As the third round of beta testing did not turn up any new problems, I'm glad to announce the official release of LDC 0.11.0. Head over to digitalmars.D.ldc for the release announcement (including download links) and further discussion:

http://forum.dlang.org/post/mailman.990.1370788529.13711.digitalmars-d-ldc@puremagic.com

 — David

On 6/9/13 10:46 AM, David Nadlinger wrote:
> As the third round of beta testing did not turn up any new problems, I'm
> glad to announce the official release of LDC 0.11.0. Head over to
> digitalmars.D.ldc for the release announcement (including download
> links) and further discussion:
>
> http://forum.dlang.org/post/mailman.990.1370788529.13711.digitalmars-d-ldc@puremagic.com

Awesome! Any ETA for 0.12.0 built with 2.063's frontend?

Andrei

On Sunday, 9 June 2013 at 16:32:55 UTC, Andrei Alexandrescu wrote:
> Any ETA for 0.12.0 built with 2.063's frontend?

Not really – if I don't get a first beta of the next version out in the next few days, it will unfortunately take *much* longer (exams mid/end August, need to prepare).

Currently working on removing some the most intrusive LLVM-specific changes from the code base, though. We should have worked on reducing that legacy technical debt anyway, as is costing us at every new frontend merge, but there is some funky business going on with CTFE in 2.063 pretty much holding up any further work.

David

On Tue, 2013-06-11 at 05:40 +0200, David Nadlinger wrote: […]
> Not really – if I don't get a first beta of the next version out in the next few days, it will unfortunately take *much* longer (exams mid/end August, need to prepare).
[…]

Is there any way of making the LDC development activity a team one rather than an individual one?

I can't volunteer myself just now as for the next three months at least I am seriously over-committed.  But I do build LDC from master/HEAD for every update and make sure it builds my presentation codes, so I guess I could claim I am already on the alpha-test team.

-- 
Russel. ============================================================================= Dr Russel Winder      t: +44 20 7585 2200   voip: sip:russel.winder@ekiga.net 41 Buckmaster Road    m: +44 7770 465 077   xmpp: russel@winder.org.uk London SW11 1EN, UK   w: www.russel.org.uk  skype: russel_winder

On Tuesday, 11 June 2013 at 08:05:55 UTC, Russel Winder wrote:
> On Tue, 2013-06-11 at 05:40 +0200, David Nadlinger wrote:
> […]
>> Not really – if I don't get a first beta of the next version out in the next few days, it will unfortunately take *much* longer (exams mid/end August, need to prepare).
> […]
>
> Is there any way of making the LDC development activity a team one
> rather than an individual one?

Oh, it absolutely is a team effort already, and that "will unfortunately take" should have been a "might unfortunately take". It's just that Kai Nacke has been focusing on Win64 and PPC64 lately, while Alexey Prokhin is planning to look into fixing our general EH codegen (including exception chaining). And somebody needs to sit down and do the boring work of getting all the test cases to work (again) after a merge has been done.

But the best way to change that would definitely be for somebody to jump in the cold water and help out with development. Doing a front-end merge (resp. the clean-up work after one) specifically is a great way to get started with development, as you usually get in contact with a number of different areas of frontend and glue code in the process. I think that's also how I got started with LDC development.

> I can't volunteer myself just now as for the next three months at least
> I am seriously over-committed.  But I do build LDC from master/HEAD for
> every update and make sure it builds my presentation codes, so I guess I
> could claim I am already on the alpha-test team.

Yes, you are, whether voluntarily so or not. :P

David

On Tuesday, 11 June 2013 at 03:40:16 UTC, David Nadlinger wrote:
> but there is some funky business going on with CTFE in 2.063 pretty much holding up any further work.

Derp. Turns out that actually calling the newly added Expression::init during startup might be a good idea.

Working on getting Phobos to compile right now.

David

"David Nadlinger" <code@klickverbot.at> wrote in message news:qikvjtuajorstsxhhnyr@forum.dlang.org...
>
> Derp. Turns out that actually calling the newly added Expression::init during startup might be a good idea.
>

That must have been interesting.  They are never dereferenced, so no immediate crash, you'd just get whichever code path was tested first...

On Saturday, 15 June 2013 at 06:32:13 UTC, Daniel Murphy wrote:
> That must have been interesting.  They are never dereferenced, so no
> immediate crash, you'd just get whichever code path was tested first...

It was – especially because I was looking for a cause in the DMD/LDC diff first before actually starting to debug the CTFE code. Thinking about it, it might actually be a good idea to add in an assertion:

https://github.com/D-Programming-Language/dmd/pull/2184

David

On Sunday, 9 June 2013 at 16:32:55 UTC, Andrei Alexandrescu wrote:
> Awesome! Any ETA for 0.12.0 built with 2.063's frontend?

Git master is on 2.063.1 now.

There were quite far-reaching internal changes, though, so the current state might not be as stable as the last release was (well, d'oh).

David