I propose tags in “Help & Discussions” section.
I think a pinned post in ideas could be useful. Listing you should
look on the mailing list and Discourse. Try to look for GitHub
usage, check if this isn’t just an alternative way to write it. If
you can’t put it on PyPi first. Read similar discussions. List
benefits, give an example.
Up-thread I proposed some new text for the pinned post, though that
didn’t seem to generate much feedback.
1 Like
Yeah, I think there shouldn’t be objections to pinning a post with all relevant information.
I would like to go through that when I propose a new idea in 2025.
1 Like
I agree with this. I think ALL ideas should go in the ideas section even if they’re not fully up to standards. To solve the “too many half-baked ideas wasting people’s time” problem I would be in favor of
- Having clearly stated criteria against which ideas/proposals can be prepared. This means if an idea is half-baked it only takes some willing person a short amount of time to say “this idea is not fully baked, please see [pinned post with criteria]. Specifically look at points 1, 4, 7…”
- There can be either a positive tag indicating a topic has reached the criteria or a negative tag indicating a topic has not reached the criteria, e.g. “needs work”.
I think a positive tag would be less work for since only some ideas would need a tag. But I think it’s a bit easier mentally to identify when a topic “needs work” than when it’s actually a fully fledged proposal.
Though I admit I don’t know how tags actually show up on the site.
2 Likes
Not to the people posting help requests.
And the problem we’re trying to address here is that Ideas is also becoming spammy and high-traffic, as a result of the level of low quality proposals and off-topic discussions. As a case in point, this thread isn’t appropriate for the Ideas category - it’s not a proposal for improving the language or stdlib…
Should it be moved to Discourse Feedback?
1 Like
I was in no way implying that the Help section should be policed or restrained.
My point is that
- for each external tool there can be a very large number of questions over how to use it, but a much smaller number of discussions (about one) over whether to include it in the stdlib, and i could make the same point about usage of a standard feature versus changes to that feature ;
- there are always, in my experience, people creating duplicate topics for help requests (in part due to often not finding the right keywords), or issues whose slight dissimilarity deserves a specific thread, while it’s much less the case in the Ideas section.
So if you mix up ideas and help, idea threads get drowned in the mass.
Yes, probably, if the mods agree.
This is exactly what’s been happening to Ideas now, but there’s disagreement about what counts as what.
I’m not aware of help requests being posted in the Ideas section, much less drowning the threads about ideas - raising to the level of a proposition or not.
In addition, a negative tag would have to be added and then potentially removed when the idea reaches maturity, whereas a positive tag is forever.
And according to people who think the majority of the threads in Ideas are bad, a positive tag would be applied less often and require less work than a negative tag.
Yes, there are two problems here:
- Those who are interested in only well-researched proposals to make changes to Python are ignoring all idea posts in the Ideas category because Ideas has become spammy to them, but,
- If we lump less-researched ideas into the Help category, then those who are interested in any ideas to make changes to Python would start ignoring idea posts in the Help cateogry as a result because Help is too spammy to them.
This again is why I think we need a new category, either:
- A “What ifs” category slotted between Help and Ideas, where people ask questions starting with “What if Python…” like “What if Python supported Goroutines?”, which is distinctly an idea to make a change to Python but without yet a concrete semantic or a path to go forward, or,
- A “Proposal” category above Ideas, where people submit well-researched ideas that are able to answer a certain predefined set of questions.
This way, categories are both clearly defined and able to be subscribed to by people of different interests.
7 Likes
I’m not aware of help requests being posted in the Ideas section,
much less drowning the threads about ideas - raising to the level
of a proposition or not.
Perhaps a way of reframing it is that a lot of newcomers to the
community interpret “ideas” to include “feature requests” for things
where they haven’t fully researched the potential implications,
haven’t entirely thought through what the implementation might look
like, or aren’t themselves willing to put in most of the work to
bring about. That is a help request either in the sense that
they’re asking someone (CPython maintainers) to do work for them
(implement their “idea”), or are in need of help in order to
understand why their “idea” doesn’t make sense.
Instead of immediately posting these notions in the ideas
category, it would be great to first post something in the help
(users) category along the lines of “please help me understand why
CPython doesn’t do/have this…” Unless you already know the answer
to that question, what you have isn’t yet something that should be
posted to the ideas category. Don’t assume you’ve just had a novel
thought that’s never crossed anyone’s mind until now, as it’s far
more likely you’re not the first person to think of it and there’s a
very good reason why it’s not already there, you’re simply missing
the required context to know what that reason is just yet.
2 Likes
‘What ifs’ are fine for research, and research is fine to do in Python Help . Help is for general discussions.
Many ‘What ifs’ would just be a result of the XY Problem, the Dunning-Kruger effect, and other similar situations. So, it will just be another Help category.
