=ENOENT) svn: E000002: Can't open file '/tmp/trunk.wvg/wc/branchdir/x': No such file or directory ]]] I'll add two issues if I get plussed. ~Neels </pre> Neels Hofmeyr 2013-05-16T22:51:51 [PATCH] SVN_UNALIGNED_ACCESS_IS_OK for ppc http://comments.gmane.org/gmane.comp.version-control.subversion.devel/142327 <pre>[[[ * subversion/include/svn_types.h: Set SVN_UNALIGNED_ACCESS_IS_OK for PowerPC. ]]] Index: subversion/include/svn_types.h =================================================================== --- subversion/include/svn_types.h(revision 1483563) +++ subversion/include/svn_types.h(working copy) &lt; at &gt;&lt; at &gt; -75,7 +75,8 &lt; at &gt;&lt; at &gt; * &lt; at &gt;since New in 1.7. */ #ifndef SVN_UNALIGNED_ACCESS_IS_OK -# if defined(_M_IX86) || defined(_M_X64) || defined(i386) || defined(__x86_64) +# if defined(_M_IX86) || defined(_M_X64) || defined(i386) || defined(__x86_64) \ + || defined __powerpc__ || defined __ppc__ # define SVN_UNALIGNED_ACCESS_IS_OK 1 # else # define SVN_UNALIGNED_ACCESS_IS_OK 0 </pre> Mattias Engdegård 2013-05-16T21:12:25 1.8.0-rc2 up for testing/signing http://comments.gmane.org/gmane.comp.version-control.subversion.devel/142285 <pre>Here it is: the second Release Candidate for Subversion 1.8.0. You can fetch the proposed tarballs from here: https://dist.apache.org/repos/dist/dev/subversion The magic rev is r1482456 This is a release candidate, meaning if all goes well it (or something very much like it) will become the final release. Please test this as if it were the final release, understanding that there will be bugs in any software we ship. &lt;insert lecture about perfect being enemy of the good here&gt; To all distributors and binary packagers: This isn't the final release. It *won't* be the final release for a few weeks yet, so please don't ship it as such. In any case, don't distribute any releases based upon these tarballs until we complete our signature gathering process and officially announce the release. Reminder that the STATUS file is still for 1.8.0, please be careful about moving changes to the approved section and only move changes that appropriate to be merged at this point (i.e. changes that would not restart our soak). If we have changes we want to put into 1.8.1, it's fine to put them in the STATUS file and vote on them, just please do not move them to the approved section. It may be helpful to refer to this section of our documentation when determining what can done to the 1.8.x branch during our soak period: http://subversion.apache.org/docs/community-guide/releasing.html#release-stabilization </pre> Ben Reser 2013-05-14T19:18:27 FSFS Repository corruption on high load on Windows (was Re: svn commit: r1082451 - /subversion/trunk/subversion/libsvn_subr/io.c) http://comments.gmane.org/gmane.comp.version-control.subversion.devel/142283 <pre> [...] Hi Bert, How did you come to conclusion that Windows doesn't have race condition between write and rename? It seems Windows actually have race condition between write and move in case of unexpected computer shutdown. In this case move operation completes after restart, but file