On Thursday, 28 May 2020 at 21:16:31 UTC, Bastiaan Veelo wrote:

>
> Your point is that the rules were followed as intended, right? Their point is that the rules should be changed to remove bias.
>

Let me put it another way. The DIP process is something that, as DIP manager, it's in my power to change by revising the documentation. But it's not up to me to decide who or how many people make decisions about language features. That step is *outside* of the DIP process. If they wanted to, they could very well decide to implement new features without ever submitting a DIP.

The reason I'm arguing semantics is that on more than one occasion I have seen people who misunderstand what the DIP process is intended to achieve (e.g., people who assume it's a community vote). When people complain that "the DIP process is broken", then somewhere down the line on reddit or discord or somewhere else I'll inevitably run into someone who considers it broken from beginning to end because they saw in the forums that "the DIP process is broken". When enough people say it, that kind of thing can spread.

I'm happy to discuss changes to the DIP process and I'm willing to make them if I can be convinced they're needed. I've done it before. We all want the best process we can have. But changing how DIPs are approved is out of bounds for me and therefore not part of the DIP process.

Anyone who would like to see changes to the decision-making process is welcome to send suggestions or proposals to me, however, as Bruce has done. Then I can put that on the agenda of the next foundation meeting.

Am 29.05.20 um 05:18 schrieb Mike Parker:
> On Thursday, 28 May 2020 at 21:16:31 UTC, Bastiaan Veelo wrote:
> 
>>
>> Your point is that the rules were followed as intended, right? Their point is that the rules should be changed to remove bias.
>>
> 
> Let me put it another way. The DIP process is something that, as DIP manager, it's in my power to change by revising the documentation. But it's not up to me to decide who or how many people make decisions about language features. That step is *outside* of the DIP process. If they wanted to, they could very well decide to implement new features without ever submitting a DIP.
> 
> The reason I'm arguing semantics is that on more than one occasion I have seen people who misunderstand what the DIP process is intended to achieve (e.g., people who assume it's a community vote). When people complain that "the DIP process is broken", then somewhere down the line on reddit or discord or somewhere else I'll inevitably run into someone who considers it broken from beginning to end because they saw in the forums that "the DIP process is broken". When enough people say it, that kind of thing can spread.
> 
> I'm happy to discuss changes to the DIP process and I'm willing to make them if I can be convinced they're needed. I've done it before. We all want the best process we can have. But changing how DIPs are approved is out of bounds for me and therefore not part of the DIP process.
> 
> Anyone who would like to see changes to the decision-making process is welcome to send suggestions or proposals to me, however, as Bruce has done. Then I can put that on the agenda of the next foundation meeting.
> 
> 

OK, now I understand why you view it that way and it makes sense. Thanks for clarifying that. Personally, I did not view it that way until now (and I believe many others also did not). For me the DIP process was the complete process you need to go through in order to get a change to the language accepted or rejected, including formal assessment.

And from the documentation of the DIP process, it certainly reads that way as I have laid out in an earlier post. If formal assessment really is not part of what we call the "DIP process", the documentation needs to be adjusted to make that clear.

However, from the perspective of anybody else than you, it doesn't really make a lot of sense. Why would you ever consider the process without the actual decision making in the end? For the community, it doesn't really matter that you personally don't have any power regarding the decision-making process. We want to talk about the whole process of getting a change into the language and it's just cumbersome to always say "the DIP process and the decision making process"...

On Friday, 29 May 2020 at 04:27:18 UTC, Johannes Loher wrote:

>
> However, from the perspective of anybody else than you, it doesn't really make a lot of sense. Why would you ever consider the process without the actual decision making in the end? For the community, it doesn't really matter that you personally don't have any power regarding the decision-making process. We want to talk about the whole process of getting a change into the language and it's just cumbersome to always say "the DIP process and the decision making process"...

Yes, I understand. What it really comes down to is that I'm just asking people to be more careful with their language. These forums do not exist in a sealed vacuum, and what we say here can easily be misinterpreted elsewhere or by someone with an interest in D who's looking over the forums. "It's messed up that a DIP author can decide on his own proposal" is very different from a broad claim like "the DIP process is broken".

Am Thu, 28 May 2020 20:59:19 +0000 schrieb Ethan:

> On Thursday, 28 May 2020 at 20:55:23 UTC, Bastiaan Veelo wrote:
>> But the DIP text itself did not convince the other language maintainer. A phone call did. It did not go through on its own merits.
> 
> Even with a board of more than two members, I'd expect them to discuss dissenting opinions before coming to a conclusion.
> 
> Discussion isn't a problem here, it's impartiality (and lack thereof).

