View mode: basic / threaded / horizontal-split · Log in · Help
April 08, 2013
DIP attack
There has been quite an explosion of DIP count recently. With 
just 3 states (Approved/Superseded/Draft) it becomes quite hard 
to have a quick overview of what actually happens on topic.

Shouldn't more precise categorization be added? I can imagine 
(Draft/Approved - Needs Implementation/Implemented/Rejected/Needs 
Improvements) as minimal useful state set. Some of DIPs rot there 
for eternity with an uncertain future.
April 08, 2013
Re: DIP attack
On 2013-04-08 10:34, Dicebot wrote:
> There has been quite an explosion of DIP count recently. With just 3
> states (Approved/Superseded/Draft) it becomes quite hard to have a quick
> overview of what actually happens on topic.
>
> Shouldn't more precise categorization be added? I can imagine
> (Draft/Approved - Needs Implementation/Implemented/Rejected/Needs
> Improvements) as minimal useful state set. Some of DIPs rot there for
> eternity with an uncertain future.

I think we need something similar to the review queue. Also, what's the 
point of DIP's if none of the core developers are looking at them. I'm 
pretty sure that I've seen several DIP's that have no comments from 
Walter or Andrei.

-- 
/Jacob Carlborg
April 08, 2013
Re: DIP attack
On Monday, 8 April 2013 at 11:16:05 UTC, Jacob Carlborg wrote:
> On 2013-04-08 10:34, Dicebot wrote:
>> There has been quite an explosion of DIP count recently. With 
>> just 3
>> states (Approved/Superseded/Draft) it becomes quite hard to 
>> have a quick
>> overview of what actually happens on topic.
>>
>> Shouldn't more precise categorization be added? I can imagine
>> (Draft/Approved - Needs 
>> Implementation/Implemented/Rejected/Needs
>> Improvements) as minimal useful state set. Some of DIPs rot 
>> there for
>> eternity with an uncertain future.
>
> I think we need something similar to the review queue. Also, 
> what's the point of DIP's if none of the core developers are 
> looking at them. I'm pretty sure that I've seen several DIP's 
> that have no comments from Walter or Andrei.

I am pretty sure about half of ones that have been added recently 
have no comments from Walter/Andrei. That worries me too, as it 
minimizes difference between DIP and random newsgroup thread.

Full review analog may be an overkill, but at least clear 
indication of DIP state is a must, in my opinion.
April 08, 2013
Re: DIP attack
On 4/8/13 4:34 AM, Dicebot wrote:
> There has been quite an explosion of DIP count recently. With just 3
> states (Approved/Superseded/Draft) it becomes quite hard to have a quick
> overview of what actually happens on topic.
>
> Shouldn't more precise categorization be added? I can imagine
> (Draft/Approved - Needs Implementation/Implemented/Rejected/Needs
> Improvements) as minimal useful state set. Some of DIPs rot there for
> eternity with an uncertain future.

A good source of inspiration: http://docs.scala-lang.org/sips/

Andrei
April 08, 2013
Re: DIP attack
On 4/8/13 7:16 AM, Jacob Carlborg wrote:
> On 2013-04-08 10:34, Dicebot wrote:
>> There has been quite an explosion of DIP count recently. With just 3
>> states (Approved/Superseded/Draft) it becomes quite hard to have a quick
>> overview of what actually happens on topic.
>>
>> Shouldn't more precise categorization be added? I can imagine
>> (Draft/Approved - Needs Implementation/Implemented/Rejected/Needs
>> Improvements) as minimal useful state set. Some of DIPs rot there for
>> eternity with an uncertain future.
>
> I think we need something similar to the review queue. Also, what's the
> point of DIP's if none of the core developers are looking at them. I'm
> pretty sure that I've seen several DIP's that have no comments from
> Walter or Andrei.

DIPs' semi-formal manner helps a lot fostering community discussion well 
before they're up for approval.

Andrei
April 08, 2013
Re: DIP attack
On 4/8/13 7:19 AM, Dicebot wrote:
> On Monday, 8 April 2013 at 11:16:05 UTC, Jacob Carlborg wrote:
>> On 2013-04-08 10:34, Dicebot wrote:
>>> There has been quite an explosion of DIP count recently. With just 3
>>> states (Approved/Superseded/Draft) it becomes quite hard to have a quick
>>> overview of what actually happens on topic.
>>>
>>> Shouldn't more precise categorization be added? I can imagine
>>> (Draft/Approved - Needs Implementation/Implemented/Rejected/Needs
>>> Improvements) as minimal useful state set. Some of DIPs rot there for
>>> eternity with an uncertain future.
>>
>> I think we need something similar to the review queue. Also, what's
>> the point of DIP's if none of the core developers are looking at them.
>> I'm pretty sure that I've seen several DIP's that have no comments
>> from Walter or Andrei.
>
> I am pretty sure about half of ones that have been added recently have
> no comments from Walter/Andrei. That worries me too, as it minimizes
> difference between DIP and random newsgroup thread.
>
> Full review analog may be an overkill, but at least clear indication of
> DIP state is a must, in my opinion.

No need to worry. Authoring a DIP indicates a willingness of the author 
to formalize a request, submit it to community scrutiny, and bring it to 
good shape before it is formally reviewed.

Andrei
April 08, 2013
Re: DIP attack
On Monday, 8 April 2013 at 15:50:35 UTC, Andrei Alexandrescu 
wrote:
> No need to worry. Authoring a DIP indicates a willingness of 
> the author to formalize a request, submit it to community 
> scrutiny, and bring it to good shape before it is formally 
> reviewed.
>
> Andrei

Sure, I understand that, but, as an author of one of DIPs I have 
no idea if my DIP was even read by you or Walter. It was 
announced and all minor issues highlighted by community were 
addressed. What now? Is still it lacking something? Is it 
preapproved and waiting for a pull request? Conceptually 
rejected? I have no idea. And know even less about other DIPs.

Usual "do pull request and then we will see" does not really work 
here, as implementing new lagnauge feature is often a 
considerable amount of efforts and some for of pre-approvement 
can really help.
April 08, 2013
Re: DIP attack
On 2013-04-08 17:46, Andrei Alexandrescu wrote:

> DIPs' semi-formal manner helps a lot fostering community discussion well
> before they're up for approval.

Yes, but currently I don't think it's much difference compared to a 
newsgroup post.

-- 
/Jacob Carlborg
April 08, 2013
Re: DIP attack
On 2013-04-08 17:26, Andrei Alexandrescu wrote:

> A good source of inspiration: http://docs.scala-lang.org/sips/

I've read this:

http://docs.scala-lang.org/sips/sip-submission.html

And as far as I know our system for DIP's doesn't work anyway near 
what's described on that page. Or are you proposing it should?

-- 
/Jacob Carlborg
April 08, 2013
Re: DIP attack
On 4/8/13 3:00 PM, Jacob Carlborg wrote:
> On 2013-04-08 17:46, Andrei Alexandrescu wrote:
>
>> DIPs' semi-formal manner helps a lot fostering community discussion well
>> before they're up for approval.
>
> Yes, but currently I don't think it's much difference compared to a
> newsgroup post.

Newsgroup posts come and go. Big difference.

Andrei
« First   ‹ Prev
1 2
Top | Discussion index | About this forum | D home