contains zero data. I understand that you intention was optimize client side operation but svn_io_write_unique() function actively used in FSFS and lead repository data corruption. I've created small program to demonstrate this problem (see attachment). On start it create four threads actively writing data to disk and one thread that uses write + move to update file atomically. If you run this program on virtual machine and then reset then with very good chance you'll get 'current' file with zeros. </pre> Ivan Zhakov 2013-05-14T17:06:22 With 1.8.0-rc1 DOA, shall we accept more approved backports? http://comments.gmane.org/gmane.comp.version-control.subversion.devel/142276 <pre>With 1.8.0-rc1 DOA, should we roll more of the 1.8.x/STATUS items into what will be rc2? There look to be quite a few that have "(1.8.1)" decorators on them, I assume because they aren't considered valuable enough to restart the soak process. But the soak is already restarted, so... Ben, can you give us a new (very near-term, please) deadline for backport approvals based on when you plan to cut rc2? </pre> C. Michael Pilato 2013-05-14T14:23:55 Nominating your r1481848 and r1481870 http://comments.gmane.org/gmane.comp.version-control.subversion.devel/142264 <pre>Hi Brane, Just saw the veto on my compiler warning commits for 1.8.x. I nominated your two commits on top of my commits (to avoid merge conflicts), which should resolve the veto, but I left it in the vetoed section to be safe. Pinging you here so you can take a look. Blair </pre> Blair Zajac 2013-05-14T06:31:51 Sparse working copy in 1.8.0 causes problems for others' merges; 1.7.x didn't complain http://comments.gmane.org/gmane.comp.version-control.subversion.devel/142258 <pre>I'm running Subversion 1.8.0-dev on Windows 7 Pro SP1. The following steps went through without error on 1.7.x, but they fail with an error on the last step when run on 1.8.0 (see below for full reproducible recipe): 1. make a copy (branch) of your trunk 2. Harry checks out the branch in full 3. Sally sparsely checks out the branch with just a subset of subtrees 4. someone adds something in trunk under one of the subtrees that Sally has excluded 5. someone removes something from trunk under the subtree added in step 4 6. Sally merges trunk into the branch (remember she has the sparse working copy) 7. Harry merges trunk into the branch BAM! Harry can't commit the merge because: svn: E155011: Commit failed (details follow): svn: E155011: Directory 'C:\testbranch1_userX\B\B1\B1a' is out of date svn: E160028: '/branches/branch1/B/B1/B1a' is out of date I suspect this has something to do with one or both of these two issues, completed in 1.8.0: - http://subversion.tigris.org/issues/show_bug.cgi?id=4305 - http://subversion.tigris.org/issues/show_bug.cgi?id=4169 Can someone confirm? Is this a new bug introduced in 1.8.0 or a correction of an oversight in 1.7.x? Either way, what is the workaround? It seems to me that a merge into a sparse working copy either shouldn't be allowed, or it should work correctly. In other words, this recipe should either fail on step 6 above (instead of 