Spreading the word about D

[Marianne Gagnon]

> > Those who're maintaining libraries, great job, don't stop!  We need a
> > solid foundation to build on!  But I think that a lot of people take one
> > look at D, see a mess of libraries on DSource (a lot of which are
> > abandoned or don't even build correctly) and sign it off as a language
> > with powerful features but a weak support system.  If we can demonstrate
> > that D is more than just that, we'll have provided proof that D is not a
> > cool gizmo, but a real, powerful tool.
> > 
> > At least, that's just my $0.02.
> > 
> 
> D doesn't even have an agreed upon standard library, and you want the
> libs on dsource to not look like a mess?!
> 
> there is no point building the roof when the foundation isn't ready yet.
>  First thing's first: D needs a standard lib. not phobos/tango/tangobos/
> downs' libs/bearophile's libs/etc... but a /STANDARD/ lib. One agreed
> upon IO API, One agreed upon concurrency API., etc...
> 
> people nowadays are accustomed to the the Java standard libs/the .net
> framework and anything less than that is perceived as unprofessional.
> A programmer friend of mine switched from Ruby to Python even though he
> thinks (like me) that the Ruby language has a the superior syntax and is
> better designed (passing the this pointer manually, wtf?)
> The main reason is of course the Python libs. you get a lot more libs in
> the standard python dist and beyond that many more 3rd party libs.
> 
> with D, you need to look for 3rd party libs for things that really
> should be in the std lib. good example is threading, either you go with
> tango which is 3rd party even if you consider it as the de-facto std
> lib, or with a different solution like downs' tools for phobos.
> 

I am personnaly only watching D from a distance and continuously am evaluating it for my own use. (I do not often post here, but perhaps my opinion may have some use since i'm just an avergae programmer trying to get the job done and not an enthusiastic)

When I look at D, my mouth waters at the awesome language design, but I shiver in fear whenever I look at libraries available. All this bunch of librairies is unstable, often unmaintained, clunky. As already mentionned often, the lack of standard library does not help. But another thing, i think, is that D is now trying to incorporate every feature in the world instead of just getting stable and letting library makers work.

So in short, I think it's too early to widely advertise D, it would IMO only give a bad impression. I think it's time to stop trying to add every possible feature (hell, java doesn't have half of D's features and can still make about any program) and instead focus on getting a bit more stable and standard. Then GDC can catch up, then other librairies can catch up and build upon a reliable base. Then IDEs can add D support and/or be developped.

When progress has been done on this side, only then can D be successfully advertised. My current position is generally "D looks very nice, but the poor developement environment makes the syntaxic improvements not worth dropping my current C++ environment"