this case Subversion has been configured with the switch "--disable-
> keychain", I assume
> this is triggering the problem.
After some discussion on irc with Johannes and input from a few others I
applied a patch that resolved the problems in r1761755.
The patch is nominated for backp
d I would be +1 on backporting such a fix to
1.9 (and if possible 1.8).
To keep translations easy I would recommend trying if we can just copy
some message support from 'svn'.
Bert
a
warning’. There are better ways to fix that problem… and not just for your
favorite RA layer, ignoring others.
These warnings on the fs layer were designed to be logged server side (e.g. in
the apache error log); not to be transferred to the client.
Bert
Sent from Mail for Windows 10
From
igurable... or depending on filesize.
We certainly want the new behavior for non-pristine working copies (on the IDEA
list for years), but I'm not sure if we always want this behavior as only
option.
This mail is partially, to just discuss this topic on the list, to make sure
everybody k
hecksum of some very
unlikely data, so users might be able to trigger this under some very unusual
circumstances.
Bert
ure reference, you might have tried building trunk@HEAD after
> locally reverting r1758069; i.e.:
>
> svn up
> svn merge -c -r1758069
>
> make check
>
> Stefan wrote on Sun, Aug 28, 2016 at 18:33:55 +0200:
> > Got approved by Bert.
> >
>
> -Original Message-
> From: Ivan Zhakov [mailto:i...@visualsvn.com]
> Sent: vrijdag 26 augustus 2016 11:58
> To: Bert Huijben <b...@qqmail.nl>
> Cc: dev@subversion.apache.org
> Subject: Re: svn commit: r1757761 -
> /subversion/trunk/subversion/libsvn_client/co
ll not always work if the file is not
explicitly closed before. On Windows it depends on apr opening the file with
delete share flags (10% perf hit in some cases) and nobody else opening the
file.
In general strictly closing before deleting is a safer procedure.
Bert
n.
I would imagine that this could really help on updates with a lot of files,
especially on threaded apache.
(The other patch handles the checkout case)
Bert
> -Original Message-
> From: Evgeny Kotkov [mailto:evgeny.kot...@visualsvn.com]
> Sent: donderdag 7 juli 2016 19:02
> To: Bert Huijben <b...@qqmail.nl>
> Cc: Subversion Development <dev@subversion.apache.org>
> Subject: Re: svn commit: r1751207 - in /subver
e the patch
I would say this introduces *more* unspecified behavior.
In practice 64 bit is reserved for each integer type provided in varargs in the
Windows AMD64 calling convention, so it doesn't change that much and the
runtime result will be +- identical
But when we use the system sprintf(), fprintf(), anyway... why not just use
"0x%p" for all calling conventions?
(Rest of the patch looks ok)
Bert
es around the world at least every year -commonly more
often- and I don't think anybody would really check for the US 2007 changes
specifically any more.
I just checked the archive for the iana timezone database
(https://www.iana.org/time-zones) and we are now on the fifth set of rules
released in 2016. There were 6 in 2015, 10 in 2014...
I don't think the 2007 rule change warrants its own article when you don't look
at it with US-only-glasses.
Bert
If we can find a browser that breaks, sure..
(Given the transfer from , which did support numbers if I remember
correctly, I'm guessing that this is all just theory)
Bert
thing that makes all these related is that we never designed for a
‘lock a complete tree’ scenario in Subversion. And I still don’t know why users
really want this, as with Subversions implementation it doesn’t really help you
in any scenario.
Bert
From: Stefan Hett
r the ra layer) reports its
changes, which is mostly based on how things are stored in an apr hash in
libsvn_repos which performs the tree diff server side.
Bert
>
> --
> Ivan Zhakov
sys' is not defined
FAIL: lock_tests.py 57: refresh timeout of DAV lock
]]]
Most (if not all) can't find sys.
After this patch (combined with the previous +- 3 patches).
It appears to be specific for the ra-dav test runs...
I'm not sure which exact patch introduced this error though.
Bert
> -Original Message-
> From: Daniel Shahaf [mailto:d...@daniel.shahaf.name]
> Sent: maandag 2 mei 2016 14:05
> To: Bert Huijben <b...@qqmail.nl>
> Cc: Stefan Fuhrmann <stef...@apache.org>; dev@subversion.apache.org
> Subject: Re: Requiring python 2.7
platforms if we change things for swig,
and some of the installed python scripts.
