Why Phobos is cool

[Avrina]

On Saturday, 27 June 2020 at 20:29:22 UTC, Walter Bright wrote:
> On 6/27/2020 5:07 AM, Avrina wrote:
>>> Many D community members have stepped up at their own 
>>> initiative to become the leader of various parts.
>> Go ahead and list some. I already have a feeling I know what 
>> the trend is going to be like.
>
> Rainer does the Win64 and VS support. Brad Roberts initiated, 
> built, and ran the test servers for many years. He did bugzilla 
> too. Iain does GDC. The LDC team is entirely self-selected. 
> Jacob does the Objective-C support. Vladimir wrote the Dforum 
> software (and many other indispensable tools). And on and on 
> for nearly every part of the D system.
>
> You want something to happen? Step up and make it happen.

There's only so much "stepping up" to make something happen can 
do. The changes that I think need to happen won't actually be put 
into place, in part because a package manager that is disjoint 
from the rest of the development can only do so much. A forum, a 
server, more servers, a bug tracker, miscellaneous tools, and an 
IDE plugin can all operate disjointly without a significant 
impact (for most of the cases) on the quality of the project. 
Part of the root of the problem with Dub lies with DMD. So simply 
having control to make decisions that would potentially cause 
breaking changes only of Dub isn't enough.

I guess you are a bit spoiled when it comes to having projects 
like LDC where someone did step up and took it upon themselves to 
create something that DMD wasn't able to provide. Not sure how 
well it would go, if for example they decided a DIP (like 
DIP1028) shouldn't be implemented and they simply strip it from 
LDC and reject the DIP themselves. Realistically I doubt this 
will happen, and that's in part because they aren't actually 
fully in control.

You care so little about Dub that you'd let just anyone take up 
the reins and do whatever they want with it. Including running it 
into the ground. The thing about individual projects is that they 
don't need Dub. I use a 100 line python script that does 
everything I need, I'm not going to waste my time with Dub. Just 
as DMD doesn't use dub (except for the library version). It's 
simply easier to not use it. No body is going to invest in 
something they can do themselves for much less effort while still 
being able to work on their own projects instead of having to 
shelf them to work on a package manager.

I think ultimately having past successes with having "someone 
step up" and do the work for you and thinking the same thing 
applies to and will work for Dub, is ultimately hurting Dub in 
the long run.

Well I'm done talking about Dub, maybe I'll link this post in 
5-10 years and we'll see what's come of Dub and your wait and see 
approach to developing it.

#dub #fixdub #dub2028