Jump to page: 1 24  
Page
Thread overview
A New Era for the D Community
May 03, 2023
Mike Parker
May 03, 2023
Ikey Doherty
May 03, 2023
Nicholas Wilson
May 03, 2023
Mike Parker
May 03, 2023
Ali Çehreli
May 03, 2023
Bradley Chatha
May 03, 2023
MissPiggy
May 03, 2023
max haughton
May 04, 2023
MissPiggy
May 03, 2023
Walter Bright
May 04, 2023
Don Allen
May 17, 2023
Theo
May 04, 2023
Ivan Kazmenko
May 05, 2023
Ogi
May 05, 2023
Mike Parker
May 13, 2023
ryuukk_
May 13, 2023
ryuukk_
Jul 05
IchorDev
Jul 05
Andrew
Jul 06
IchorDev
May 14, 2023
M.M.
May 06, 2023
ikelaiah
May 07, 2023
monkyyy
May 09, 2023
FeepingCreature
May 13, 2023
monkyyy
May 09, 2023
Paul Backus
May 26, 2023
Dejan Lekic
May 03, 2023

At DConf '22, Roberto Ierusalimschy, the lead designer of the Lua programming language, was our guest keynote speaker. At the end of his talk, Robert Schadek asked him how he organized the Lua community. His answer: Is it too ugly if I say I don't?

In my talk at the same conference, I highlighted part of the history of the D community's evolution. I used the metaphor of a pioneer settlement that evolved into a village and then into a town. In those early pioneering days, Walter's approach to organizing the community was the same as Roberto's. He didn't need to do anything. The community organized itself and built from scratch the foundations of the ecosystem we have today. If you discovered D and expected only to be a user and never a contributor, then D just wasn't for you. Of course, as the community grew and evolved, and built up more of the ecosystem, there was increasing space for noncontributors. That, in turn, led to shifting expectations.

At different points in D's history, Walter, Andrei, and motivated community members took steps to adapt to changing expectations. For several years they were able to keep up reasonably well. To give just one example, in the earliest days, users posted bug reports in the forums or by emailing Walter, and contributors emailed him patches. In response to calls for better bug management, a community member volunteered to maintain a Bugzilla instance. Then later, when people were wondering why they had to submit patches to Bugzilla when GitHub existed, Walter was persuaded to put the compiler's source on GitHub.

By the time Walter and Andrei established the D Language Foundation in 2015, they had settled into a very loose management style. Motivated community members volunteered to create services, or take charge of something in the ecosystem, and Walter and Andrei would give their blessing. I can't speak to what their goals were with the Foundation, but they largely continued that approach to managing the ecosystem. Unfortunately, as the community continued to grow and expectations continued to evolve, that approach was unsustainable.

Ten years ago, if the forums went down, everyone knew to contact Vladimir Panteleev. Today, many people don't know, and probably don't care, that he pays for and maintains the server on which the forums are hosted. The forums are on dlang.org, the official website of the D Language Foundation, so the DLF is responsible for getting things back up. When a bug report remains open for years, it doesn't matter that it's because it hasn't come to the attention of someone with the time, ability, and motivation to fix it. The DLF is responsible for organizing resources to address reported issues, and if we can't, that's on us.

One of the biggest complaints I've heard over the past few years is some form of "lack of leadership/management/vision". It's painful to hear, as I know that everyone involved with the DLF is personally invested. We're here because we love what we're doing. Yet, the criticism is on the mark.

Yes, there have been improvements over the past few years. The quarterly DLF meetings with industry representatives, initially proposed by Nicholas Wilson, have been productive. The monthly meetings that grew out of those have led to several positive changes. Symmetry's sponsorship of the Pull Request and Issue Manager positions held by Razvan Nitu and Dennis Korpel has been a huge boon. We've begun migrating some ecosystem services to DLF servers (it's going very slowly, but it's happening). We've done several good things that I could enumerate here. But collectively, it's the equivalent of being surrounded by small fires and running around to put them out at random. We simply do not have a structured system of management.

Several times over the past couple of years, we discussed what to do about it. We read books, watched presentations, and asked for advice. We got nowhere.

Then, in July of last year, Paul Toth of Ucora reached out with an offer. Had I known at the time that it was going to change everything, I would have put it at the top of my priority list. Alas, it wasn't until November that things started moving.

Ucora’s long-term vision is to change the way the world works. As part of their mission, they provide organizations with the tools they need to transform the way they work. IVY, their organizational development program, is a simple but innovative approach to workflow. Ucora has been using D for several years. They're invested in D's success, and so they want the DLF to be successful. Paul offered to put the DLF team through the IVY program at no charge. We accepted, and every Friday for the past 14 weeks we've been having sessions with Saeed Sabeti, Ucora's Director of Organizational Development. May 5th will be our 15th and final session.