Bert
Sent from Mail for Windows 10
From: Stefan Fuhrmann
Sent: maandag 2 mei 2016 07:02
To: dev@subversion.apache.org
Subject: Re: svn commit: r1741742 - in
/subversion/trunk:contrib/client-side/svn-merge-vendor.py
Isn’t this a documented default value?
If it is, changing the behavior will likely break many users that don’t use
sasl.
Bert
Sent from my Windows 10 phone
From: Daniel Shahaf
Sent: zaterdag 30 april 2016 02:53
To: dev@subversion.apache.org
Subject: [PATCH/RFC] use-sasl=true in --without-sasl
Not entirely sure, but I think you should still publish your pgp key to the
major key stores. Once you put your fingerprint on id.apache.org, it knows how
to fetch your key from there.
Bert
Sent from Mail for Windows 10
From: Stefan
Sent: donderdag 28 april 2016 01:15
To: dev
issues)
Bert
Sent from Mail for Windows 10
From: Ivan Zhakov
Sent: zaterdag 23 april 2016 20:08
To: Branko Čibej
Cc: Subversion Development
Subject: Re: Really funny test failure on Windows
On 23 April 2016 at 18:43, Branko Čibej <br...@apache.org> wrote:
> This just happened when I wa
+1 on backport. Feel free to call it a documentation/trivial change that
doesn't need full votes.
Bert
> -Original Message-
> From: kot...@apache.org [mailto:kot...@apache.org]
> Sent: donderdag 21 april 2016 15:51
> To: comm...@subversion.apache.org
> Subj
ten it.
I think I missed that the context isn't complete and the format string
specifies a local path first, and a repos_relpath later.
Bert
ore information about conventions we use, such as our preferred
> log message format, see our community guide:
> http://subversion.apache.org/docs/community-guide/
In this specific case I like the heads-up on the mailinglist though.
But +1 on removing it.
Bert
his change. Good fix.
Perhaps we can extend it later on to verify in which mode we currently run the
test suite (assuming the suite sets up the apache server)
Bert
> -Original Message-
> From: Johan Corveleyn [mailto:jcor...@gmail.com]
> Sent: donderdag 3 maart 2016 09:19
> To: Bert Huijben <b...@qqmail.nl>
> Cc: Branko Čibej <br...@apache.org>; Subversion Development
> <dev@subversion.apache.org>
> Subjec
answered now... This is not really fixable in a backwards compatible
way as there is no way to express errors in the schema. (How do we express an
error to access a file somewhere deep inside a tree? That would make the entire
tree invalid as expressed in todays schema)
Bert
o care for handling error conditions.
(Sometimes it is easier to work around things by just performing things
recursively on an ancestor directory)
Bert
t; working copy would break in a different way.
>
> I guess we would need some "change locale" operation, which would at least
> update the saved locale in the .svn directory.
There is no saved locale in the .svn directory...
Bert
ound out cleanup is not going to help here... we just can't access
this file (or directory, or symlink), so we can't delete it or anything to
help you.
Bert
were already moved to the standard
locations.
I just removed the file in r1732848.
(Copied last version to the bottom of this mail for easy reference)
Bert
[[
ra_serf status
==
This library is an RA-layer implementation of a WebDAV client that uses
Serf.
Serf's homepage is
[Moving thread to dev@s.a.o from users@]
> -Original Message-
> From: Bert Huijben [mailto:b...@qqmail.nl]
> Sent: maandag 22 februari 2016 13:21
> To: 'Daniel Shahaf' <d...@daniel.shahaf.name>; 'Michal Matyl'
> <michal.ma...@zf.com>
> Cc: us...@subve
t_conflict_option_merged_text ||
> + (option_id ==
> svn_client_conflict_option_update_any_moved_away_children
> +&& incoming_change == svn_wc_conflict_action_edit)) &&
The new variable incoming_change is used uninitialized here, breaking the build
Are you sure the pool arguments are in the right order here?
The usual order is result_pool, scratch_pool while most of the calls here
appear to use the opposite order.
Bert
Sent from Mail for Windows 10
From: stef...@apache.org
Sent: maandag 15 februari 2016 22:47
To: comm
on recursive operations.
