WEBVTT

00:00.000 --> 00:10.600
Awesome, so we're almost at time. This next session is about recognition and why it matters

00:10.600 --> 00:16.320
for sustainability. Projects rely on far more than code itself. And when that work stays

00:16.320 --> 00:23.880
invisible, people tend to leave. Please welcome Jim Madge.

00:24.880 --> 00:31.880
Thank you. Right, so self-raising Lazarus, all contributors and how open source

00:31.880 --> 00:38.880
can rise again. So I thought that title was very clever and funny, but I've realised

00:38.880 --> 00:44.880
it's actually maybe not very easy to understand. So this is a depiction of the raising of Lazarus.

00:44.880 --> 00:52.880
This is a miracle of Jesus. So friend of his Lazarus was dead four days and Jesus arrived

00:53.880 --> 00:59.880
and rose in. Now, in the English-speaking world, we have something called self-raising flower.

00:59.880 --> 01:06.880
That is flower which has eleven-ing agents in. So when you make a batter or dough, it rises

01:06.880 --> 01:14.880
on its own. It's got everything it needs to rise. Hence self-raising Lazarus as a joke.

01:15.880 --> 01:21.880
Now, this is Malcolm, no, this is not Malcolm Tucker. This is Peter Crowdy who plays a character called

01:21.880 --> 01:27.880
Malcolm Tucker in the show called The Thick of It. And in one scene, I like very much.

01:27.880 --> 01:32.880
He says, this is a sense of depth to our sensibilities. He says, you're looking at Lazarus

01:32.880 --> 01:41.880
to eat arts and not just plain Lazarus, I'm self-raising Lazarus. So not only is it a difficult

01:41.880 --> 01:51.880
to understand, it's a joke that I still. But I like the title because one, I really like that

01:51.880 --> 02:00.880
quote, I think it's funny. And also it reminded me of how oposos projects can sort of survive a death

02:00.880 --> 02:06.880
and actually they've got the stuff they need to come back. So yeah, this is a story about

02:07.880 --> 02:13.880
open source in the small community projects. It's about volunteer labor. It's not about

02:13.880 --> 02:20.880
big commercial back projects. It's about the day-to-day struggle of volunteers and then

02:20.880 --> 02:27.880
actual kind of turnover. But yeah, so it's a little about the difficulties, but there's

02:27.880 --> 02:33.880
also some joy and I think it's about how projects can come back and like self-raising

02:33.880 --> 02:39.880
flower, how oposos means all the ingredients are there because something's come back.

02:39.880 --> 02:45.880
So a bit about me, I love free and open source software and everything open. I work as a

02:45.880 --> 02:53.880
research software engineer in the UK. I'm a steering committee member for the terrain

02:53.880 --> 03:01.880
way, which is a collaboratively written guide for open reproducible ethical

03:01.880 --> 03:06.880
data science. More recently, I decided to make some contributions to

03:06.880 --> 03:14.880
Jupiter book, which is a subject of Jupiter. And most recently, I've decided to take on

03:14.880 --> 03:22.880
this, being a maintainer of all contributors. This is me. So I can 94. My mum's taking the

03:22.880 --> 03:27.880
picture and I'm telling her there's this man from Finland on the use net group who's made

03:27.880 --> 03:34.880
a free operating system that we should really have a look at. So we need to know a bit

03:34.880 --> 03:41.880
about all contributors. All contributors is a project all about recognising the

03:41.880 --> 03:46.880
contributors to our open source projects. And the diversity of the types of

03:46.880 --> 03:52.880
contributions they make and important that means it's not just about code. It's a

03:52.880 --> 03:59.880
number of things. There's a specification which enforces things like the

03:59.880 --> 04:06.880
record of contributors must be prominently displayed. It encourages categorisation

04:06.880 --> 04:12.880
so to show the types of contributions people make. And it also discourages certain things

04:12.880 --> 04:19.880
such as trying to scrutinise the quality or size of contribution. It's not about

04:19.880 --> 04:25.880
who's the best contributor, who's the most active contributor. It's about what people

04:25.880 --> 04:34.880
have brought to a project. In practice, that's a JSON file. And so managing that by hand

04:35.880 --> 04:41.880
is not fun. It's our approach. So there are sets of tools also. There's a CLI which

04:41.880 --> 04:48.880
doesn't most job of managing this for you. And I think probably the way people interact

04:48.880 --> 04:55.880
with this, the most, it's probably the bot. So there's a, there's a get her bot.

