[Python-ideas] Re: PEP's shouldn't require a sponsor

[Paul Moore]

On Sat, 27 Jul 2019 at 03:51, Kyle Stanley  wrote:
>
> Eric V. Smith wrote:
> > In addition, I find it hard to believe someone couldn't find a sponsor
> > for a well-written PEP. I'm happy to sponsor such a PEP, even if I think
> > it will be rejected. Rejected PEPs serve a useful purpose, too, if only
> > to point to when the same issue comes up in the future.
>
> Do most of the other core developers also share this perspective? Even
> though PEPs were not intended to be intimidating, they definitely can be
> for those who are less familiar with the process. I can imagine that many
> people would think that a "sponsor" would mean fully convincing someone
> to be completely on board with their idea.

Personally, my position is a bit more nuanced. I'm happy enough to
sponsor a PEP, but most people's "PEP ideas" are actually still
insufficiently well thought out to be worth a PEP, and typically my
first comment as a sponsor would be "you don't need a sponsor yet, you
need to refine your proposal a lot first". However, from the way a lot
of threads on python-ideas go, it's clear that a lot of people aren't
really aware of how much work is needed to get a proposal to the point
where it's ready for a PEP (and when they get feedback to that effect
from the list, they get frustrated at the "negative feedback").

So while I'm happy enough to sponsor a (well-written) PEP, I'm not
anywhere near as willing to mentor someone in how to develop a
proposal to the point where it's ready for submission as a PEP
(because I simply don't have the time). And people tend not to
appreciate the difference between those two tasks.

Paul
___
Python-ideas mailing list -- python-ideas@python.org
To unsubscribe send an email to python-ideas-le...@python.org
https://mail.python.org/mailman3/lists/python-ideas.python.org/
Message archived at 
https://mail.python.org/archives/list/python-ideas@python.org/message/AYMID2BLLKOUXQA765RDBHPCIJHMSMK6/
Code of Conduct: http://python.org/psf/codeofconduct/

[Python-ideas] Re: PEP's shouldn't require a sponsor

[Chris Angelico]

On Sat, Jul 27, 2019 at 12:50 PM Kyle Stanley  wrote:
>
> Eric V. Smith wrote:
> > In addition, I find it hard to believe someone couldn't find a sponsor
> > for a well-written PEP. I'm happy to sponsor such a PEP, even if I think
> > it will be rejected. Rejected PEPs serve a useful purpose, too, if only
> > to point to when the same issue comes up in the future.
>
> Do most of the other core developers also share this perspective? Even
> though PEPs were not intended to be intimidating, they definitely can be
> for those who are less familiar with the process. I can imagine that many
> people would think that a "sponsor" would mean fully convincing someone
> to be completely on board with their idea.
>
> As someone who only more recently began contributing to Python, my previous 
> perception of PEPs were these monolithic technical
> documents that were well approved by the entire community. I'm slowly
> starting to see them more as simply being well structured proposals after
> having seen more of them.

Not a core dev, but from my perspective, PEPs are far too "iconic".
People make their first posts to python-ideas under the impression
that they should be writing PEPs. No, that's not the case; start with
discussion (which doesn't require a sponsor), and *then* start talking
about a PEP. By the time you get that far along with a proposal,
either your idea has enough support for a core dev to say "yeah, I'll
sponsor that" (even if s/he doesn't actually agree with the proposal),
or you know you're asking for something controversial (in which case
your first hurdle is to convince a core dev).

ChrisA
___
Python-ideas mailing list -- python-ideas@python.org
To unsubscribe send an email to python-ideas-le...@python.org
https://mail.python.org/mailman3/lists/python-ideas.python.org/
Message archived at 
https://mail.python.org/archives/list/python-ideas@python.org/message/6ED2GZP6PFP3REWMDIFYE7X6OS4OMVI5/
Code of Conduct: http://python.org/psf/codeofconduct/

[Python-ideas] Re: PEP's shouldn't require a sponsor

[Kyle Stanley]

Eric V. Smith wrote:
> In addition, I find it hard to believe someone couldn't find a sponsor 
> for a well-written PEP. I'm happy to sponsor such a PEP, even if I think 
> it will be rejected. Rejected PEPs serve a useful purpose, too, if only 
> to point to when the same issue comes up in the future.