Bert
Sent from Mail for Windows 10
From: Stefan Sperling
Sent: maandag 15 februari 2016 18:17
To: Bert Huijben
Cc: dev@subversion.apache.org
Subject: Re: svn commit: r1730546 - in
/subversion/trunk/subversion:include/private/svn_wc_private.h
libsvn_client/resolved.clibsvn_wc
_depth_empty,
> + FALSE, FALSE, TRUE,
> + conflict_choice,
> + NULL, NULL, /* legacy conflict_func/baton
> */
> + ctx->cancel_func,
> + ctx->cancel_baton,
> + ctx->notify_func2,
> + ctx->notify_baton2,
> + scratch_pool);
> + err = svn_error_compose_create(err, svn_wc__release_write_lock(ctx-
> >wc_ctx,
> +
> lock_abspath,
> +
> scratch_pool));
> + svn_io_sleep_for_timestamps(local_abspath, scratch_pool);
> + SVN_ERR(err);
Same thing here... I don't think this sleep operation is necessary.
Same conditions, in case it is really necessary: add output argument to note
the cases where they are really necessary, etc. (Never needed on directories,
things that are already deleted, etc. etc.)
Bert
ecific for 'targets' argument in one of our .i
files, but no matching implementation.
I'm trying to find why, but the only things I can find would indicate that this
is just based on how swig implements array mappings for perl.
Bert
are helper functions on the client to create an Ra session from that,
which automatically hook a few other features that are now ignored.
Bert
itself. And given that the diff parse code has
to read the file multiple times we will need a tempfile anyway.
Bert
Sent from Outlook Mail for Windows 10 phone
From: Daniel Shahaf
Sent: zondag 7 februari 2016 01:22
To: Andreas Scherer; dev@subversion.apache.org
Subject: Re: [RFE] Make 'svn patch
ting the stack to the old state. With an inner function this is not
guaranteed to work. (I'm guessing many compilers would just inline the single
call, but then...)
Bert
h '/foo').
In my experience using svndumpfilter without --drop-*, but without
--renumber-revs is the most likely cause of these problems. Forgetting to
add that flag just creates invalid dumpfiles, so I don't think we should
have allowed that scenario. (Where is my time machine?)
BTW: svndumpfilter existed since Subversion 1.0, so it certainly predates
2009. Not sure about when its specific features were introduced though.
Bert
> -Original Message-
> From: Stefan Fuhrmann [mailto:stef...@apache.org]
> Sent: maandag 1 februari 2016 07:29
> To: dev@subversion.apache.org
> Subject: Re: svn commit: r1727785 -
> /subversion/trunk/subversion/libsvn_ra_svn/marshal.c
>
>
>
> On 31.01.
cleanup)?
I think this commit has the wrong log message.
Bert
>
> Modified:
> subversion/trunk/subversion/libsvn_ra_svn/marshal.c
>
> Modified: subversion/trunk/subversion/libsvn_ra_svn/marshal.c
> URL:
> http://svn.apache.org/viewvc/subversio
ow use _() to localize the option here, instead of
in the callee.
I was assuming the N_() in the array was that it wasn't evaluated there, but
added to the localization strings anyway... to allow later code to use the
localized variants via the api.
We currently don't provide access to our gettext instance via our api, so I
have no idea how localization should work when we are used as a library and
don't provide the localized values.
Bert
ut the member variable... I don't see a problem with the
accessor function. It might not be needed later on after more cleanup as
callers most likely already have all/most these pointers itself)
Bert
Do not check DIRENT->SIZE due known problem in ra_serf.
> This
>test for different bug.
Wasn't it easier to 'just' fix ra_serf here?
If I remember correctly serf always retrieves the node kind on retrieved nodes,
so this shouldn't be hard to fix.
Bert
use, but it
takes someone to maintain such a thing.
(I did exactly this for C#/.Net in the SharpSvn project)
Bert
Hi Greg,
Any news on this?
I’m trying to find the list this discussion moved to.
Bert
From: Greg Stein [mailto:gst...@gmail.com]
Sent: dinsdag 1 december 2015 22:20
To: Tony Stevenson <pct...@apache.org>
Cc: David Nalley <ke4...@apache.org
+1: stefan2
When I look at the patch (and the log message), it looks at a very simple
change... but I have absolutely no idea what the side effects are.
