Poll: A PSF grant for running a “Core Dev Mentorship Program”


(Tal Einat) #1

EDIT: Please use the newest poll below in post #7!

Please take a moment to vote your opinion on my suggestion for a PSF-funded “Core Dev Mentorship Program”, recently brought up here on Discourse and on the mailing list. This poll will inform the PSF’s decision whether to make such a grant.

  • -1
  • 0
  • 1

0 voters

Rationale: The PSF asked to know what the core dev community thinks of this proposal. The discussions so far have seen few participants and fewer clear opinions for or against it. I hope that this poll will provide a general overview of the opinions about it.

FWIW, I’ll consider even just a few core devs being against it as a sign that it shouldn’t be done.


(Antoine Pitrou) #2

Since you chose to make the votes private, I will state for the record that I voted 0.


(Tal Einat) #3

Bah! I didn’t mean to do that. I now see that the poll is very problematic…


(Tal Einat) #4

The original poll accidentally had the votes hidden (it’s the default!) and only shows the average (!). Here’s a new poll instead. Sorry!

EDIT: Sorry, this poll is also no good; use the next one down!

  • -1
  • 0
  • 1

0 voters


#5

I voted earlier but preferred the vote to be private, so I’m not voting in the new poll.


(Tal Einat) #6

Ack! Now the poll is just showing who voted, and the average.


(Tal Einat) #7

Okay, new poll, hopefully as close as possible to our old mailing-list style voting:

  • +1
  • +0.5
  • +0
  • -0
  • -0.5
  • -1

0 voters

(Really sorry for this mess!)


(Victor Stinner) #8

After the discussion, now I’m no more sure if it’s better to use a grant for mentoring or just to let core devs do whatever their want (but provide a report of their work).


(Steve Dower) #9

Fairly sure those aren’t mutually exclusive, and if a core dev can do more mentoring with a grant than without one, why not? It doesn’t have to stop anyone else from mentoring too.


(Petr Viktorin) #10

Ah! This comment made me realize what to say here – something I’ve been thinking about unsuccessfully for the past few days :slight_smile:

This vote is not like those on the mailing list. I need to register two votes, each with a different hat on.

As just a core dev, I say go ahead! Since there’s PSF money involved, be careful about accounting and reporting. Run it as an experiment: what’s the hypothesis (what exactly do we expect to happen when we put money on mentoring)? How well will that hold up when it’s actually tried? How can the next run be improved?
I trust you’ll manage to do that well. That would make my vote +0.5.

As an employee of Red Hat, a PSF sponsor, I’m quite uneasy. In my job I’m asked to prioritize customer issues, but those come in waves, so most of the time the team has significant “spare” capacity to work upstream, on things like core development and mentoring. Since we do that, Red Hat does not generally donate to non-profits that pay people for development tasks (which reviewing patches certainly is). If PSF would be to do this regularly, Red Hat would probably need to revisit the sponsorship.
That would make the vote -1 (but maybe it would be OK to try once, then find alternate ways to fund if it works).

I don’t want either of these votes attached to just my name. One is from Petr the individual, ignoring his employer’s interests. The other from Petr the Red Hatter.


(Tal Einat) #11

If that’s the issue then I can remove the reviewing of patches and PRs from the proposed activities in the “mentorship program”. It would be less efficient, but I believe it would still effective.

Reviewing patches is not nearly the most important part of how I propose to do this mentoring. I’ve addressed this recently on the mailing list:

Reviewing PRs is also a way to help and train contributors. It’s not very different from mentoring, depending on your definition of mentoring :slight_smile:

It’s very different from my method of mentoring. I do many additional things, among them:

  • first and foremost, those mentored having a long-term plan and partner
  • settings expectations and providing encouragement
  • helping choose appropriate tasks
  • answering little “silly” questions that many feel uncomfortable to raise in public forums
  • explaining lots of small details about process, tools, priorities and more; for example, what kind of changes are backported and how far back
  • maintaining momentum over a significant period of time

Reviewing PRs is one significant piece of my mentoring, but IMO not the most important one.


