Re: Towards 2.0.6

2016-12-18 Thread Stefan Bodewig
On 2016-11-07, Dominik Psenner wrote:

> LOG4NET-487 is about the configuration option and the issue I would
> like to see in the release. The reason is, not having this
> configurable costs performance and creating a mutex always is
> troublesome to some configurations. The latter resulted in
> LOG4NET-506.

Obviously I haven't managed to invest the time needed to fix it. The
main blocker for myself has been that I may be able to write something
that compiles and doesn't break the existing tests, but I don't think
that's enough for introducing a feature like this shortly before cutting
a release. And I really don't know how to properly test the variations
needed.

Having said that, I'll update the site sources in svn for a release
candidate and unless anybody yells will tag and build an RC and call for
a vote the coming days.

Stefan


Re: Towards 2.0.6

2016-11-09 Thread Dominik Psenner

On 2016-11-09 10:11, Stefan Bodewig wrote:

On 2016-11-07, Dominik Psenner wrote:


LOG4NET-487 is about the configuration option and the issue I would
like to see in the release. The reason is, not having this
configurable costs performance and creating a mutex always is
troublesome to some configurations.

I should be able to carve out some time the coming days and will look
into LOG4NET-487 unless anybody beats me to it. Would you be OK with me
calling a release vote after that?


It is something I would like to see in 2.0.6, which doesn't imply that 
it must be part of 2.0.6. I'm going to vote for 2.0.6 even if 
LOG4NET-487 is not part of it, so long 2.0.6 passes my checks.


Cheers


Re: Towards 2.0.6

2016-11-09 Thread Stefan Bodewig
On 2016-11-07, Dominik Psenner wrote:

> LOG4NET-487 is about the configuration option and the issue I would
> like to see in the release. The reason is, not having this
> configurable costs performance and creating a mutex always is
> troublesome to some configurations.

I should be able to carve out some time the coming days and will look
into LOG4NET-487 unless anybody beats me to it. Would you be OK with me
calling a release vote after that?

Stefan


Re: Towards 2.0.6

2016-11-06 Thread Stefan Bodewig
So this would be something you'd like to see adressed before the
release, right?

I think this is https://issues.apache.org/jira/browse/LOG4NET-485 which
is linked by https://issues.apache.org/jira/browse/LOG4NET-506 - only
the later is still open.

And there is https://issues.apache.org/jira/browse/LOG4NET-487

LOG4NET-485 is resolved with fix version 1.2.14 and the correspondig
commit (svn revision 1711838) really is part of 1.2.14.

https://svn.apache.org/viewvc/logging/log4net/tags/1.2.14/src/Appender/RollingFileAppender.cs?r1=1707180=1711838

Stefan

On 2016-11-06, Dominik Psenner wrote:

> We had introduced a mutex to secure rolling operations in the rolling file
> appender. It would be nice if we could make the rolling locking
> configurable (none, lock, mutex), defaulting to none. There should be an
> issue for that, too.

> On 6 Nov 2016 2:19 p.m., "Joe" <jocular...@hotmail.com> wrote:

> I just had a quick look through the commits since 1.2.15 and I don't see
> anything that looks particularly risky.
> I'd vote for cutting an RC.

> -Original Message-
> From: Stefan Bodewig [mailto:bode...@apache.org]
> Sent: 06 November 2016 11:14
> To: log4net-dev@logging.apache.org
> Subject: Towards 2.0.6

> Hi all

> I'd like to get 2.0.6 released in order to help out people who want to use
> log4net with .NET Core.

> We've had some changes since I built the last test assemblies, but AFAIK
> nobody has given them a try anyway. I wonder whether creating test
> assemblies is worth the effort or whether I should simply proceed with
> cutting a release candidate and call for a vote.

> So what do you think, should I just cut an RC? Is there anything floating
> around that should go into 2.0.6?

> Stefan


RE: Towards 2.0.6

2016-11-06 Thread Dominik Psenner
We had introduced a mutex to secure rolling operations in the rolling file
appender. It would be nice if we could make the rolling locking
configurable (none, lock, mutex), defaulting to none. There should be an
issue for that, too.

On 6 Nov 2016 2:19 p.m., "Joe" <jocular...@hotmail.com> wrote:

I just had a quick look through the commits since 1.2.15 and I don't see
anything that looks particularly risky.
I'd vote for cutting an RC.

-Original Message-
From: Stefan Bodewig [mailto:bode...@apache.org]
Sent: 06 November 2016 11:14
To: log4net-dev@logging.apache.org
Subject: Towards 2.0.6

Hi all

I'd like to get 2.0.6 released in order to help out people who want to use
log4net with .NET Core.

We've had some changes since I built the last test assemblies, but AFAIK
nobody has given them a try anyway. I wonder whether creating test
assemblies is worth the effort or whether I should simply proceed with
cutting a release candidate and call for a vote.

So what do you think, should I just cut an RC? Is there anything floating
around that should go into 2.0.6?

Stefan


RE: Towards 2.0.6

2016-11-06 Thread Joe
I just had a quick look through the commits since 1.2.15 and I don't see 
anything that looks particularly risky.
I'd vote for cutting an RC.

-Original Message-
From: Stefan Bodewig [mailto:bode...@apache.org] 
Sent: 06 November 2016 11:14
To: log4net-dev@logging.apache.org
Subject: Towards 2.0.6

Hi all

I'd like to get 2.0.6 released in order to help out people who want to use 
log4net with .NET Core.

We've had some changes since I built the last test assemblies, but AFAIK nobody 
has given them a try anyway. I wonder whether creating test assemblies is worth 
the effort or whether I should simply proceed with cutting a release candidate 
and call for a vote.

So what do you think, should I just cut an RC? Is there anything floating 
around that should go into 2.0.6?

Stefan