Do most of the other core developers also share this perspective? Even 
though PEPs were not intended to be intimidating, they definitely can be 
for those who are less familiar with the process. I can imagine that many
people would think that a "sponsor" would mean fully convincing someone
to be completely on board with their idea.

As someone who only more recently began contributing to Python, my previous 
perception of PEPs were these monolithic technical 
documents that were well approved by the entire community. I'm slowly
starting to see them more as simply being well structured proposals after
having seen more of them.

To many outside of the development community though, such as those proposing 
ideas, their impression of a PEP is probably based on the
massive ones such as PEP 8. Although it was purely comical, I think PEP 401
helped me quite a lot to see them as less intimidating. PEP 581 is a good
example of an actual approved one that's easily digestible.
___
Python-ideas mailing list -- python-ideas@python.org
To unsubscribe send an email to python-ideas-le...@python.org
https://mail.python.org/mailman3/lists/python-ideas.python.org/
Message archived at 
https://mail.python.org/archives/list/python-ideas@python.org/message/4WVJMMCFKQOUBSO452K45KIARIJ2Y6KB/
Code of Conduct: http://python.org/psf/codeofconduct/

[Python-ideas] Re: PEP's shouldn't require a sponsor

[Eric V. Smith]

On 7/26/2019 11:21 AM, Geoffrey Spear wrote:


On Thu, Jul 25, 2019 at 8:17 AM Batuhan Taskaya > wrote:


Why do i need to convince a core developer for my PEP? AFAIK the
steering council can include non core developers (i know it isn't
that current case but for the future this is important). And if the
last authority who will approve my PEP is the steering council i
just need to convince them not core developers. 



To convince a majority of the steering council and 0 core developers to 
make a change, you'd need a hypothetical future steering council with a 
non-core-developer majority.


Even if we stipulate that this would ever happen (which seems 
exceedingly unlikely), presumably that majority would change the policy 
to allow themselves to sponsor PEPs.


(In my opinion if there's even a single steering council member who 
isn't a core developer, they'd probably be able to convince the rest of 
the council to add a special exception allowing them to sponsor PEPs; I 
find it hard to imagine someone being trusted enough to sit on the 
council without the rest of the council thinking they can be trusted to 
sponsor a PEP...)


I agree with all of these points.

In addition, I find it hard to believe someone couldn't find a sponsor 
for a well-written PEP. I'm happy to sponsor such a PEP, even if I think 
it will be rejected. Rejected PEPs serve a useful purpose, too, if only 
to point to when the same issue comes up in the future.


Just be aware that writing a PEP so that it gets to the accept/reject 
stage can take months of work. At least it has in my case.


What you won't find is someone to sponsor a PEP that says something like 
"remove the GIL", without any details of how to do it. And it's good 
that no one would sponsor that: it's just noise! This is what the "you 
must have a sponsor" rule is trying to prevent. It's not trying to 
prevent well thought out ideas that might not get accepted.


Eric
___
Python-ideas mailing list -- python-ideas@python.org
To unsubscribe send an email to python-ideas-le...@python.org
https://mail.python.org/mailman3/lists/python-ideas.python.org/
Message archived at 
https://mail.python.org/archives/list/python-ideas@python.org/message/OWZSXAD72VYBNVZ63UWGLNGGYBE6VEW4/
Code of Conduct: http://python.org/psf/codeofconduct/

[Python-ideas] Re: PEP's shouldn't require a sponsor

[Geoffrey Spear]

On Thu, Jul 25, 2019 at 8:17 AM Batuhan Taskaya 
wrote:

> Why do i need to convince a core developer for my PEP? AFAIK the steering
> council can include non core developers (i know it isn't that current case
> but for the future this is important). And if the last authority who will
> approve my PEP is the steering council i just need to convince them not
> core developers.
>

To convince a majority of the steering council and 0 core developers to
make a change, you'd need a hypothetical future steering council with a
non-core-developer majority.

Even if we stipulate that this would ever happen (which seems exceedingly
unlikely), presumably that majority would change the policy to allow
themselves to sponsor PEPs.