Yes, because it is.
Y’all may disagree with that and want to change that, but that’s what the Ideas section is, I’ll quote the description once again : “Would you like to change something in Python? This might be your feedback forum.”
If the “idea” applies to Python itself, implementing it is subordinate to the relevance and merits of the idea. That would qualify as help, most likely, but it happens after the discussion in Ideas.
That’s not help. Or the CoreDevelopment section is also a help section, and the PEP section is too, and all threads about Ideas are as well. Such a broad definition doesn’t make sense. A Help or support section is about help for using Python, otherwise every single need for communication can be broadly considered as help.
And that’s why finding the relevant thread, among all the ideas for changes to Python, is something we need to optimize, by placing them all - and only them - in a single searchable namespace 
2 Likes
The description did not come down from the heavens to define what the section is. It is a blurb someone wrote and it’s clear that it doesn’t fully specify the purpose of the category[1]. It should be clarified.
the word “might” is doing a lot of work there ↩︎
1 Like
Absolutely, and even if it were, we would have all right to change it.
But the fact is that it has been there for a long time, and that the Ideas section has been used as such - and is used as such.
Y’all think that encompasses too many things, so be it, but what I’m saying is that this distinction, this particular boundary, is a very important feature when it comes to searching past topics. Hence why all that currently fits into Ideas should stay separated from the rest, even if we choose to split all that in two.
On a sidenote, there already is a process by which modifications to Python go through : the PEPs. And all modifications to Python don’t necessarily go through PEPs.
What you’re all putting in place will turn a 1- to 2-step process into a 2- to 3-step process. It will require more talk and administrative actions to apply this longer process. And it may also have the consequence of barring or discouraging people from participating in the evolution of Python, a problem which, in my opinion and not pointing fingers at anyone, is already not that great.
1 Like
No, it isn’t. The Ideas section started life as the python-ideas mailing list and its remit goes back to that list.
So the description is wrong. We can change that. The only thing preventing us just doing so is the fact that we’re not 100% sure how to make the message clearer while still being polite (“No stupid ideas please” isn’t acceptable, for example 
), and people are arguing over where we should direct the people we’re trying to get to not post to Ideas.
Python is a volunteer organisation. There’s an implied commitment along with any suggestion to change the language or stdlib, that the proposer is willing to do at least some of the work themselves. That may not be implementation, but it definitely includes research, managing discussion, etc.
And yes, again that “implied commitment” might be too implicit for the current community of people posting in Ideas. Python’s user base has changed a lot, and new users are much more likely to be either programmers using Python in a proprietary environment, with no experience of the open source community, or people forced to use Python who may not want to, and who wish Python was more like their language of choice. Again, maybe we need to spell out our expectations more clearly.
But regardless, those are the scope of the Ideas group and our expectations of people participating in it. Changing them amounts to replacing the Ideas group with something with a different role. Which may be fine - but it will no longer be a useful step on the way to getting an idea implemented in Python, as the few remaining core devs and “old hands” will depart and there will be no guidance as to how to progress an idea.
Given the existence of the historical mailing list archives and the PEP repository, that’s not actually possible. And I’m not sure it’d help anyway. Most of the problem is caused by people posting unresearched ideas. Making it easier to research won’t help people who don’t even realise (or care, in some cases) that doing research up front is an expectation in the first place.
3 Likes
Can Discourse have categories so that only a select group of people can move topics there? Then one could have a special section for “sponsored ideas”, separate from the free-for-all ideas section we currently have.
The description did not come down from the heavens to define what
the section is. It is a blurb someone wrote and it’s clear that it
doesn’t fully specify the purpose of the category[1]. It should be clarified.
Also the “you” because many are missing that it’s asking “are you
willing to do what’s necessary to make this happen?” rather than
“would you like someone to do this for you?”
You don’t change things by making suggestions, you change things
by presenting the work you’re planning to do so long as there’s
consensus from the maintainers that they’re likely to approve your
plan.
the word
“might” is doing a lot of work there ↩︎
1 Like
We already have that process, though. Propose an idea, get a core dev sponsor, that core dev will help you prepare a PEP that can be posted to the PEP discussions category.
What you are suggesting is in effect adding a second “get a sponsor, move on a stage” step, for people who aren’t motivated enough to do the work to make progress on the existing route.
1 Like
I understood the problem to be that the current “ideas” category has become low S/N and is therefore not very accessible. What I imagined is that everything remains pretty much exactly as it is now, except there is a vetted subcategory, say, where actually viable ideas can be highlighted. Of course, that only works if there are sufficiently many people in the privileged group still sifting through the usual “ideas” category.
Ideally, this could be done with a tag, but I don’t think Discourse has restricted tags.
1 Like