04:55.880 --> 05:02.880
And just by leaving comments in poor requests or issues asking the bot to add someone

05:02.880 --> 05:07.880
it will, it will come along and they can report requests making those changes for you.

05:07.880 --> 05:11.880
And I think, if I'm, that's, that's caught in my speech.

05:11.880 --> 05:16.880
It's the access communities where you might have maintainers, community managers who

05:16.880 --> 05:22.880
want to get involved in this sort of thing, but wouldn't be so happy using the CLI.

05:22.880 --> 05:28.880
The project has maybe not a huge, but I think quite a sort of passionate

05:28.880 --> 05:34.880
views, and I think there's some common things we can say about the people who use this project.

05:34.880 --> 05:41.880
The, the projects who really want their contributors to feel valued and, and recognize

05:41.880 --> 05:46.880
they're often projects where there's a lot of work which isn't represented in the

05:46.880 --> 05:51.880
committee history. So I think lots of community management work, promotional work,

05:51.880 --> 05:55.880
maybe finance and sustainability, that sort of thing doesn't, doesn't show up in

05:55.880 --> 06:00.880
it, but is, is critical, really, really important. I think there's sort of an interesting

06:00.880 --> 06:06.880
cluster of projects which are kind of academic to adjacent to academia.

06:06.880 --> 06:10.880
And I think maybe that's what's something to do with. There are, there are similar

06:10.880 --> 06:16.880
conversations going on there about what taxonomies should be used for crediting people

06:16.880 --> 06:22.880
with research, how do you acknowledge people who are contrary to research, but are not

06:22.880 --> 06:29.880
as of other paper. There's an interesting kind of overlap there. And because you can

06:29.880 --> 06:36.880
generate with the, with the, with the CLI, these nice HTML tables, sharing all the

06:36.880 --> 06:41.880
contributors, I think, your projects where there's some significant presence on the web,

06:41.880 --> 06:47.880
whether that's the thing you're making is, is web based or or a documentation site.

06:47.880 --> 06:53.880
They tend to use it a lot. As an example, if the internet is still working, the

06:53.880 --> 07:00.880
training way, which I mentioned earlier, which I work on.

07:00.880 --> 07:09.880
Yes, this is the on this page on the training way where we have a table of, of

07:09.880 --> 07:15.880
the contributors, the little emoji tell you what sort of contribution types they've made.

07:15.880 --> 07:22.880
So that's a little example of what it might have got like.

07:22.880 --> 07:28.880
But all is not well. And if you were paying attention on GitHub, you probably would have

07:28.880 --> 07:35.880
noticed some bad signs. It's been quite a while since there were any releases.

07:35.880 --> 07:43.880
So the pensies are now quite old. There were a lot of issues hanging around, whether

07:43.880 --> 07:48.880
that book reports or feature requests. I think things weren't getting done.

07:48.880 --> 07:56.880
In general, not much engagement from the maintainers. And I think sort of pulling on the

07:56.880 --> 08:00.880
idea, like I said, this is like value. There are people who really care about this.

08:00.880 --> 08:06.880
They read it like this. So how does that happen? What's going on? Part of that, I think,

08:06.880 --> 08:11.880
is two few people. It's a man's of work. This is number of people.

08:11.880 --> 08:17.880
It just wasn't making sense. Of the people that do contribute. Of course, they've got

08:17.880 --> 08:21.880
lives. They've got things to do. I don't think we can hold it against them for saying that

08:21.880 --> 08:26.880
their family or their work might come before this sort of hobby.

08:26.880 --> 08:31.880
Burn out a lot of interest. I think maybe for this particular case probably more on the

08:31.880 --> 08:38.880
loss of interest side. It's completely fine to really jump into something and have a

08:38.880 --> 08:42.880
lot of fun. If it starts getting boring, if it starts getting like a chore,

08:42.880 --> 08:49.880
I think that's fine. That's just natural. People are going to want to move on.

08:49.880 --> 08:53.880
Something else with that time.

08:53.880 --> 09:02.880
Another important point which I've heard Mike McCoy talk about this room in

09:02.880 --> 09:08.880
previous years is which I quite like. It's this idea of having a pipeline to

09:08.880 --> 09:12.880
take people who are users, to an image of contributors, and actually hopefully

09:12.880 --> 09:18.880
turn them into maintainers. I think that just wasn't happening. You could see

09:18.880 --> 09:23.880
there were occasionally some users who would make occasional contributions, but

09:23.880 --> 09:28.880
very few people sort of got to the point where they were really active in

09:28.880 --> 09:33.880
maintaining the project beyond something like I want to add a feature which I

09:33.880 --> 09:36.880
personally would like to see.