7) or it should go all the way through correctly, including step 7. Here's the complete recipe for Windows (it creates the repository from scratch, so the only assumption is that you have a C: drive): cd /d C:\ svnadmin create test_repo svn checkout file:///C:/test_repo svntestrepo cd svntestrepo md trunk cd trunk md B md C cd .. md branches svn add branches svn add trunk svn commit -m "" svn copy "^/trunk" "^/branches/branch1" -m "" svn checkout "^/branches/branch1" ..\testbranch1_userX svn checkout --depth empty "^/branches/branch1" ..\testbranch1_userY cd ..\testbranch1_userY svn update --set-depth infinity C svn checkout "^/trunk" ..\testtrunk cd ..\testtrunk\B md B1 cd B1 echo B1 test &gt; test.txt md B1a cd B1a echo B1a test &gt; test.txt cd ..\.. svn add B1 cd .. svn commit -m "" cd C md C1 cd C1 echo C1 test &gt; test.txt cd .. svn add C1 cd .. svn commit -m "" cd B\B1 svn delete B1a cd ..\.. svn commit -m "" cd ..\testbranch1_userY svn merge "^/trunk" svn commit -m "" cd ..\testbranch1_userX svn update svn merge "^/trunk" svn commit -m "" Thanks! </pre> Bob Cardillo 2013-05-14T01:58:03 Repair corrupt repository http://comments.gmane.org/gmane.comp.version-control.subversion.devel/142238 <pre>Hi I am looking for expert help in Brisbane, Australia to repair a corrupt repository. Svnadmin verify returns "Decompression of svndiff data failed" at an earlier revision. We use SVN for document management and don't have the expertise and time internally to resolve. Regards Brad Waldon / Director Waldon Services Pty Ltd E: brad&lt; at &gt;waldonservices.com&lt;mailto:brad&lt; at &gt;waldonservices.com&gt; W: www.waldonservices.com&lt;http://www.waldonservices.com/&gt; </pre> Brad Waldon 2013-05-13T08:31:51 [svnbench] Revision: 1481674 compiled May 13 2013, 00:22:05 on x86_64-unknown-linux-gnu http://comments.gmane.org/gmane.comp.version-control.subversion.devel/142236 <pre>1.7.0&lt; at &gt;1181106 vs. trunk&lt; at &gt;1481633 Started at Mon May 13 00:26:11 UTC 2013 *DISCLAIMER* - This tests only file://-URL access on a GNU/Linux VM. This is intended to measure changes in performance of the local working copy layer, *only*. These results are *not* generally true for everyone. Charts of this data are available at http://svn-qavm.apache.org/charts/ Averaged-total results across all runs: --------------------------------------- Compare trunk&lt; at &gt;1481633 to 1.7.0 N avg operation 18/9 0.55|-33.711 TOTAL RUN 1K/530 0.86| -0.003 add 36/18 0.76| -0.207 checkout 144/72 0.67| -0.649 commit 18/9 0.87| -0.003 copy 18/9 0.82| -0.049 delete 90/45 0.15| -3.703 info 36/18 0.51| -1.065 merge 936/516 0.80| -0.002 mkdir 48/21 0.83| -0.002 propdel 13K/6K 0.73| -0.002 proplist 13K/6K 0.78| -0.002 propset 1K/591 0.75| -0.003 ps 36/18 1.82| +0.008 resolve 36/18 0.83| -0.033 resolved 252/126 0.70| -0.053 status 18/9 0.70| -0.331 switch 252/126 0.77| -0.157 update (legend: "1.23|+0.45" means: slower by factor 1.23 and by 0.45 seconds; factor &lt; 1 and seconds &lt; 0 means 'trunk&lt; at &gt;1481633' is faster. "2/3" means: '1.7.0' has 2 timings on record, the other has 3.) Above totals split into