This is the final test release before we go to [core] with the new
crond. Please signoff, one for each architecture.


Aside from the upstream changes, two packaging items:

* I've decided to leave the crond logrotate script here for now,
unless/until aaron (syslog-ng maintainer) thinks we should take it into
that package. I'd rather tackle that effort in a separate step, rather
than have it block launching the new 4.x dcron series.


* I've added an optdepend on smtp-server to address FS #12662.

Upstream release log below:

dcron 4.4 released. Changes:

* Finished mailjobs were being left as zombie processes. Fixed.

* When using crond with logging-to-file, user jobs could only log some
events if they had write access to the log. Fixed this by having crond
keep a file descriptor open to the log; also added a SIGHUP handler
to make crond re-open the logfile. The sample logrotate script now
sends that signal.

* More sensible command-line parsing by crontab.

* Add prune-cronstamps to extra; document extra/*; general improvement
of README and manpages.

* Portability improvements, and defs.h now has fuller comments about
requirements.

* Makefile improvements: `make` now caches variables for `make install`;
don't stomp CFLAGS environment variable, and added
BINDIR,SBINDIR,MANDIR.


* Thanks to Juergen Daubert for testing and suggestions.

- P

On Mon, Jan 18, 2010 at 9:33 PM, Paul Mattal <paul@mattal.com> wrote:
> This is the final test release before we go to [core] with the new crond.
> Please signoff, one for each architecture.
>
> Aside from the upstream changes, two packaging items:
>
> * I've decided to leave the crond logrotate script here for now,
> unless/until aaron (syslog-ng maintainer) thinks we should take it into that
> package. I'd rather tackle that effort in a separate step, rather than have
> it block launching the new 4.x dcron series.
>
> * I've added an optdepend on smtp-server to address FS #12662.

Seems OK here on x86_64. No more zombie sendmail processes either.

-Dan

On 01/19/2010 11:35 PM, Dan McGee wrote:

On Mon, Jan 18, 2010 at 9:33 PM, Paul Mattal<paul@mattal.com> wrote:

This is the final test release before we go to [core] with the new crond.
Please signoff, one for each architecture.

Aside from the upstream changes, two packaging items:

* I've decided to leave the crond logrotate script here for now,
unless/until aaron (syslog-ng maintainer) thinks we should take it into that
package. I'd rather tackle that effort in a separate step, rather than have
it block launching the new 4.x dcron series.

* I've added an optdepend on smtp-server to address FS #12662.


Seems OK here on x86_64. No more zombie sendmail processes either.


Thanks, Dan.

Can I get one for i686? Anyone?

- P

On 01/21/2010 09:46 AM, Paul Mattal wrote:

On 01/19/2010 11:35 PM, Dan McGee wrote:

On Mon, Jan 18, 2010 at 9:33 PM, Paul Mattal<paul@mattal.com> wrote:

This is the final test release before we go to [core] with the new
crond.
Please signoff, one for each architecture.

Aside from the upstream changes, two packaging items:

* I've decided to leave the crond logrotate script here for now,
unless/until aaron (syslog-ng maintainer) thinks we should take it
into that
package. I'd rather tackle that effort in a separate step, rather
than have
it block launching the new 4.x dcron series.

* I've added an optdepend on smtp-server to address FS #12662.


Seems OK here on x86_64. No more zombie sendmail processes either.


Thanks, Dan.

Can I get one for i686? Anyone?


Since nobody has signed off i686, I guess I'll wait until it's a week
without the second signoff and then push it anyway.


That would be on Monday 1/25.

- P

On 24/01/10 15:06, Paul Mattal wrote:

On 01/21/2010 09:46 AM, Paul Mattal wrote:

On 01/19/2010 11:35 PM, Dan McGee wrote:

On Mon, Jan 18, 2010 at 9:33 PM, Paul Mattal<paul@mattal.com> wrote:

This is the final test release before we go to [core] with the new
crond.
Please signoff, one for each architecture.

Aside from the upstream changes, two packaging items:

* I've decided to leave the crond logrotate script here for now,
unless/until aaron (syslog-ng maintainer) thinks we should take it
into that
package. I'd rather tackle that effort in a separate step, rather
than have
it block launching the new 4.x dcron series.

* I've added an optdepend on smtp-server to address FS #12662.


Seems OK here on x86_64. No more zombie sendmail processes either.


Thanks, Dan.

Can I get one for i686? Anyone?


Since nobody has signed off i686, I guess I'll wait until it's a week
without the second signoff and then push it anyway.

That would be on Monday 1/25.


Signoff i686. I do not do any fancy crons though but the basics seem to
work.


Allan

On Sun, Jan 24, 2010 at 12:53 AM, Allan McRae <allan@archlinux.org> wrote:
> On 24/01/10 15:06, Paul Mattal wrote:
>>
>> On 01/21/2010 09:46 AM, Paul Mattal wrote:
>>>
>>> On 01/19/2010 11:35 PM, Dan McGee wrote:
>>>>
>>>> On Mon, Jan 18, 2010 at 9:33 PM, Paul Mattal<paul@mattal.com> wrote:
>>>>>
>>>>> This is the final test release before we go to [core] with the new
>>>>> crond.
>>>>> Please signoff, one for each architecture.
>>>>>
>>>>> Aside from the upstream changes, two packaging items:
>>>>>
>>>>> * I've decided to leave the crond logrotate script here for now,
>>>>> unless/until aaron (syslog-ng maintainer) thinks we should take it
>>>>> into that
>>>>> package. I'd rather tackle that effort in a separate step, rather
>>>>> than have
>>>>> it block launching the new 4.x dcron series.
>>>>>
>>>>> * I've added an optdepend on smtp-server to address FS #12662.
>>>>
>>>> Seems OK here on x86_64. No more zombie sendmail processes either.
>>>
>>> Thanks, Dan.
>>>
>>> Can I get one for i686? Anyone?
>>
>> Since nobody has signed off i686, I guess I'll wait until it's a week
>> without the second signoff and then push it anyway.
>>
>> That would be on Monday 1/25.
>
> Signoff i686. *I do not do any fancy crons though but the basics seem to
> work.
>
> Allan
>

same here. signoff x86_64. BTW, you should set your PACKAGER variable
in makepkg.conf.