09:36.880 --> 09:41.880
Focus a little bit. Bus factor. I think it was a problem.

09:41.880 --> 09:46.880
The GitHub organization, a few and five people were able to pay maintenance role

09:46.880 --> 09:51.880
of those. I think approximately zero were active.

09:51.880 --> 09:55.880
Crowd in which is the platform for managing translations. There were two people

09:55.880 --> 10:04.880
to manage that. The worst thing hosting the domain and DNS was completely

10:04.880 --> 10:08.880
reliant on one person.

10:08.880 --> 10:12.880
This sort of talks about a resurrection.

10:12.880 --> 10:20.880
A resurrection has to be a death. The sort of trigger for this was the

10:20.880 --> 10:24.880
the all countries as website disappeared.

10:24.880 --> 10:27.880
People started noticing.

10:27.880 --> 10:32.880
Lea Weser, who is the founder of PiOpenSy, was one of the people

10:32.880 --> 10:35.880
that noticed and kind of started a conversation.

10:35.880 --> 10:38.880
I saw on MasterDawn and it sort of broke out into various

10:38.880 --> 10:41.880
groups. People saying, that's really disappointing.

10:41.880 --> 10:45.880
We feel this is important. We don't want it to go away.

10:45.880 --> 10:49.880
There are people saying, well, what can I do?

10:49.880 --> 10:52.880
I'd be willing to step up and help.

10:52.880 --> 10:56.880
It's a small team that has started to form.

10:56.880 --> 11:02.880
I think this is roughly who we are at the moment.

11:02.880 --> 11:05.880
There's there. I am there.

11:05.880 --> 11:10.880
Of these people, Josh has contributed to the project before.

11:10.880 --> 11:15.880
Maximilian has contributed and has done some maintenance work

11:15.880 --> 11:18.880
in the past, but it's sort of gone dormant.

11:18.880 --> 11:21.880
Yeah, he's keen to help again.

11:21.880 --> 11:25.880
We wouldn't be possible to all without support

11:25.880 --> 11:28.880
from previous maintainers and they can help for

11:28.880 --> 11:31.880
and they are involved in helping us sort out things,

11:31.880 --> 11:35.880
but they're not in a position to come back

11:35.880 --> 11:38.880
and play a very active maintenance role.

11:39.880 --> 11:42.880
So we got together.

11:42.880 --> 11:46.880
Christine and I talked about what's the plan

11:46.880 --> 11:48.880
and what are we going to do?

11:48.880 --> 11:53.880
The first thing we decided to approach was documentation.

11:53.880 --> 11:58.880
The reason we came to that was it seemed like the urgent thing

11:58.880 --> 12:01.880
that the doc says is down.

12:01.880 --> 12:04.880
That's fixed that.

12:05.880 --> 12:08.880
I think we kind of felt documentation is quite a good way

12:08.880 --> 12:10.880
to start learning a project.

12:10.880 --> 12:13.880
You get to see a bit of different things.

12:13.880 --> 12:16.880
It was also, perhaps, a way to avoid

12:16.880 --> 12:20.880
jumping into JavaScript and a code base we didn't understand

12:20.880 --> 12:21.880
right away.

12:21.880 --> 12:24.880
So we're all much more comfortable with Markdown

12:24.880 --> 12:27.880
generating a static site.

12:27.880 --> 12:31.880
So we got dug in and there were a few challenges there.

12:31.880 --> 12:36.880
So I think not surprising given that there hadn't been maintenance.

12:36.880 --> 12:39.880
A number of dependencies were very old,

12:39.880 --> 12:44.880
several major versions old full of vulnerabilities.

12:44.880 --> 12:47.880
The site was built using docu source,

12:47.880 --> 12:50.880
which I had not seen before,

12:50.880 --> 12:54.880
and that was a number of major versions out.

12:54.880 --> 12:57.880
So it was going to require some changes anyway

12:57.880 --> 13:00.880
and some people in the team were kind of opinionated.

13:00.880 --> 13:03.880
I don't really like this, I'm not sure, for Rastro.

13:03.880 --> 13:06.880
This feels like an opportunity will switch.

13:06.880 --> 13:09.880
And so the similarly in the documentation,

13:09.880 --> 13:12.880
there were some things which were now old enough

13:12.880 --> 13:14.880
that they were just kind of out of date.