To put it more explicitly, discussion between review board members is expected.

Discussion between DIP author and review board considering technical details: I'm not sure. Ideally a DIP should convey all important information on its own. A direct technical discussion between DIP author and review board is then not necessary.

-- 
Johannes

On Fri, May 29, 2020 at 3:10 AM Ethan via Digitalmars-d < digitalmars-d@puremagic.com> wrote:

> On Thursday, 28 May 2020 at 02:39:16 UTC, Manu wrote:
> > OMFG... Wow... what on earth happened here while I wasn't looking!
>
> I wasn't looking either. I don't have a ton of time to devote to non-work activities at the moment.
>
> I have to say though. I don't think I've ever seen the D community this unified.
>

Yeah, I'm just desperately struggling to keep on top of my real work :/ This was a weird thing.

On Thursday, 28 May 2020 at 22:54:07 UTC, Andrei Alexandrescu wrote:
>
> Quite quickly it goes to the same fixed point I mentioned a few days ago: We need a few more folks of Walter's caliber. Whom we don't have. (They're in awfully short supply is part of the problem.) Define roles all we want, shuffle human resources (real or imaginary) all we want, strategize all we want, we can't get around this obvious obstacle. Much of the current kerfuffle is the panic resulting from the fact that he is just like anyone liable to make mistakes.

I think that the obvious problem of this drama and also the D project as a whole, is that the project doesn't scale beyond a startup like project run by to close buddies. The company that succeeds hire a professional manager (typically a CEO) and the ones that don't just keep on going. There will be occasions where the founders will not agree but the third party ensures the interest of the company. I've seen this dozens of times right now and it is often when the technical staff tries to go on project management without any experience. Either the company sends some of the programmers to some kind of leadership course or similar or hire a professional, then things usually improve, a little bit at least. The company that doesn't act and believe programmers have some hereditary project management abilities, always fails with the targets and the project as a whole.

What this project needs is more people on board and that's not necessarily need to be a technical expert but someone whose responsibility to ensure that the D language becomes successful. This is the difference between Rust and D. Rust has professional managers and can afford it, D has not the resources for this.

The D project has two directions now, either let this where the community decide the direction of D or D should be run by some kind board. I'm not so sure about if the community should have too much to say because that might hinder the progress of D. People will always disagree so it is much more difficult. You Andrei is always welcome back. Would you reconsider even if it would less time as the board would increase?

Also the panic around DIP 1028 also shows the difference in expectations of the community input.

On Friday, 29 May 2020 at 05:03:25 UTC, Mike Parker wrote:
> On Friday, 29 May 2020 at 04:27:18 UTC, Johannes Loher wrote:
>  "It's messed up that a DIP author can decide on his own proposal" is very different from a broad claim like "the DIP process is broken".

Not that different. All the more reason for the DIP process to be fixed than leave it the way it is. You are just arguing semantics at this point to try and avoid bad PR. You care more about public imagine than due process. That's backwards.

On Friday, 29 May 2020 at 11:14:31 UTC, Gregory wrote:

>
> Not that different. All the more reason for the DIP process to be fixed than leave it the way it is. You are just arguing semantics at this point to try and avoid bad PR. You care more about public imagine than due process. That's backwards.

Well, for one, caring about D's image is part of what I do. For another, I'm not arguing against changes to the decision-making process (or changes to any part of the DIP process), so I really don't know what you're on about.

On Friday, 29 May 2020 at 11:23:22 UTC, Mike Parker wrote:
> On Friday, 29 May 2020 at 11:14:31 UTC, Gregory wrote:
>
>>
>> Not that different. All the more reason for the DIP process to be fixed than leave it the way it is. You are just arguing semantics at this point to try and avoid bad PR. You care more about public imagine than due process. That's backwards.
>
> Well, for one, caring about D's image is part of what I do. For another, I'm not arguing against changes to the decision-making process (or changes to any part of the DIP process), so I really don't know what you're on about.

So there's no problem with the process, but you aren't against fixing the process? Which is it?

On Friday, 29 May 2020 at 13:12:22 UTC, Gregory wrote:

>
> So there's no problem with the process, but you aren't against fixing the process? Which is it?

If you read my previous posts in this thread, I have said I don't believe the process is currently broken. But I also said this:

"I'm happy to discuss changes to the DIP process and I'm willing to make them if I can be convinced they're needed. I've done it before. We all want the best process we can have."

https://forum.dlang.org/post/eeiijnpjevxwlfesccnb@forum.dlang.org

So, yes, it really is both. I don't see a problem in the process now, but I'm open to making changes that are in my purview if someone convinces me there *is* a problem or has ideas on how to improve it.