Hi all,

On 19 Jun 2014, at 15:50, David Nadlinger wrote:
> I think it is long overdue that we decide how to go on about doing the
> merge of 2.064. At this point, it is probably more "stable" (in the
> sense of bug-free) than master, simply because of all the fixed
> frontend bugs.

I fixed the Travis test timeout issue and merged the merge-2.064 branch to master. I hope everybody is okay with this, but we were definitely treating it as our de-facto main branch already. The libcurl heisenbug may come back to break the build, but that's still much better than the confusion that the previous situation created.

Best,
David

On Sunday, 22 June 2014 at 17:33:49 UTC, David Nadlinger via digitalmars-d-ldc wrote:
> Hi all,
>
> On 19 Jun 2014, at 15:50, David Nadlinger wrote:
>> I think it is long overdue that we decide how to go on about doing the
>> merge of 2.064. At this point, it is probably more "stable" (in the
>> sense of bug-free) than master, simply because of all the fixed
>> frontend bugs.
>
> I fixed the Travis test timeout issue and merged the merge-2.064 branch to master. I hope everybody is okay with this, but we were definitely treating it as our de-facto main branch already. The libcurl heisenbug may come back to break the build, but that's still much better than the confusion that the previous situation created.
>
> Best,
> David

Hi David!

Sorry for the long delay. Shouldn't we create a release based on the current state? The heisenbug was not reported by any user - only the Travis build failed. We could then merge the merge-2.065 branch. I think this branch contains only a couple of problems - in addition to the existing code emission problems.

Regards,
Kai

Hi Kai,

On 22 Jun 2014, at 20:23, Kai Nacke via digitalmars-d-ldc wrote:
> On Sunday, 22 June 2014 at 17:33:49 UTC, David Nadlinger via digitalmars-d-ldc wrote:
>> The libcurl heisenbug may come back to break the build, but that's still much better than the confusion that the previous situation created.

Guess what happened for the 3.1 release builds right after I posted this message… What do we really do about this now?

> Sorry for the long delay. Shouldn't we create a release based on the current state?

I think this would be a good idea, yes. Can you get things rolling? I'm unfortunately rather busy the next few weeks starting tomorrow, but I'd certainly happy to build the OS X packages again.

Best,
David

On Sunday, 22 June 2014 at 21:08:17 UTC, David Nadlinger via digitalmars-d-ldc wrote:
> Hi Kai,
>
> On 22 Jun 2014, at 20:23, Kai Nacke via digitalmars-d-ldc wrote:
>> On Sunday, 22 June 2014 at 17:33:49 UTC, David Nadlinger via digitalmars-d-ldc wrote:
>>> The libcurl heisenbug may come back to break the build, but that's still much better than the confusion that the previous situation created.
>
> Guess what happened for the 3.1 release builds right after I posted this message… What do we really do about this now?

:-) I see.

>> Sorry for the long delay. Shouldn't we create a release based on the current state?
>
> I think this would be a good idea, yes. Can you get things rolling? I'm unfortunately rather busy the next few weeks starting tomorrow, but I'd certainly happy to build the OS X packages again.

Yes, I start the release process today. I'll ping you when I had set the tags...

Regards,
Kai