(In my opinion if there's even a single steering council member who isn't a
core developer, they'd probably be able to convince the rest of the council
to add a special exception allowing them to sponsor PEPs; I find it hard to
imagine someone being trusted enough to sit on the council without the rest
of the council thinking they can be trusted to sponsor a PEP...)
___
Python-ideas mailing list -- python-ideas@python.org
To unsubscribe send an email to python-ideas-le...@python.org
https://mail.python.org/mailman3/lists/python-ideas.python.org/
Message archived at 
https://mail.python.org/archives/list/python-ideas@python.org/message/2V7IIPMNQCKJFXMKZXZQK6WJ4URKDHR4/
Code of Conduct: http://python.org/psf/codeofconduct/

[Python-ideas] Re: PEP's shouldn't require a sponsor

[Serhiy Storchaka]

25.07.19 15:14, Batuhan Taskaya пише:
Why do i need to convince a core developer for my PEP? AFAIK the 
steering council can include non core developers (i know it isn't that 
current case but for the future this is important). And if the last 
authority who will approve my PEP is the steering council i just need to 
convince them not core developers.


Sponsors can stay (and they should because guidance is important) but 
thy shouldn't be mandatory.  Let everyone to send their peps.


Only core developers have permissions to merge in the Python 
repositories. So you need at least one core developer on your side to 
make an official PEP.


I hope you do not want to allow everybody to merge in the Python 
repositories.

___
Python-ideas mailing list -- python-ideas@python.org
To unsubscribe send an email to python-ideas-le...@python.org
https://mail.python.org/mailman3/lists/python-ideas.python.org/
Message archived at 
https://mail.python.org/archives/list/python-ideas@python.org/message/4G2C5G355VQEDEEKB7PFX4ZF2XORHPAC/
Code of Conduct: http://python.org/psf/codeofconduct/

[Python-ideas] Re: PEP's shouldn't require a sponsor

[Brett Cannon]

On Thu, Jul 25, 2019 at 5:21 AM Rhodri James  wrote:

> On 25/07/2019 13:14, Batuhan Taskaya wrote:
> > Why do i need to convince a core developer for my PEP? AFAIK the steering
> > council can include non core developers (i know it isn't that current
> case
> > but for the future this is important). And if the last authority who will
> > approve my PEP is the steering council i just need to convince them not
> > core developers.
> >
> > Sponsors can stay (and they should because guidance is important) but thy
> > shouldn't be mandatory.  Let everyone to send their peps.
>
> Consider it this way; if you can't convince a single core developer to
> back your idea, your chances of getting general support, never mind the
> steering council, are very limited.
>

I'll make it even more concrete: if a single core dev won't sign off on a
PEP idea then the steering council won't consider it. It should be
understood that I came up with the sponsorship idea as a PEP author, core
dev, and steering council member (which the steering council signed off on):

- As a PEP editor because there's only so many of us and so I want PEPs to
be as "done" as they can be by the time they come to the peps repo so
there's less copy-editing
- As a core dev because I don't want a discussion on python-dev to start
until a PEP is in great shape and there won't be rehashing of previous
ideas or any critical points missing
- As a member of the steering council because if not a single core dev
likes an idea then I'm not interested in considering a PEP because the
steering council may help break stalemates but we are not about to against
the entire Python core team (which is what we would be asked to do if not a
single core dev wanted to back a PEP)

In other words the hurdle of finding a sponsor is very deliberate.
___
Python-ideas mailing list -- python-ideas@python.org
To unsubscribe send an email to python-ideas-le...@python.org
https://mail.python.org/mailman3/lists/python-ideas.python.org/
Message archived at 
https://mail.python.org/archives/list/python-ideas@python.org/message/HPWNGP4MCHBJ3VXQO4ADXWUUGIRWB7TI/
Code of Conduct: http://python.org/psf/codeofconduct/

[Python-ideas] Re: PEP's shouldn't require a sponsor

[Andre Roberge]

On Thu, Jul 25, 2019 at 9:32 AM Batuhan Taskaya 
wrote:

> What i see is when you post the ideas channel and it is something that
> doesnt change much on the frontside people dont care. And when people dont
> care, they forgot. PEP reviewing process is way better than posting to
> ideas and try to convince people.
>
>>
>>
You are proposing a change to the content of PEP 1 (
https://www.python.org/dev/peps/pep-0001/). If you look at the history of
that PEP, you will see that its content has changed over time. It used to
be that non-core developers could write PEPs that were accepted: I know
this very well since I've never been a core developer and have written a
PEP that was accepted (https://www.python.org/dev/peps/pep-3111/) some
thirteen years ago. This was then; now, I personally do see the need to
have PEPs sponsored by core developers - you obviously have a different
opinion.