Given your descriptions, I would like to see it backported, but can you help us
review this change?
Thanks,
Bert
le_checksum2(), but reference node by @a root and
> + * @a path.
> + */
> +svn_error_t *
> svn_fs_file_checksum(svn_checksum_t **checksum,
> svn_checksum_kind_t kind,
> svn_fs_root_t *root,
> @@ -2422,21 +2456,29 @@ svn_fs_file_md5_checksum(unsigned char d
Bert
while (index_from && index_to
> + && sig_from_prefix[index_from] == sig_to_prefix[index_to])
> +{
> + sig_from_prefix[index_from] = sig_to_prefix[index_to] = '\0';
> + --index_from;
> + --index_to;
> +}
> +
I'm wondering, is it safe to do this this way... or should we use the safer
svn_uri_dirname() to split of whole components from a url at a time.
Are there cases where this might match half a directory (or host) name?
Bert
mergeinfo for merge of r1721488 into '.':
U .
--- Merging r1721648 into '.':
Gsubversion\bindings\swig\INSTALL
Asubversion\bindings\swig\include\proxy.py
Gsubversion\bindings\swig\include\proxy.swg
Cbuild\ac-macros\swig.m4
--- Recording mergeinfo for merge of r1721648 into '.':
G .
Summary of conflicts:
Text conflicts: 1
]]
Bert
that aren’t completely
updated.
Perhaps something around the Subversion Corporation if ever? (Don’t think so…)
Bert
From: Greg Stein
Sent: vrijdag 1 januari 2016 14:43
To: dev@subversion.apache.org
Subject: votes? [HISTORY HELP?]
Hey all,
There have been a few occasions when I've noted
ust exist'?
When I accidentally tried to use a wrong directory some weeks ago (while
working on libsvn_fs_git) I found that when that file is missing we assume that
the filesystem is BDB.
Bert
ual Studio 2015 Update 1 on Windows 8.1 Enterprise x64.
(32 bit binaries)
apr 1.5.1
apr-util 1.5.4
apr_memcache 1.5.4
db 4.4.20
expat 2.1.0
httpd 2.4.17
mod_dav 2.4.17
openssl 1.0.2e
sasl 2.1.26
serf 1.3.8
sqlite 3.9.2
zlib 1.2.8
[fsfs | bdb] x [local | svn | serf]
Signatures committed.
Bert
ual Studio 2015 Update 1 on Windows 8.1 Enterprise x64.
(32 bit binaries)
apr 1.5.1
apr-util 1.5.4
apr_memcache 1.5.4
db 4.4.20
expat 2.1.0
httpd 2.4.17
mod_dav 2.4.17
openssl 1.0.2e
sasl 2.1.26
serf 1.3.8
sqlite 3.9.2
zlib 1.2.8
[fsfs | bdb] x [local | svn | serf]
Bert
> -Original Message-
> From: Daniel Shahaf [mailto:d...@daniel.shahaf.name]
> Sent: zaterdag 12 december 2015 09:20
> To: Julian Foad <julianf...@gmail.com>
> Cc: Philip Martin <philip.mar...@wandisco.com>; Bert Huijben
> <b...@qqmail.nl>; dev@subv
Are both hash tables created with the same hash function? (I think stefan2
introduced some variants).
Otherwise I would expect some key values to be changed somewhere after adding
to the first hashtable… But I don’t think this is really a likely scenario.
Bert
Sent from Outlook Mail
> -Original Message-
> From: Evgeny Kotkov [mailto:evgeny.kot...@visualsvn.com]
> Sent: maandag 7 december 2015 12:11
> To: Bert Huijben <b...@qqmail.nl>
> Cc: Subversion Development <dev@subversion.apache.org>
> Subject: Re: svn commit: r1718167 -
t dav_svn__status(request_rec *r)
> the request. Ideally we would iterate over all processes but that
> would need some MPM support, so we settle for simply showing the
> process ID. */
> -#include
>ap_rprintf(r, "Server process id: %d\n", (int)getpid());
> #endif
>
> ]]]
I would be +1 on applying this patch under the obvious fix rule.
