Alternate the Language Summit between PyCon US and EuroPython

Core Development
1

I propose we alternate the Python Language Summit between PyCon US and EuroPython, with the 2026 summit to be held at EuroPython.

Since 2009, the summit is a one-day event that takes place at PyCon US before the main conference days. It has twice been held at EuroPython (EP) in 2010 and 2011.

Python is an international community, and not all potential attendees can travel to the US each year.

We have a core team sprint each year. 2016-2018 were in the US, after which it has alternated between the US and Europe (with the two lockdown years online).

Remembering the PSF mission is ā€œto support and facilitate the growth of a diverse and international community of Python programmersā€, weā€™ve been discussing also alternating the language summit between PyCon US and EuroPython.

Last month in Prague, I spoke with Artur Czepiel, chair of the EuroPython Society which organises EP, and they would be happy to host us. Similar to PyCon US, EP also has a couple of days of summits and tutorials before the main event.

Travel

Some of the PSF funding goes towards a CPython core development budget. A travel grant is available for the core team from this budget to help them travel to PyCon US (so it doesnā€™t come out of the regular financial aid budget).

If we alternate US/Europe, would this travel grant only be available for the conference hosting the summit? Or could a core team member choose which one?

Tickets

Free PyCon US tickets are available from the PSF/PyCon US org (and not from this budget).

Similarly, the EPS already offers free EP tickets for the core team via the Guido van Rossum Core Developer Grant.

Other ideas

  • Hold a summit at both PyCon US and EuroPython each year.
  • Hold a summit at an APAC conference.

What are your thoughts?

I think it would be a good idea. EuroPython is a wonderful conference, with excellent talks, high core team attendance, a CPython core panel, and sprints. Many PyCon US summit attendees also attend EuroPython. Ideally Iā€™d like to see a summit at both, but alternating would be a good start.

PS Artur also asked if there are other things the EPS/EP can do to help core development. Any ideas?

29 Likes
2

Just want to drop a huge +1 on this. I think this proposal is fair, more inclusive, and even more relevant given the current political climate thatā€™s making it challenging for many on our team to feel comfortable travelling to the US.

Also, FWIW, EuroPython is a fantastic conference. Well worth the trip, even without the Language Summit.

9 Likes
3

Maybe not just a good start! What prevents us from just adding a Language Summit session to EuroPython and have two LSā€™s every year[1]? Is the concern that we might not have enough topics or that topics would overlap? Having two might even have some benefits, but allowing for fewer topics but more discussion on those topics?

  1. Iā€™d say the same about the core dev sprint, but letā€™s leave it at the LS for now ā†©ļøŽ

3 Likes
4

Iā€™ve always previously had the impression that as an active committer/core dev, thereā€™s an impression that we should try and attend the LS. I think having two per year might put an (implict) expectation that people try and attend both. Assuming you go to the conference as well, thatā€™s a fortnight of holiday and several thousand pounds total expenses, which are big investments to make.

This being said, Iā€™ve never been to the Summit, so Iā€™m clearly derilict in my duties!

A

1 Like
5

Iā€™m a bit worried that this could lead to two mostly-non-overlapping groups attending the different summits, possibly with each forming their own ā€œconsensusā€ about different issues.

I for one would feel pressure to attend both.

8 Likes
6

While itā€™s perhaps encouraged for core devs to attend the LS, in the sense that this is an event primarily[1] for folks with commit bit, the reality is that thereā€™s no way even a majority of core devs can attend. I think weā€™ve been pretty careful lately not to make any binding decisions at the LS exactly because we recognize that not all core dev voices can be heard at the in-person event.

  1. but not solely ā†©ļøŽ

2 Likes
7

Maybe we need to rethink what the goal is of the LS? Itā€™s at least different now than when it was started.

2 Likes
8

One concern is the travel grants from the core dev budget. If there are two summits, I expect it wouldnā€™t stretch to both. And from the top post:

If we alternate US/Europe, would this travel grant only be available for the conference hosting the summit? Or could a core team member choose which one?

9

The language summit already seems to have moved towards being an opportunity for the broader community to tell core devs about stuff, so I donā€™t see a big issue with having two (but I already donā€™t feel incredible pressure to attend).

The core dev sprints is the important one to try and get to. That time spent in person makes working together online much easier for the rest of the year.

