On Sun, Feb 11, 2018 at 4:33 PM Daniel Shahaf
wrote:
> Troy Curtis Jr wrote on Sun, 11 Feb 2018 16:10 +:
> > I'm asking for comments because the actual diff output given when
> > 'diffoptions' is None, is changed.
> ...
> > Thoughts?
>
> 1. We have two consumers of FileDiff in tools/. Check
On 12.02.2018 16:00, Daniel Shahaf wrote:
> The shelving functionality is experimental:
>
> /** Shelve a change.
> ⋮
> * @since New in 1.10.
> * @warning EXPERIMENTAL.
> */
> SVN_EXPERIMENTAL
> svn_error_t *
> svn_client_shelve(const char *name,
>
On 2/12/2018 1:04 PM, Karl Fogel wrote:
* tools/hook-scripts/mailer/mailer.py: If the Subversion Python bindings
could not be found for import at all, have the error message state
that clearly, instead of duplicating the error message used for
importing too old a version of the bindings.
On 02/12/2018 04:04 PM, Karl Fogel wrote:
> Hi all. I think one of the import error messages in
> tools/hook-scripts/mailer/mailer.py is misleading (I discovered this while
> debugging the problem that Troy Curtis Jr. solved in r1823802). Any
> objections if I install this patch?
>
> [[[
> Giv
Hi all. I think one of the import error messages in
tools/hook-scripts/mailer/mailer.py is misleading (I discovered this while
debugging the problem that Troy Curtis Jr. solved in r1823802). Any objections
if I install this patch?
[[[
Give correct error message on failure to import Subversion
Daniel Shahaf wrote:
2. We might include the change in an alpha/beta release to give it more testing.
(#2 applies to FSFS's new verify-before-commit knob, too.)
I noticed a 'verify-before-commit' option is included in fsfs.conf
nowadays, "enabled by default in debug builds". (On unix that's
Daniel Shahaf wrote:
julianf...@apache.org wrote on Mon, 12 Feb 2018 13:17 +:
+++ subversion/trunk/subversion/svn/shelf-cmd.c Mon Feb 12 13:17:16 2018
@@ -71,6 +71,36 @@ friendly_duration_str(apr_time_t duratio
+#ifndef WIN32
+/* Run CMD with ARGS.
+ * Send its stdout to the parent's stdou
The shelving functionality is experimental:
/** Shelve a change.
⋮
* @since New in 1.10.
* @warning EXPERIMENTAL.
*/
SVN_EXPERIMENTAL
svn_error_t *
svn_client_shelve(const char *name,
const apr_array_header_t *paths,
s
julianf...@apache.org wrote on Mon, 12 Feb 2018 13:17 +:
> +++ subversion/trunk/subversion/svn/shelf-cmd.c Mon Feb 12 13:17:16 2018
> @@ -71,6 +71,36 @@ friendly_duration_str(apr_time_t duratio
> +#ifndef WIN32
> +/* Run CMD with ARGS.
> + * Send its stdout to the parent's stdout. Disconnect it
On 12.02.2018 14:27, Stefan Sperling wrote:
> On Mon, Feb 12, 2018 at 02:24:56PM +0100, Branko Čibej wrote:
>> Out of interest: why are you using APR without threads?
> To ensure that this configuration is still being tested.
That's good.
https://ci.apache.org/builders/svn-x64-macosx-apr1.3-nothr
On Mon, Feb 12, 2018 at 02:24:56PM +0100, Branko Čibej wrote:
> Out of interest: why are you using APR without threads?
To ensure that this configuration is still being tested.
On 12.02.2018 13:07, Stefan Sperling wrote:
> On Mon, Feb 12, 2018 at 12:50:12PM +0100, Stefan Sperling wrote:
>> The svn-bb-openbsd bot has found a crash in the object pool code.
>>
>> It is triggered when running RA serf tests on OpenBSD. The httpd
>> work processes crash due to what looks like a
Philip Martin wrote:
Daniel Shahaf writes:
Neither of these calls correctly quotes the path that's interpolated
into the command line. Consequently, the wcroot_abspath would be
executed as shell commands. Could that be fixed please?
Should probably do utf8 to local conversion as well.
Cha
On Mon, Feb 12, 2018 at 12:50:12PM +0100, Stefan Sperling wrote:
> The svn-bb-openbsd bot has found a crash in the object pool code.
>
> It is triggered when running RA serf tests on OpenBSD. The httpd
> work processes crash due to what looks like a use-after-free (0xdf
> means this byte of memory
The svn-bb-openbsd bot has found a crash in the object pool code.
It is triggered when running RA serf tests on OpenBSD. The httpd
work processes crash due to what looks like a use-after-free (0xdf
means this byte of memory has been freed).
I can reproduce this on sparc64 and amd64 machines.
The
15 matches
Mail list logo