Bert
> -Original Message-
> From: Daniel Shahaf [mailto:danie...@apache.org]
> Sent: zaterdag 28 november 2015 18:32
> To: Philip Martin <philip.mar...@wandisco.com>
> Cc: Bert Huijben <b...@qqmail.nl>; 'Branko Čibej' <br...@apache.org>;
> dev@subversion.
les are kept alive by apr pools, in which the original file
handle was allocated.
And in case of JavaHL these would be kept alive by java objects... So I guess
some explicit dispose/release/close call is missing somewhere.
Once the pool is properly cleaned out/destroyed most likely the files will also
be deleted, as that is an operation we commonly delay to pool cleanup.
Bert
> -Original Message-
> From: Philip Martin [mailto:philip.mar...@wandisco.com]
> Sent: vrijdag 27 november 2015 13:46
> To: Bert Huijben <b...@qqmail.nl>
> Cc: 'Daniel Shahaf' <d...@daniel.shahaf.name>; 'Branko Čibej'
> <br...@apache.org>; dev@subv
> -Original Message-
> From: Bert Huijben [mailto:b...@qqmail.nl]
> Sent: vrijdag 27 november 2015 10:28
> To: 'Daniel Shahaf' <d...@daniel.shahaf.name>; 'Philip Martin'
> <philip.mar...@wandisco.com>
> Cc: 'Branko Čibej' <br...@apache.org>; dev@subv
> -Original Message-
> From: Daniel Shahaf [mailto:d...@daniel.shahaf.name]
> Sent: vrijdag 27 november 2015 08:50
> To: Philip Martin <philip.mar...@wandisco.com>
> Cc: Bert Huijben <b...@qqmail.nl>; 'Branko Čibej' <br...@apache.org>;
> dev@subv
> -Original Message-
> From: Ivan Zhakov [mailto:i...@visualsvn.com]
> Sent: donderdag 26 november 2015 09:19
> To: dev@subversion.apache.org; Bert Huijben <b...@qqmail.nl>
> Subject: Re: svn commit: r1716575 - in
> /subversion/trunk/subversion/libsvn_ra_serf: c
> -Original Message-
> From: Ivan Zhakov [mailto:i...@visualsvn.com]
> Sent: donderdag 26 november 2015 09:54
> To: Bert Huijben <b...@qqmail.nl>
> Cc: dev@subversion.apache.org
> Subject: Re: svn commit: r1716575 - in
> /subversion/trunk/subversion/libsvn_
> -Original Message-
> From: Ivan Zhakov [mailto:i...@visualsvn.com]
> Sent: donderdag 26 november 2015 09:54
> To: Bert Huijben <b...@qqmail.nl>
> Cc: dev@subversion.apache.org
> Subject: Re: svn commit: r1716575 - in
> /subversion/trunk/subversion/libsvn_
ing for
> requests that specify ?r=WORKINGREV in the url, e.g.
>
> https://svn.apache.org/repos/asf/subversion/trunk/README?r=1716593
I think you should document this as not sending ?p=<> in the url.
Once you add a ?p=, an additional = specifying the same or lower revision is
100% safe.
Bert
> -Original Message-
> From: Philip Martin [mailto:phi...@codematters.co.uk]
> Sent: donderdag 26 november 2015 19:37
> To: Bert Huijben <b...@qqmail.nl>
> Cc: 'Branko Čibej' <br...@apache.org>; dev@subversion.apache.org
> Subject: Re: svn commit:
ctly
> reproducible for him. Should I ask for an "svn status -u" output?
This should be perfectly reproducable when you call status on a directory that
contains subdirectories that you are not allowed to read (via a mod_authz_svn
config file or similar svnserve config).
svn status -u explicitly ignores absent nodes.
Bert
> -Original Message-
> From: Branko Čibej [mailto:br...@apache.org]
> Sent: woensdag 25 november 2015 23:43
> To: dev@subversion.apache.org
> Subject: Re: svn commit: r1716548 -
> /subversion/trunk/subversion/tests/libsvn_fs/fs-test.c
>
> On 25.11.2015 2
The question is: Why isn't the base revision properly saved in the txn when the
HEAD revision is a no-op commit?
I'm guessing that we safe something else and then accidentally fetch the last
modified revision of that and use it as the BASE revision of the transaction.