separate &lt;dir-levels&gt;x&lt;dir-spread&gt; runs: ---------------------------------------------------------------- Compare trunk&lt; at &gt;1481633,5x5 to 1.7.0,5x5 N avg operation 6/3 0.56|-89.059 TOTAL RUN 942/456 0.88| -0.002 add 12/6 0.81| -0.436 checkout 48/24 0.71| -1.505 commit 6/3 1.47| +0.007 copy 6/3 0.88| -0.079 delete 30/15 0.14|-10.888 info 12/6 0.54| -2.651 merge 852/470 0.81| -0.002 mkdir 48/20 0.82| -0.002 propdel 12K/6K 0.74| -0.002 proplist 12K/6K 0.80| -0.002 propset 1K/552 0.76| -0.002 ps 12/6 3.85| +0.024 resolve 12/6 0.84| -0.081 resolved 84/42 0.73| -0.115 status 6/3 0.75| -0.709 switch 84/42 0.85| -0.253 update (legend: "1.23|+0.45" means: slower by factor 1.23 and by 0.45 seconds; factor &lt; 1 and seconds &lt; 0 means 'trunk&lt; at &gt;1481633,5x5' is faster. "2/3" means: '1.7.0,5x5' has 2 timings on record, the other has 3.) Compare trunk&lt; at &gt;1481633,100x1 to 1.7.0,100x1 N avg operation 6/3 0.48| -9.469 TOTAL RUN 168/71 0.77| -0.004 add 12/6 0.46| -0.130 checkout 48/24 0.42| -0.339 commit 6/3 0.66| -0.008 copy 6/3 0.59| -0.055 delete 30/15 0.40| -0.189 info 12/6 0.33| -0.411 merge 84/46 0.79| -0.003 mkdir 654/337 0.58| -0.005 proplist 528/273 0.59| -0.006 propset 42/33 0.61| -0.006 ps 12/6 1.34| +0.003 resolve 12/6 0.88| -0.009 resolved 84/42 0.59| -0.034 status 6/3 0.44| -0.231 switch 84/42 0.45| -0.189 update (legend: "1.23|+0.45" means: slower by factor 1.23 and by 0.45 seconds; factor &lt; 1 and seconds &lt; 0 means 'trunk&lt; at &gt;1481633,100x1' is faster. "2/3" means: '1.7.0,100x1' has 2 timings on record, the other has 3.) Compare trunk&lt; at &gt;1481633,1x100 to 1.7.0,1x100 N avg operation 6/3 0.50| -2.604 TOTAL RUN 6/3 0.54| -0.029 add 12/6 0.50| -0.054 checkout 48/24 0.45| -0.103 commit 6/3 0.69| -0.007 copy 6/3 0.56| -0.015 delete 30/15 0.59| -0.033 info 12/6 0.35| -0.133 merge 222/111 0.49| -0.006 proplist 252/126 0.51| -0.007 propset 12/6 0.50| -0.007 ps 12/6 0.79| -0.003 resolve 12/6 0.58| -0.010 resolved 84/42 0.50| -0.011 status 6/3 0.47| -0.054 switch 84/42 0.55| -0.028 update (legend: "1.23|+0.45" means: slower by factor 1.23 and by 0.45 seconds; factor &lt; 1 and seconds &lt; 0 means 'trunk&lt; at &gt;1481633,1x100' is faster. "2/3" means: '1.7.0,1x100' has 2 timings on record, the other has 3.) More detail: ------------ Timings for 1.7.0,5x5 N min max avg operation (unit is seconds) 6 192.14 212.78 201.26 TOTAL RUN 942 0.01 1.60 0.02 add 12 0.02 4.88 2.27 checkout 48 1.07 12.86 5.11 commit 6 0.01 0.02 0.02 copy 6 0.61 0.70 0.68 delete 30 6.32 31.50 12.60 info 12 5.33 8.49 5.76 merge 852 0.01 0.04 0.01 mkdir 48 0.01 0.02 0.01 propdel 12K 0.01 0.03 0.01 proplist 12K 0.01 0.06 0.01 propset 1K 0.01 0.03 0.01 ps 12 0.01 0.01 0.01 resolve 12 0.40 0.59 0.49 resolved 84 0.17 0.83 0.43 status 6 2.71 2.88 2.78 switch 84 0.21 4.85 1.65 update -- Timings for trunk&lt; at &gt;1481633,5x5 N min max avg operation (unit is seconds) 3 108.62 117.45 112.21 TOTAL RUN 456 0.01 1.46 0.02 add 6 0.02 3.69 1.83 checkout 24 0.57 10.48 3.60 commit 3 0.02 0.03 0.02 copy 3 0.59 0.61 0.60 delete 15 0.60 3.41 1.72 info 6 2.38 4.09 3.11 merge 470 0.01 