5 Likes
10

Iā€™m +1 on alternating, but I also asked for this to be considered before. :grin: I also think itā€™s just fair for those who canā€™t make the trip to the US for whatever reason but could make it to Europe (or APAC if that ever happens).

5 Likes
11

Some further potential advantages I see ā€œalternatingā€ would have over ā€œbothā€ would be:

  • it gives the conferences a potential opportunity to have another event that they host in their language summit ā€œoffā€ year without needing to book an extra summit room over what they currently do
  • core devs may appreciate not having to choose between the language summit and another event in the ā€œoffā€ year (since both EP and PyCon US generally host multiple summits)
  • still only need to find one set of summit coordinators each year (who only need to wrangle one set of talk proposals) rather than doubling that part of the workload
6 Likes
12

Alternating (rather than having two summits per year) would also increase the (perceived) legitimacy of the non-PyConUS Language Summit. If we keep having it at PyCon US every year itā€™s very easy to see EuroPythonā€™s Language Summit as an annex, an adjunct, an also-ran. Making the summits alternate makes it more explicit that they should be treated by everyone as equal and equivalent.

13 Likes
13

If we continue to alternate the core dev sprints as well, 2026 would have the dev sprints in North America and the language summit in Europe. That also makes it easier for folks to attend at least one core dev event each year if geographic location is an issue.

I suspect Hugo thought of that, but I didnā€™t see an explicit mention of it.

9 Likes
14

Iā€™m +1 on alternating the annual Language Summit between EP and PyCon US. I would love to see a Language Summit in Asia-Pacific someday.

I would love to see CPython mini-summits, perhaps topic focused, at conferences throughout the world.

5 Likes
15

Iā€™m up for having the LS at EP as well but Iā€™m not too opinionated if alternating or having it at both EP/PyCon US. I can see pros/cons for both solutions.

Whatever we decide, we should consider to allow remote/online participation (still under invite). This could be an additional burden and technically challenged, but it allows people who cannot travel not to miss the discussions (and provided remote attendees are OK in waking up early/going to sleep late for that day).

Usually the agenda is published before the event, they can jump in for the parts who have an interest in.

This defies a bit the principle ā€œletā€™s meet face to faceā€ but it could help folks who cannot travel for any reason.

4 Likes
16

I have attended lots of remote/online meetings and always found that ā€˜hybridā€™ of in person and online are the worst kind, as either kind of attendee.

As someone who mightā€™ve attended online before, and hopes to attend in the future, I would prefer that it be entirely online or entirely in person. Per the comments above, no binding decisions should be made at the LS, so those interested who canā€™t attend the summit have the usual options of participating in discussions afterwards, e.g. here on Discourse or on a PEP.

A

2 Likes
17

What you are saying is true, even with the best setup unfortunately.

Doing completely online is off the table I guess. I really like what Steve said.

This is true for both the sprint, LS and conferences in general. In the spirit of increase this, then Iā€™m leaning towards having two LS per year, one at PyCon and the other one at EP.

If we want to push it further, doing a third one at PyCon APAC.

Maybe we need to stress not to feel pressure to attend both.

2 Likes
18

Cambridge sprint follow-up: Open a poll about alternating the language summit between PyCon US and EuroPython Ā· Issue #309 Ā· python/steering-council Ā· GitHub

5 Likes
19

Am -1 with LS moving, I think country rotation is a lot for EuroPython, personally I would be demotivated by too many visas to catch up with core things. I think keeping one thing in one country and moving the annual sprint maybe was a good compromise. Its possible that if you came to PyCon US and missed annual core sprint, you could console yourself by attending core sprints at PyCon US for that year, well thats me in a year when am exhausted with visas.

20

While I understand visas are an issue for some, the converse of what you said is also true: having LS in US every year demotivates people from Europe/Rest of world to travel to LS for the same visa reasons/personal reasons. What youā€™re suggesting is to maintain the status quo, which is already unfair for people living outside the US. Rotation around major regions where core devs reside would make it fair for people from around the world.

I think country rotation is a lot for EuroPython,

Traditionally, EP has been in the Schengen area (with a few exceptions). I am not a visa adviser, but doesnā€™t that mean a single Schengen visa would suffice for all EP for the visa validity period? You donā€™t have to get a new visa for every Schengen country IIUC.

6 Likes