Ber
> -Original Message-
> From: stef...@apache.org [mailto:stef...@apache.org]
> Sent: woensdag 25 november 2015 14:32
> To: comm...@subversion.apache.org
> Subject: svn commit: r1716394 -
> /subversion/trunk/subversion/tests/cmdline/svnrdump_tests.py
>
> Author: stefan2
> Date: Wed Nov 25
> -Original Message-
> From: Ivan Zhakov [mailto:i...@visualsvn.com]
> Sent: maandag 23 november 2015 15:31
> To: dev@subversion.apache.org; Bert Huijben <b...@qqmail.nl>
> Subject: Re: svn commit: r1715832 - /subversion/trunk/win-tests.py
>
> On 23 Nove
There is currently no way how a user with released software can encounter this
problem.
BTW I think I have a solution for the replay problem at the cost of a few lb of
ram.
Bert
Sent from Outlook Mail for Windows 10 phone
From: Julian Foad
Sent: vrijdag 20 november 2015 12:04
To: Bert
With the right tooling both operations should be equivalent. Perhaps it is
easier to spend time on that.
Bert
Sent from Outlook Mail for Windows 10 phone
From: Philip Martin
Sent: vrijdag 20 november 2015 12:21
To: Ivan Zhakov
Cc: Daniel Shahaf;dev@subversion.apache.org
Subject: Re: svn+ssh
with an update report that wants to fetch too many files.
Bert
Sent from Outlook Mail for Windows 10 phone
From: Ivan Zhakov
Sent: vrijdag 20 november 2015 10:27
To: Bert Huijben
Cc: dev@subversion.apache.org
Subject: Re: svn commit:
r1715228-/subversion/trunk/subversion/libsvn_ra_serf/util.c
On 20
authentication could be introduced directly
in the svnserve protocol.
Both these options would also fix the problems raised.
Bert
Sent from Outlook Mail for Windows 10 phone
From: Mark Phippard
Sent: vrijdag 20 november 2015 15:20
To: Bert Huijben
Cc: Philip Martin;Ivan Zhakov;Daniel Shahaf;dev
ks fine, but currently libssh2 still lacks a few of the more recently added
cypher types of ssh, with shorter handshake times.
Bert
> -Original Message-
> From: Ivan Zhakov [mailto:i...@visualsvn.com]
> Sent: vrijdag 20 november 2015 10:12
> To: Bert Huijben <b...@qqmail.nl>
> Cc: dev@subversion.apache.org; d...@serf.apache.org
> Subject: Re: svn commit: r1715228 -
> /subversion/trunk/subve
tion is
necessary even when earlier requests succeeded, depending on the auth scheme).
If you want to try this for yourself, you need http 2.4.17 with mod_http2 (or
newer) and OpenSSL 1.0.2 or newer. (If you patch Subversion a bit further you
should be able to use h2c, which avoids the openssl dependency).
Bert
> -Original Message-
> From: Bert Huijben [mailto:b...@qqmail.nl]
> Sent: donderdag 19 november 2015 23:41
> To: 'Ivan Zhakov' <i...@visualsvn.com>
> Cc: dev@subversion.apache.org
> Subject: RE: svn commit: r1715228 -
> /subversion/trunk/subve
> -Original Message-
> From: Ivan Zhakov [mailto:i...@visualsvn.com]
> Sent: donderdag 19 november 2015 23:12
> To: Bert Huijben <b...@qqmail.nl>
> Cc: dev@subversion.apache.org
> Subject: Re: svn commit: r1715228 -
> /subversion/trunk/subversion/libsvn_ra_serf
nless Content-Length or Content-Type is set.
This should not affect normal Subversion usage, but it makes our http usage a
bit cleaner.
Bert
ry, 1, True,
> + return run_command(svnmover_binary, 1, False,
> *(_with_auth(_with_config_dir(varargs
You missed changing the comment here that documents that you explicitly use
binary output.
Bert
where it should fetch a pristine from:
BASE or WORKING (or anything inbetween). We can never release the code until at
least those things are guaranteed stable.
I don't think we can just make a final decision on the final future approach
here yet...
Bert
I think we should keep the functions in the private namespace until we really
expose this in our api/tools. Otherwise: +1.
