DUB RC 0.9.24-rc.1 ready for testing

[Andrei Alexandrescu via Digitalmars-d]

On 7/21/15 3:00 AM, Martin Nowak wrote:
> On Monday, 20 July 2015 at 23:18:34 UTC, Andrei Alexandrescu wrote:
>> On 7/20/15 5:30 PM, Martin Nowak wrote:
>>> On Thursday, 16 July 2015 at 08:28:08 UTC, Suliman wrote:
>>>> In what version of DMD do you plan to include dub and vibe?
>>>
>>> It doesn't make sense to include vibe.d.
>>
>> I think it does - this has been discussed before. -- Andrei
>
> It has, in length
> http://forum.dlang.org/post/mdnrus$188e$1@digitalmars.com, but you
> remain one of the very few to think it is a good idea to distribute
> vibe.d with dmd.

Probably I need to better explain why I think we should try that.

It all starts with a high level thought. We want to accelerate D 
adoption rate way beyond what it is now. Radically, like 10x. We've done 
a number of things, many of which helped. But there's this thought - if 
we keep on doing what we've been doing, we'll keep on getting the 
results we've been getting. (There could be changes of phase, synergy, 
cumulative effects etc. but just waiting for those to happen doesn't 
sound like the best tactics.)

So I keep an eye on radical new things we could try - things we have not 
done before, and that have worked for others. Some might just not work, 
but we don't know if we don't just try.

At the first D meetup in the Silicon Valley, Vic (an accomplished 
entrepreneur who has been following up D'd path) discussed some ideas 
for improving D's adoption. He mentioned some other languages have 
improved adoption by means of a strong application (e.g. Rails for Ruby) 
and suggested we make vibe.d, which is one of D's most compelling 
publicly available applications, more prominent among D's toolchain. He 
mentioned that many folks start with the high-level need ("I need a web 
framework") and accept the language as an artifact.

I think that's a good idea to try, for several reasons. First, it will 
enhance vibe.d's visibility (there are quite a few D users who haven't 
heard of vibe). Second, it consolidates things and makes it easy for 
folks who want to get vibe.d - no more version incompatibilities, 
multiple downloads, things that don't mesh etc. Third, it provides even 
non-vibe users with a good example of a large framework written in D 
that they may use for inspiration and good language use.

> It doesn't make sense because dub is the enabling tool for the whole
> package ecosystem, with which vibe.d is fully integrated (dub was
> formerly called vpm - vibe package manager).
> Copying a vibe.d version into the distribution creates a lot of problems
> without solving anything.

I agree it does not "solve" anything the same way Rails does not "solve" 
anything for Ruby.

> - what about vibe.d's dependencies
> - how would dub know about the distributed vibe.d package
> - how to use multiple vibe.d versions in parallel

These may be framed two ways. One is, these are arguments to not 
integrate vibe.d with D. The other is, we buy into the vision that we 
should try bundling vibe with D, and as a matter of course we need to 
solve some practical matters. Clearly these all are solvable.

> If your long-term goal is to integrate vibe into phobos, fine,
> though I'm not a fan of this strategy b/c an independent package
> ecosystem can grow faster.
> Simply copying a dub package into the distribution doesn't help anyone.

I think it could help. This is like one of those business ideas 
discussions - there are lots of reasons for which any isn't going to 
work. The point is finding ways to make it work.


Andrei