This has been a transformative experience. In part, it's been a journey of self-discovery. In discovering ourselves, we've learned more about each other. We now have a deeper insight into what motivates and demotivates each of us, why we're devoting a chunk of our lives to D, and how that knowledge can help us in our work. Importantly, we've also discovered our vision for the DLF and learned how to view our work through the IVY Lens.

The result is that we now have the tools we need to build up that structured system of management that we've been missing, and a much clearer view of how to get there. In the coming months, that's precisely what we're going to focus on. We'll be having regular meetings outside of our monthlies to make it happen.

You're going to hear more about IVY as time goes by, and eventually, we're going to start employing it more broadly in the community. We now have a better idea of how to more effectively guide contributors so that they can be more efficient and stay motivated. Before we get to that point, we've got a lot of decisions to make and a lot of work to do internally to provide a foundation on which we can build.

I'm not exaggerating when I say that this is going to be the most significant change in the D community in the 20 years I've been a part of it. I expect we're going to encounter bumps along the way, but that's okay. We now have a clear vision and purpose, and that makes all the difference.

Those of us who participated in the IVY program are grateful to Ucora for presenting us with the opportunity. Not only will IVY help us in guiding D, but we can take what we've learned and apply it to other areas of our lives. They've given us a valuable gift.

And on a personal note, I want to thank all of the DLF associates who participated. Going into it, everyone was hopeful something good would come of it, but I think most of the group were unsure if it would be helpful. None of us knew what to expect. I may have been the only one truly excited about it, and still, I was concerned I was being overly optimistic. As we continued the sessions, it was fun to see when someone finally "got it". In the end, we all did.

Our enthusiasm is high, and we're ready to get going. I think you'll like where we're headed.

May 03, 2023

Sounds awesome Mike, really does. As long as transparency remains a key objective, I think everyone is on board.

May 03, 2023
Very interesting.

I just want to say, I'm having repeated unrelated CI failures for: https://github.com/dlang/phobos/pull/8699

Yesterday I finally said right stuff this I don't have the energy to fight the CI on this anymore and asked Razvan to kinda take over as the changes on my end are fundamentally done.

This may not seem very important, but to Symmetry, without Razvan I'd honestly just close that PR and leave 500ms reduction in compile times when you import std.regex on the table. Having this extra support in the community who can tackle this sort of issues is both highly appreciated and highly motivational for unblocking people in the future.

Much appreciation all round, long may it continue!
May 03, 2023

On Wednesday, 3 May 2023 at 11:13:34 UTC, Mike Parker wrote:

>

Yes, there have been improvements over the past few years. The quarterly DLF meetings with industry representatives, initially proposed by Nicholas Wilson, have been productive.

From memory that was merely an AGM at DConf, but sure, I'll take credit for that (despite knowing next to nothing about management)!

May 03, 2023

On Wednesday, 3 May 2023 at 11:37:31 UTC, Nicholas Wilson wrote:

>

From memory that was merely an AGM at DConf, but sure, I'll take credit for that (despite knowing next to nothing about management)!

The AGM was your idea, but so were the regular meetings. You pushed it to Ali in 2018. I have a second-hand email record!

May 03, 2023
On 5/3/23 05:08, Mike Parker wrote:
> On Wednesday, 3 May 2023 at 11:37:31 UTC, Nicholas Wilson wrote:
> 
>> From memory that was merely an AGM at DConf, but sure, I'll take credit for that (despite knowing next to nothing about management)!
> 
> The AGM was your idea, but so were the regular meetings. You pushed it to Ali in 2018. I have a second-hand email record!

Wow! I wouldn't have remembered but it's an email I wrote on November 3, 2018, which starts with "Nicholas Wilson visited the Bay Area for an LLVM conference and I had a couple of chances to see him."

Good old days! :)

Ali

May 03, 2023

On 5/3/23 7:13 AM, Mike Parker wrote:

>

Our enthusiasm is high, and we're ready to get going. I think you'll like where we're headed.

This all sounds awesome!

-Steve

May 03, 2023

On Wednesday, 3 May 2023 at 11:13:34 UTC, Mike Parker wrote:

>

...

Exciting news!

May 03, 2023
On Wednesday, 3 May 2023 at 11:13:34 UTC, Mike Parker wrote:
>
> Our enthusiasm is high, and we're ready to get going. I think you'll like where we're headed.

IVY may increase the probability of a particular outcome, but it cannot predict a particular outcome.

We still can't even predict weather forecasts all that well either.

You simply cannot ever know the position and velocity of every atom in the universe.

> You're going to hear more about IVY as time goes by, and eventually,
> we're going to start employing it more broadly in the community.

Whether you can in fact 'deploy' IVY into the 'broader community'... well..that remains to be seen.

It's more likely, I think, that D's 'future' will be increasingly determined by the priorities of a select group of corporations, rather than any psychological/behavioural science.
May 03, 2023
This initiative has my full support.
« First   ‹ Prev
1 2 3 4