Bert
From: Julian Foad
Sent: dinsdag 10 november 2015 16:28
To: dev
Subject: Merge 'svnmover' demo tool to trunk
The work on the 'move-tracking-2' branch currently
current time in
> - * microseconds since 00:00:00 January 1, 1970 UTC. So it is
> - * extremely unlikely that a transaction name will be reused.
> + * sequence and reuse transaction names.
I would have said it 'may' reuse transaction names.
I don't think we promise that they will actually be reused or not. And perhaps
the format may change again in the future.
Bert
ed to:
>
> http://subversion.tigris.org/issues/show_bug.cgi?id=4598
Yes this looks like a typical example of that bug.
The original instance of this bug also related to a CVS import.
Bert
>
> --
> Vincent Lefèvre <vinc...@vinc17.net> - Web: <https://www.vinc17.net/>
>
> -Original Message-
> From: i...@visualsvn.com [mailto:i...@visualsvn.com] On Behalf Of Ivan
> Zhakov
> Sent: zaterdag 31 oktober 2015 18:10
> To: Bert Huijben <b...@qqmail.nl>
> Cc: dev@subversion.apache.org
> Subject: Re: svn commit: r1711582 - in /
ure that path is file. */
> + if (svn_fs_fs__dag_node_kind(node2) != svn_node_file)
> +return svn_error_createf
> + (SVN_ERR_FS_GENERAL, NULL, _("'%s' is not a file"), path2);
> +
I don't think you want to backport this, but returning SVN_ERR_FS_NOT_FILE
would make more sense here, than returning SVN_ERR_FS_GENERAL.
And I would like to see the regression test checking that for all filesystem
layers for 1.10... but that is a bigger change.
Bert
lient/merge.c tests/cmdline/merge_automatic_tests.py
> >>> +SVN_ERR(svn_stream_open_readonly(_stream,
> mine_abspath,
> >>> + scratch_pool, scratch_pool));
> >>> +
> >>> + if (!special &&
on
not-changed files)
· 1.5-1.8 behavior required for ‘svn annotate -g’ (Wants to know all
revisions. Uses that to trigger behavior. Deltas sometimes between unexpected
revisions)
Api only used for blame or otherwise determining actual changes.
Bert
From: b
cases... while
we already have the code to fix those cases.
We should revert the behavior where it makes sense. Reverting it everywhere
'just because' doesn't make sense.
Bert
Sent from Mail for Windows 10
From: Johan Corveleyn
Sent: dinsdag 27 oktober 2015 02:53
To: Bert Huijben
Cc
> -Original Message-
> From: Evgeny Kotkov [mailto:evgeny.kot...@visualsvn.com]
> Sent: dinsdag 27 oktober 2015 14:38
> To: Bert Huijben <b...@qqmail.nl>
> Cc: Johan Corveleyn <jcor...@gmail.com>; Stefan Fuhrmann
> <stefan.fuhrm...@wandisco.com>; Julia
> -Original Message-
> From: Ivan Zhakov [mailto:i...@visualsvn.com]
> Sent: maandag 26 oktober 2015 09:05
> To: dev@subversion.apache.org; Bert Huijben <b...@qqmail.nl>
> Subject: Re: svn commit: r1705843 - in /subversion/trunk/subversion:
> libsvn_client/e
> -Original Message-
> From: Ivan Zhakov [mailto:i...@visualsvn.com]
> Sent: maandag 26 oktober 2015 09:05
> To: dev@subversion.apache.org; Bert Huijben <b...@qqmail.nl>
> Subject: Re: svn commit: r1705843 - in /subversion/trunk/subversion:
> libsvn_client/e
> -Original Message-
> From: Ivan Zhakov [mailto:i...@visualsvn.com]
> Sent: maandag 26 oktober 2015 19:37
> To: dev@subversion.apache.org; Bert Huijben <b...@qqmail.nl>
> Subject: Re: svn commit: r1710631 - in /subversion/trunk/subversion:
> include/svn_fs
Accept-Encoding",
> - "svndiff1;q=0.9,svndiff;q=0.8");
> + /* Do not advertise svndiff1 support if we're not interested in
> + compression. */
> + serf_bucket_headers_setn(headers, "Accept-Encoding",
> "svndiff;q=0.9");
Same here.
Bert
> }
>
>return SVN_NO_ERROR;
>
101 - 200 of 1808 matches
Mail list logo