At the very least, you should go over the history of PEP 1, and find out
the discussions (on python-ideas and python-dev, as well as possibly other
sources) that led to the change which you wish to revert. You might want to
provide a rationale as to what such a reversion is warranted, including a
summary of the arguments that led to the current situation (sponsorship
required) and how the entire process would be improved by not requiring
this sponsorship.

Following the practice for discussions that lead to PEPs, you should also
include the arguments offered by others on this list who do not appear to
support such a change and explain why and how the alternative you propose
is better.

I assume that your underlying motivation is that you have in mind some
changes to the Python language or its libraries, and that you would wish to
submit a PEP without requiring a sponsor. Since core developers are
responsible for implementing and supporting such changes, your suggestion
to change PEP 1 should include an explanation as to how core developers
would benefit from the change you propose and how allowing anyone, no
matter how familiar they are with the Python code and standard library, to
be able to submit PEP that would require the core developers to spend
valuable time considering them seriously without the ideas included in the
PEP having been discussed more casually on some public forums such as this
one.

André Roberge
___
Python-ideas mailing list -- python-ideas@python.org
To unsubscribe send an email to python-ideas-le...@python.org
https://mail.python.org/mailman3/lists/python-ideas.python.org/
Message archived at 
https://mail.python.org/archives/list/python-ideas@python.org/message/HOSFLAOUA4PRO6YC7ISXJKTN5Y2C7LZF/
Code of Conduct: http://python.org/psf/codeofconduct/

[Python-ideas] Re: PEP's shouldn't require a sponsor

[Ethan Furman]

On 07/25/2019 05:31 AM, Batuhan Taskaya wrote:


What i see is when you post the ideas channel and it is something
 that doesnt change much on the frontside people dont care. And
 when people dont care, they forgot. PEP reviewing process is way
 better than posting to ideas and try to convince people.


If someone posts a PEP to python-dev as the first step, they will be told to 
take it to ideas for discussion.

If someone wants to post a PEP-like document to python-ideas they are welcome 
to do so -- possibly the only thing lacking to make it an official PEP is the 
PEP number, but that can be added later once you've got the support of a 
core-dev.

However, writing a PEP-like document is a serious investment of energy and 
time, and most ideas can be discarded long before with minimal discussion.

--
~Ethan~
___
Python-ideas mailing list -- python-ideas@python.org
To unsubscribe send an email to python-ideas-le...@python.org
https://mail.python.org/mailman3/lists/python-ideas.python.org/
Message archived at 
https://mail.python.org/archives/list/python-ideas@python.org/message/RGYVJ33BNAN3VH4Q4ML72XTGMNTKMUS2/
Code of Conduct: http://python.org/psf/codeofconduct/

[Python-ideas] Re: PEP's shouldn't require a sponsor

[Chris Angelico]

On Thu, Jul 25, 2019 at 10:34 PM Batuhan Taskaya  wrote:
>
> What i see is when you post the ideas channel and it is something that doesnt 
> change much on the frontside people dont care. And when people dont care, 
> they forgot. PEP reviewing process is way better than posting to ideas and 
> try to convince people.
>

Then convince a core dev that your proposal is important enough to
write a PEP about. If you can't do that, odds are your proposal
wouldn't be accepted anyway - like Rhodri said.