0.02 0.01 mkdir 20 0.01 0.01 0.01 propdel 6K 0.01 0.02 0.01 proplist 6K 0.01 0.06 0.01 propset 552 0.01 0.02 0.01 ps 6 0.03 0.04 0.03 resolve 6 0.29 0.55 0.41 resolved 42 0.12 0.55 0.32 status 3 2.03 2.15 2.07 switch 42 0.20 4.11 1.39 update -- Compare trunk&lt; at &gt;1481633,5x5 to 1.7.0,5x5 N min max avg operation 6/3 0.57|-83.515 0.55|-95.335 0.56|-89.059 TOTAL RUN 942/456 0.93| -0.001 0.91| -0.137 0.88| -0.002 add 12/6 0.90| -0.002 0.76| -1.194 0.81| -0.436 checkout 48/24 0.53| -0.502 0.82| -2.376 0.71| -1.505 commit 6/3 1.18| +0.003 1.91| +0.015 1.47| +0.007 copy 6/3 0.96| -0.022 0.87| -0.090 0.88| -0.079 delete 30/15 0.10| -5.723 0.11|-28.091 0.14|-10.888 info 12/6 0.45| -2.955 0.48| -4.402 0.54| -2.651 merge 852/470 0.94| -0.001 0.42| -0.024 0.81| -0.002 mkdir 48/20 0.89| -0.001 0.71| -0.006 0.82| -0.002 propdel 12K/6K 0.85| -0.001 0.47| -0.018 0.74| -0.002 proplist 12K/6K 0.87| -0.001 0.85| -0.010 0.80| -0.002 propset 1K/552 0.88| -0.001 0.58| -0.013 0.76| -0.002 ps 12/6 3.10| +0.017 4.45| +0.031 3.85| +0.024 resolve 12/6 0.74| -0.106 0.93| -0.040 0.84| -0.081 resolved 84/42 0.69| -0.053 0.67| -0.277 0.73| -0.115 status 6/3 0.75| -0.678 0.75| -0.727 0.75| -0.709 switch 84/42 0.93| -0.014 0.85| -0.744 0.85| -0.253 update (legend: "1.23|+0.45" means: slower by factor 1.23 and by 0.45 seconds; factor &lt; 1 and seconds &lt; 0 means 'trunk&lt; at &gt;1481633,5x5' is faster. "2/3" means: '1.7.0,5x5' has 2 timings on record, the other has 3.) Timings for 1.7.0,100x1 N min max avg operation (unit is seconds) 6 13.34 22.49 18.31 TOTAL RUN 168 0.01 0.13 0.02 add 12 0.02 0.74 0.24 checkout 48 0.12 1.41 0.59 commit 6 0.02 0.03 0.02 copy 6 0.09 0.22 0.13 delete 30 0.15 0.80 0.32 info 12 0.46 0.91 0.62 merge 84 0.01 0.03 0.02 mkdir 654 0.01 0.03 0.01 proplist 528 0.01 0.03 0.01 propset 42 0.01 0.03 0.01 ps 12 0.01 0.02 0.01 resolve 12 0.05 0.11 0.07 resolved 84 0.04 0.20 0.08 status 6 0.33 0.65 0.41 switch 84 0.04 0.86 0.34 update -- Timings for trunk&lt; at &gt;1481633,100x1 N min max avg operation (unit is seconds) 3 8.67 9.04 8.84 TOTAL RUN 71 0.01 0.10 0.01 add 6 0.02 0.21 0.11 checkout 24 0.09 0.48 0.25 commit 3 0.01 0.02 0.02 copy 3 0.08 0.08 0.08 delete 15 0.10 0.19 0.13 info 6 0.19 0.23 0.21 merge 46 0.01 0.02 0.01 mkdir 1 0.01 0.01 0.01 propdel 337 0.01 0.01 0.01 proplist 273 0.01 0.01 0.01 propset 33 0.01 0.01 0.01 ps 6 0.01 0.01 0.01 resolve 6 0.06 0.07 0.06 resolved 42 0.03 0.07 0.05 status 3 0.18 0.18 0.18 switch 42 0.04 0.52 0.15 update -- Compare trunk&lt; at &gt;1481633,100x1 to 1.7.0,100x1 N min max avg operation 6/3 0.65| -4.673 0.40|-13.457 0.48| -9.469 TOTAL RUN 168/71 0.92| -0.001 0.78| -0.028 0.77| -0.004 add 12/6 0.93| -0.001 0.29| -0.525 0.46| -0.130 checkout 48/24 0.76| -0.028 0.34| -0.936 0.42| -0.339 commit 6/3 0.90| -0.002 0.50| -0.017 0.66| -0.008 copy 6/3 0.81| -0.017 0.36| -0.139 0.59| -0.055 delete 30/15 0.64| -0.053 0.23| -0.617 0.40| -0.189 info 12/6 0.41| -0.274 0.26| -0.676 0.33| -0.411 merge 84/46 0.95| -0.000 0.56| -0.015 0.79| -0.003 mkdir 654/337 0.82| -0.001 0.35| -0.017 0.58| -0.005 proplist 528/273 0.85| -0.001 0.35| -0.020 0.59| -0.006 