Back from the dead - How is D going?

[Don]

Bruno Medeiros wrote:
> Thanks everyone for the links and comments. I see a lot has happened 
> meanwhile, and I have a lot to catch up (annotations, safety features, 
> concurrency changes like thread-local by default and whatnot, and other 
> minor things)
> 
> http://www.prowiki.org/wiki4d/wiki.cgi?LanguageDevel#FutureDirections
> Reading this I like a lot of the minor changes coming in. Like static 
> arrays becoming value types. Or AAs and complex types moving from 
> language to library, I think its good to make the language simpler 
> whenever no significant amount of power or usefulness is lost (although 
> I guessing the powers that be only agreed to move this to the library 
> because now D's metaprogramming features, as well as new struct features 
> and others, make it more feasible to implement these types in the library).

Exactly. They're also an important test: are the built-in features 
*really* good enough to implement all this stuff?

> "Remove C-style declarations" -> again, quite good, simplifying language 
> without any loss in power.
> 
> "Remove typedef" Hum, Im curious about this one, but I'll read the 
> appropriate threads later when I have the time.
> 
> I also like this blog post as well:
> http://www.jfbillingsley.com/blog/?p=53
> 
> Lutger wrote:
>  >
>  > Hi, welcome back!
>  >
>  > Here are a couple of bullet points for you, not sure what you missed and
>  > what not:
>  >
>  > - Andrei Alexandrescu is writing (has almost completed) a book on D, 
> to be
>  > published around before april 2010 iirc
>  >
>  > - Around the same time, D2 should be finished.
>  >
> 
> Interesting. Does that mean that the D's design with regards to 
> concurrency will be considered finished? (or that at least no major 
> changes are expected?) And that D's designers (Andrei & co.) are happy 
> with it?

Anything which is written in Andrei's book is a permanent feature of the 
language. Hence the current rush to fix up the loose ends and remove 
cruft. There are still a few unresolved issues with operator overloading 
and with concurrency, and with aspects of the safe-D system.

> I was thinking that D2 would take longer to be finished, but it is good 
> that it gets done soon. This way maybe we can start moving away from the 
> current limbo state were the core majority of libraries and tools are 
> pending in D1, which seems to be one of the great roadblocks at the 
> moment for D's growth and development.

Definitely. After the book is published, the emphasis will switch to the 
libraries, bug fixes, and the rest of the tool chain.