[log4cxx] C++ Version

[Robert Middleton]

The current version of next_stable fails when compiled in C++11 mode
due to std::make_unique not being in that version.  This is pretty
easy to fix(Herb Sutter has an implementation of this for C++11 which
seems to be the same as GCC's: https://herbsutter.com/gotw/_102/)

I see three options here:
1. Continue to use C++11 for next_stable, reverting any make_unique calls
2. Use a custom log4cxx::make_unique function for C++11 compilers
3. Use C++14(or later)

Does anybody have any thoughts on this?  At this point I feel it is
perfectly reasonable to do any of these options, as my general rule of
thumb is to support whatever compiler the 'oldstable' version of
Debian uses(so anything released in the past ~5 years).  RHEL has a
much longer release cycle; I believe RHEL 7(which is still supported
for at least 1.5 years) would only be able to be compiled with a C++11
compiler.

-Robert Middleton

Re: Review Request: LOG4J2-3628 Migrate from maven-changes-plugin to a merge-conflict-free Markdown-based solution

[Gary Gregory]

I do like all the details in the Jira ticket! :-)

Gary

On Mon, Nov 7, 2022, 16:03 Volkan Yazıcı  wrote:

> In the context of migrating away from JIRA to GitHub Issues, I want to
> replace the current maven-changes-plugin setup with a merge-conflict-free
> Markdown-based solution. I have elaborated the details in LOG4J2-3628
> .
>
> Since this will change the behaviour of how we introduce changes and make
> releases, I would appreciate it if committers could skim through it and
> share their feedback.
>

Re: Review Request: LOG4J2-3628 Migrate from maven-changes-plugin to a merge-conflict-free Markdown-based solution

[Carter Kozak]

On Tue, Nov 8, 2022, at 14:28, Matt Sicker wrote:
> Any bug tracker we use which isn’t hosted by Apache will likely have similar 
> issues.

The Apache Jira instance has the same issue at this point, even more broadly. I 
agree that the mailing list is a reasonable option for folks to report issues. 
If the existing user or dev lists become inundated with bug reports, we can 
always spin up a new mailing list for tracking and triage, but I suspect that 
will be the exception rather than the rule.

Carter Kozak

Re: Review Request: LOG4J2-3628 Migrate from maven-changes-plugin to a merge-conflict-free Markdown-based solution

[Matt Sicker]

This is why I’ve suggested accepting bug reports via email in addition to 
GitHub. Any bug tracker we use which isn’t hosted by Apache will likely have 
similar issues.

> On Nov 7, 2022, at 8:58 PM, Gary Gregory  wrote:
> 
> Just FYI,
> 
> There is a release note discussion going on on the Maven list and someone
> said the following when I mentioned our plan to switch to GitHub Issues:
> 
> "So you mean using ONLY gh for issues?
> So by doing this, you will exclude people living in countries banned
> from Github.
> Is it acceptable from an Apache Foundation POV?"
> 
> Gary
> 
> On Mon, Nov 7, 2022, 16:03 Volkan Yazıcı  wrote:
> 
>> In the context of migrating away from JIRA to GitHub Issues, I want to
>> replace the current maven-changes-plugin setup with a merge-conflict-free
>> Markdown-based solution. I have elaborated the details in LOG4J2-3628
>> .
>> 
>> Since this will change the behaviour of how we introduce changes and make
>> releases, I would appreciate it if committers could skim through it and
>> share their feedback.
>>