propset 42/33 0.81| -0.002 0.34| -0.020 0.61| -0.006 ps 12/6 1.44| +0.003 0.72| -0.005 1.34| +0.003 resolve 12/6 1.28| +0.014 0.59| -0.045 0.88| -0.009 resolved 84/42 0.71| -0.012 0.35| -0.131 0.59| -0.034 status 6/3 0.54| -0.153 0.28| -0.467 0.44| -0.231 switch 84/42 0.89| -0.005 0.61| -0.336 0.45| -0.189 update (legend: "1.23|+0.45" means: slower by factor 1.23 and by 0.45 seconds; factor &lt; 1 and seconds &lt; 0 means 'trunk&lt; at &gt;1481633,100x1' is faster. "2/3" means: '1.7.0,100x1' has 2 timings on record, the other has 3.) Timings for 1.7.0,1x100 N min max avg operation (unit is seconds) 6 3.09 7.20 5.16 TOTAL RUN 6 0.04 0.09 0.06 add 12 0.02 0.27 0.11 checkout 48 0.03 0.56 0.19 commit 6 0.01 0.03 0.02 copy 6 0.02 0.05 0.03 delete 30 0.04 0.18 0.08 info 12 0.12 0.30 0.21 merge 222 0.01 0.02 0.01 proplist 252 0.01 0.02 0.01 propset 12 0.01 0.02 0.01 ps 12 0.01 0.02 0.01 resolve 12 0.01 0.04 0.02 resolved 84 0.01 0.05 0.02 status 6 0.06 0.15 0.10 switch 84 0.02 0.25 0.06 update -- Timings for trunk&lt; at &gt;1481633,1x100 N min max avg operation (unit is seconds) 3 2.54 2.57 2.55 TOTAL RUN 3 0.03 0.03 0.03 add 6 0.02 0.09 0.05 checkout 24 0.02 0.19 0.08 commit 3 0.01 0.02 0.02 copy 3 0.02 0.02 0.02 delete 15 0.04 0.08 0.05 info 6 0.06 0.09 0.07 merge 111 0.01 0.01 0.01 proplist 126 0.01 0.01 0.01 propset 6 0.01 0.01 0.01 ps 6 0.01 0.01 0.01 resolve 6 0.01 0.02 0.01 resolved 42 0.01 0.02 0.01 status 3 0.05 0.05 0.05 switch 42 0.02 0.09 0.03 update -- Compare trunk&lt; at &gt;1481633,1x100 to 1.7.0,1x100 N min max avg operation 6/3 0.82| -0.549 0.36| -4.636 0.50| -2.604 TOTAL RUN 6/3 0.92| -0.003 0.38| -0.056 0.54| -0.029 add 12/6 0.92| -0.001 0.35| -0.176 0.50| -0.054 checkout 48/24 0.64| -0.012 0.35| -0.361 0.45| -0.103 commit 6/3 1.02| +0.000 0.56| -0.013 0.69| -0.007 copy 6/3 0.91| -0.002 0.42| -0.027 0.56| -0.015 delete 30/15 1.00| -0.000 0.42| -0.105 0.59| -0.033 info 12/6 0.49| -0.060 0.29| -0.210 0.35| -0.133 merge 222/111 0.83| -0.001 0.39| -0.011 0.49| -0.006 proplist 252/126 0.85| -0.001 0.38| -0.013 0.51| -0.007 propset 12/6 0.85| -0.001 0.35| -0.013 0.50| -0.007 ps 12/6 1.30| +0.002 0.61| -0.007 0.79| -0.003 resolve 12/6 0.82| -0.002 0.47| -0.019 0.58| -0.010 resolved 84/42 0.83| -0.002 0.36| -0.029 0.50| -0.011 status 6/3 0.79| -0.013 0.33| -0.099 0.47| -0.054 switch 84/42 0.99| -0.000 0.36| -0.162 0.55| -0.028 update (legend: "1.23|+0.45" means: slower by factor 1.23 and by 0.45 seconds; factor &lt; 1 and seconds &lt; 0 means 'trunk&lt; at &gt;1481633,1x100' is faster. "2/3" means: '1.7.0,1x100' has 2 timings on record, the other has 3.) Timings for 1.7.0 N min max avg operation (unit is seconds) 18 3.09 212.78 74.91 TOTAL RUN 1K 0.01 1.60 0.02 add 36 0.02 4.88 0.87 checkout 144 0.03 12.86 1.96 commit 18 0.01 0.03 0.02 copy 18 0.02 0.70 0.28 delete 90 0.04 31.50 4.33 info 36 0.12 8.49 2.20 merge 936 0.01 0.04 0.01 mkdir 48 0.01 0.02 0.01 propdel 13K 0.01 0.03 0.01 proplist 13K 0.01 0.06 0.01 propset 1K 0.01 0.03 0.01 ps 36 0.01 0.02 0.01 resolve 36 0.01 0.59 0.20 resolved 252 0.01 0.83 0.18 status 18 0.06 2.88 1.10 switch 252 0.02 4.85 0.68 update -- Timings