On Tuesday, 8 March 2016 at 21:32:28 UTC, Andrei Alexandrescu wrote:
>
> Again people discuss the GC in D. -- Andrei

Are you guys making progress with rcstring and lifetimes?

On 3/8/2016 7:54 AM, Dmitry Olshansky wrote:
> This more or less means that we (as in D enthusiasts) have some more time to
> carve up some "market" share. Till C++20 I guess.

Back in the olden (Zortech C++) days, lots of people told me that the next release of Microsoft C++ was going to be so awesome, it would bury us. That we might as well go buy our caskets and reserve a burial plot right now. This drumbeat would continue right up until the release.

The release was always underwhelming, and ZTC++ would find even more customers.

The same pundits would then say, yeah, that release wasn't that great. But the *next* release, that one will bury Zortech!

And so it would go on for every release.

Microsoft only buried us after we stopped working on it.

On Tuesday, 8 March 2016 at 18:24:54 UTC, Minas Mina wrote:
> I honestly don't care about those. Boost has them. Modules are far more important for me.

Just found out that Modules weren't supposed to be scheduled for C++17 anyway, but was already meant to be a separate TS. Seems like the C++ standard process will be more continuous than before.

On Tuesday, 8 March 2016 at 22:31:39 UTC, Walter Bright wrote:
> Back in the olden (Zortech C++) days, lots of people told me that the next release of Microsoft C++ was going to be so awesome, it would bury us. That we might as well go buy our caskets and reserve a burial plot right now. This drumbeat would continue right up until the release.

 Didn't MS spout the same 'awesome' OS they were working on that would have all the same features or better, and prevent OS2 from being adopted?

On Tuesday, 8 March 2016 at 22:53:42 UTC, Era Scarecrow wrote:
>  Didn't MS spout the same 'awesome' OS they were working on that would have all the same features or better, and prevent OS2 from being adopted?

Windows NT was known for having a decent kernel. Vapour-ware was a common marketing strategy in the 80s and 90s, though. Not Microsoft specific.

On 3/8/2016 2:53 PM, Era Scarecrow wrote:
>   Didn't MS spout the same 'awesome' OS they were working on that would have all
> the same features or better, and prevent OS2 from being adopted?

It wasn't Microsoft telling me that their next C++ release would bury Zortech. It was everyone else.

On Tuesday, 8 March 2016 at 22:53:01 UTC, Ola Fosheim Grøstad wrote:
> On Tuesday, 8 March 2016 at 18:24:54 UTC, Minas Mina wrote:
>> I honestly don't care about those. Boost has them. Modules are far more important for me.
>
> Just found out that Modules weren't supposed to be scheduled for C++17 anyway, but was already meant to be a separate TS. Seems like the C++ standard process will be more continuous than before.

Is the modules specification finalized though? As far as I know there were two implementations, clang's and vc++'s.

On Wednesday, 9 March 2016 at 06:08:57 UTC, Minas Mina wrote:
> Is the modules specification finalized though? As far as I know there were two implementations, clang's and vc++'s.

I don't think it finalized. BTW, I think some of the Microsoft people have deleted some of their reddit comments. I can no longer find the statement where it was claimed that VC++ was left alone for two years in the Ballmer era. But it is rather obvious that VC++ needs more work than the other compilers, not sure how that affects the standardization process.