(Petr Viktorin) #12

It’s not – reviewing is just the most clear-cut part of mentoring.
Time to do interesting development-related things like mentoring is a major benefit of working where I do. If PSF was able to pay a living wage for that, it would take away a lot of motivation for my day job (supporting 10-year-old systems). It would be great for me as an individual if I got the mentoring gig, but that wouldn’t be good for the Red Hat/PSF relationship.


(Tal Einat) #13

That’s awesome!

Can you, or someone else working with you or in a similar position, run such a mentoring program sometime soon? If so then there’s certainly no need for me to request such a grant from the PSF, and I’d be happy to help with brainstorming and organization of it in my free time.

As it is, as I mentioned in the original post, all of the currently available mentoring for CPython appears to be too little.


(Petr Viktorin) #14

I (with my individual contributor hat on) spend this time in different ways (most of which aren’t visible on CPython core dev channels). But if a part-time paid mentorship job was available from the PSF, I’d be very tempted to take that instead of my current day job.

Alas, that is the case. But I (with my Red Hat on) need to warn that PSF might not want to compete with its sponsors for human resources. It’s just a point to consider – I can’t speak for the company and I have no say in the sponsorship deal – but I think it’s important to consider it.

With either hat on: this situation is not ideal, but I can’t see a good solution.


(Steve Dower) #15

From my POV similar to @encukou (also two hats, also a sponsorship deal, also not involved in it personally), I think provided the grant is clearly for “community building” rather than “development work” there shouldn’t be an issue. The PSF’s mission is community building.

It probably also needs to be clearly stated that the PSF does not offer “part-time paid mentorship jobs,” but may provide a grant for a specified purpose with specific goals. a.k.a., if Tal screws this up then there won’t be any more grants for this purpose :slight_smile:

Of course, when companies are acquired, things change, so it’s entirely possible that donations could dry up anyway under any excuse that can be found…


(Petr Viktorin) #16

Yes, that should addresses my concerns!

It seems everyone interested now understands this potential conflict of interest.
Tal, I wish you best of luck with the grant! May your effort be fruitful.


(Marc-André Lemburg) #17

Just want to clarify this: The PSF’s mission is not only community building. Supporting development work is one of its key missions (but unfortunately, only rarely used):

I would very much like to see more PSF money flow into Python development. Note that the only reason this is not happening a lot is the lack of grant requests in this area, not that the PSF does not want to fund such activities.

I can understand Petr’s concerns related to Red Hat, but don’t follow RH’s reasoning here. In the end, sponsorship money is used for a good purpose and to support open source. It should really be up to the open source project on how to achieve that. Clearly, Red Hat will directly benefit from such mentorship as well, since it’ll increase the number of core developers and thus reduce the efforts Red Hat has to put into supporting Python for its customers. That’s a standard win-win situation.


(Petr Viktorin) #18

Supporting open source is exactly what RH wants! What RH probably does not want is the PSF becoming a middleman. RH would rather hire another core dev than pay the PSF to pay a core dev.
RH tries very hard to make its business interests align with the interests of open-source projects – but in the end it’s the business end that decides where money goes.

FWIW, I don’t claim to represent Red Hat in this discussion – I’m commenting on its motivations as I see them. Also, note that I am officially free to act in the interests of the Python project, even if it’s against the interests of my employer (see the last paragraph of Conflict of Interest in RH’s code of conduct).


(Antoine Pitrou) #19

As a counter example, RH is a sponsor of the Linux Foundation, which employs Linus Torvalds and Greg Kroah-Hartman.

Regardless, it’s all up to the PSF, but I think it would be silly to refuse funding a developer just because it might annoy one of the sponsors. Hopefully the PSF has enough sponsors that it doesn’t become shackled to the wills of one particular corporate backer…


(Petr Viktorin) #20

Indeed. Unless there’s further misunderstanding of the potential conflict of interest, let’s drop the issue.