for trunk&lt; at &gt;1481633 N min max avg operation (unit is seconds) 9 2.54 117.45 41.20 TOTAL RUN 530 0.01 1.46 0.02 add 18 0.02 3.69 0.67 checkout 72 0.02 10.48 1.31 commit 9 0.01 0.03 0.02 copy 9 0.02 0.61 0.23 delete 45 0.04 3.41 0.63 info 18 0.06 4.09 1.13 merge 516 0.01 0.02 0.01 mkdir 21 0.01 0.01 0.01 propdel 6K 0.01 0.02 0.01 proplist 6K 0.01 0.06 0.01 propset 591 0.01 0.02 0.01 ps 18 0.01 0.04 0.02 resolve 18 0.01 0.55 0.16 resolved 126 0.01 0.55 0.13 status 9 0.05 2.15 0.77 switch 126 0.02 4.11 0.53 update -- Compare trunk&lt; at &gt;1481633 to 1.7.0 N min max avg operation 18/9 0.82| -0.549 0.55|-95.335 0.55|-33.711 TOTAL RUN 1K/530 0.93| -0.001 0.91| -0.137 0.86| -0.003 add 36/18 0.91| -0.002 0.76| -1.194 0.76| -0.207 checkout 144/72 0.64| -0.012 0.82| -2.376 0.67| -0.649 commit 18/9 1.02| +0.000 0.93| -0.002 0.87| -0.003 copy 18/9 0.91| -0.002 0.87| -0.090 0.82| -0.049 delete 90/45 1.00| -0.000 0.11|-28.091 0.15| -3.703 info 36/18 0.49| -0.060 0.48| -4.402 0.51| -1.065 merge 936/516 0.94| -0.001 0.45| -0.022 0.80| -0.002 mkdir 48/21 0.89| -0.001 0.71| -0.006 0.83| -0.002 propdel 13K/6K 0.84| -0.001 0.47| -0.018 0.73| -0.002 proplist 13K/6K 0.86| -0.001 0.85| -0.010 0.78| -0.002 propset 1K/591 0.86| -0.001 0.58| -0.013 0.75| -0.003 ps 36/18 1.30| +0.002 2.23| +0.022 1.82| +0.008 resolve 36/18 0.82| -0.002 0.93| -0.040 0.83| -0.033 resolved 252/126 0.83| -0.002 0.67| -0.277 0.70| -0.053 status 18/9 0.79| -0.013 0.75| -0.727 0.70| -0.331 switch 252/126 0.99| -0.000 0.85| -0.744 0.77| -0.157 update (legend: "1.23|+0.45" means: slower by factor 1.23 and by 0.45 seconds; factor &lt; 1 and seconds &lt; 0 means 'trunk&lt; at &gt;1481633' is faster. "2/3" means: '1.7.0' has 2 timings on record, the other has 3.) Had started at Mon May 13 00:26:11 UTC 2013, done at Mon May 13 00:36:13 UTC 2013 wrote chart file: /home/neels/svnbench/charts/.busy/compare_1.7.0_trunk&lt; at &gt;last12.svg wrote chart file: /home/neels/svnbench/charts/.busy/compare_1.7.0_trunk&lt; at &gt;last12,5x5.svg wrote chart file: /home/neels/svnbench/charts/.busy/compare_1.7.0_trunk&lt; at &gt;last12,1x100.svg wrote chart file: /home/neels/svnbench/charts/.busy/compare_1.7.0_trunk&lt; at &gt;last12,100x1.svg </pre> neels< at >apache.org 2013-05-13T00:36:39 Problems building trunk (and probably 1.8.x) on FreeBSD 9.0 http://comments.gmane.org/gmane.comp.version-control.subversion.devel/142210 <pre>I was trying to build trunk on some ASF Infrastructure machines that have FreeBSD 9.0 on them. I ran into the following error: [[[ /bin/sh /home/breser/trunk/libtool --tag=CC --silent --mode=compile gcc -std=c89 -g -O2 -g -O2 -I./subversion/include -I./subversion -I/usr/local/include/apr-1 -I/usr/local/include/apr-1 -I/usr/local/include -I/usr/local/include/db48 -I/home/breser/trunk/sqlite-amalgamation -o subversion/libsvn_subr/sqlite.lo -c subversion/libsvn_subr/sqlite.c subversion/libsvn_subr/sqlite.c:65:2: error: #error SQLite is too old </pre> Ben Reser 2013-05-10T20:55:16 Search Engine Search the mailing list at Gmane query http://search.gmane.org/?group=$group=gmane.comp.version-control.subversion.devel