ChrisA
___
Python-ideas mailing list -- python-ideas@python.org
To unsubscribe send an email to python-ideas-le...@python.org
https://mail.python.org/mailman3/lists/python-ideas.python.org/
Message archived at 
https://mail.python.org/archives/list/python-ideas@python.org/message/CI2I425VGYYJX4IMCI5KULHN7MIL6XDX/
Code of Conduct: http://python.org/psf/codeofconduct/

[Python-ideas] Re: PEP's shouldn't require a sponsor

[Chris Angelico]

On Thu, Jul 25, 2019 at 10:16 PM Batuhan Taskaya  wrote:
>
> Why do i need to convince a core developer for my PEP? AFAIK the steering 
> council can include non core developers (i know it isn't that current case 
> but for the future this is important). And if the last authority who will 
> approve my PEP is the steering council i just need to convince them not core 
> developers.
>
> Sponsors can stay (and they should because guidance is important) but thy 
> shouldn't be mandatory.  Let everyone to send their peps.
>

A PEP shouldn't be the *first* step in a proposal - it's a document
designed to gather together a lot of details, possibly arguments and
counter-arguments, and the justifications for decisions. Instead,
start by just writing up your idea as a python-ideas post; anyone can
do that, and then if you get enough support on -ideas, you can go
forward with writing a PEP, if it's even necessary.

ChrisA
___
Python-ideas mailing list -- python-ideas@python.org
To unsubscribe send an email to python-ideas-le...@python.org
https://mail.python.org/mailman3/lists/python-ideas.python.org/
Message archived at 
https://mail.python.org/archives/list/python-ideas@python.org/message/MTGC2W5XE45J5TJUV4XXYFT5UTP3MBPT/
Code of Conduct: http://python.org/psf/codeofconduct/

[Python-ideas] Re: PEP's shouldn't require a sponsor

[Batuhan Taskaya]

What i see is when you post the ideas channel and it is something that
doesnt change much on the frontside people dont care. And when people dont
care, they forgot. PEP reviewing process is way better than posting to
ideas and try to convince people.

On Thu, Jul 25, 2019 at 3:21 PM Rhodri James  wrote:

> On 25/07/2019 13:14, Batuhan Taskaya wrote:
> > Why do i need to convince a core developer for my PEP? AFAIK the steering
> > council can include non core developers (i know it isn't that current
> case
> > but for the future this is important). And if the last authority who will
> > approve my PEP is the steering council i just need to convince them not
> > core developers.
> >
> > Sponsors can stay (and they should because guidance is important) but thy
> > shouldn't be mandatory.  Let everyone to send their peps.
>
> Consider it this way; if you can't convince a single core developer to
> back your idea, your chances of getting general support, never mind the
> steering council, are very limited.
>
> --
> Rhodri James *-* Kynesim Ltd
> ___
> Python-ideas mailing list -- python-ideas@python.org
> To unsubscribe send an email to python-ideas-le...@python.org
> https://mail.python.org/mailman3/lists/python-ideas.python.org/
> Message archived at
> https://mail.python.org/archives/list/python-ideas@python.org/message/USFGXFOSUE6BDAVZNLO7BBSC25OY7CQD/
> Code of Conduct: http://python.org/psf/codeofconduct/
>
___
Python-ideas mailing list -- python-ideas@python.org
To unsubscribe send an email to python-ideas-le...@python.org
https://mail.python.org/mailman3/lists/python-ideas.python.org/
Message archived at 
https://mail.python.org/archives/list/python-ideas@python.org/message/HQ67MYCIU23F5RLKTUH6WWEKRC2CB6BO/
Code of Conduct: http://python.org/psf/codeofconduct/

[Python-ideas] Re: PEP's shouldn't require a sponsor

[Rhodri James]

On 25/07/2019 13:14, Batuhan Taskaya wrote:

Why do i need to convince a core developer for my PEP? AFAIK the steering
council can include non core developers (i know it isn't that current case
but for the future this is important). And if the last authority who will
approve my PEP is the steering council i just need to convince them not
core developers.

Sponsors can stay (and they should because guidance is important) but thy
shouldn't be mandatory.  Let everyone to send their peps.


Consider it this way; if you can't convince a single core developer to 
back your idea, your chances of getting general support, never mind the 
steering council, are very limited.


--
Rhodri James *-* Kynesim Ltd
___
Python-ideas mailing list -- python-ideas@python.org
To unsubscribe send an email to python-ideas-le...@python.org
https://mail.python.org/mailman3/lists/python-ideas.python.org/
Message archived at 
https://mail.python.org/archives/list/python-ideas@python.org/message/USFGXFOSUE6BDAVZNLO7BBSC25OY7CQD/
Code of Conduct: http://python.org/psf/codeofconduct/