13:14.880 --> 13:18.880
So it wasn't the easiest task,

13:18.880 --> 13:21.880
but we could see a route through.

13:21.880 --> 13:26.880
And yeah, we got stuck in project boards,

13:26.880 --> 13:28.880
got working on that.

13:28.880 --> 13:32.880
We also wanted to make sure we were sort of building in more resilience,

13:32.880 --> 13:34.880
because that seemed to be a problem.

13:34.880 --> 13:38.880
And I think a lot of the work was really sending emails

13:38.880 --> 13:42.880
being very patient for people who we completely relied on

13:42.880 --> 13:46.880
in some cases the only person who had access to the DNS.

13:46.880 --> 13:48.880
We just need to work with this person,

13:48.880 --> 13:51.880
but there was a very time pour.

13:51.880 --> 13:55.880
So it has to be sort of quite a lot of efforts and patience

13:55.880 --> 13:59.880
and understanding on our side about the speed at which

13:59.880 --> 14:01.880
those things like move.

14:01.880 --> 14:05.880
I'm happy to say where we are now,

14:05.880 --> 14:09.880
the site is back, so it's online,

14:09.880 --> 14:15.880
and it's sort of important that it's back under our control.

14:15.880 --> 14:21.880
So I think this is updated and any changes we want to make.

14:21.880 --> 14:25.880
That's fine, we're no longer sort of relying on people

14:25.880 --> 14:27.880
who are sort of difficult to contact,

14:27.880 --> 14:29.880
if we're to get the attention of.

14:29.880 --> 14:33.880
We are, as of maybe two weeks ago,

14:33.880 --> 14:37.880
we're starting now to shift focus to the CLI.

14:37.880 --> 14:39.880
Some of the problems there, lots of other dependencies,

14:39.880 --> 14:43.880
a lot of untangling bits of code that have been deprecated,

14:43.880 --> 14:47.880
finding replacements, and getting control of bits and infrastructure there.

14:47.880 --> 14:49.880
So I think we're quite hopeful that,

14:49.880 --> 14:53.880
or at least going to have a patch for these where the update is

14:53.880 --> 14:57.880
when you install this or no longer shout out to you about

14:57.880 --> 15:01.880
millions of vulnerabilities and the URL packages.

15:01.880 --> 15:04.880
The bus factor has significantly improved,

15:04.880 --> 15:07.880
I mean, I think domain and DNS two people,

15:07.880 --> 15:10.880
so it doesn't really feel terribly comfortable,

15:10.880 --> 15:13.880
but we're in a much better position than we were.

15:13.880 --> 15:16.880
Unfortunately, I'm really kind of,

15:16.880 --> 15:20.880
I'm disappointed, I'm kind of upset that in this migration,

15:20.880 --> 15:24.880
we decided getting a website back up and back under control

15:24.880 --> 15:28.880
was worth temporarily losing the translations.

15:28.880 --> 15:32.880
I think that bothers me, because it's really important for accessibility,

15:32.880 --> 15:35.880
and that's something we really want to move on to,

15:35.880 --> 15:39.880
but it just seemed like the best compromise at the time.

15:39.880 --> 15:42.880
Yes, so where we're going now,

15:42.880 --> 15:45.880
a short term, get those translations back,

15:45.880 --> 15:48.880
like I said, make a release of the CLI,

15:48.880 --> 15:51.880
and just make sure we actually go access to all of the infrastructure,

15:51.880 --> 15:55.880
the various bits of things that the bot run,

15:55.880 --> 15:58.880
build and see eyes and things,

15:58.880 --> 16:04.880
longer term, I think we really need to think about this issue of,

16:04.880 --> 16:06.880
how do you get your users,

16:06.880 --> 16:09.880
and turn them into contributors so that they're sort of

16:09.880 --> 16:11.880
that to the engaging and reporting,

16:11.880 --> 16:14.880
and that there's a much clearer route to,

16:14.880 --> 16:17.880
if you want to get involved with this matters to you,

16:17.880 --> 16:18.880
you can get involved,

16:18.880 --> 16:22.880
and hopefully avoid this sort of low again.

16:22.880 --> 16:25.880
Yeah, and then eventually move on to thinking about,

16:25.880 --> 16:27.880
where do we want this to go?

16:27.880 --> 16:30.880
Can we start picking up some of these bugs that have been around for ages?

16:30.880 --> 16:32.880
What features do we want to add?

16:32.880 --> 16:34.880
Yeah, so there's some challenges,

16:34.880 --> 16:37.880
we're not all super familiar with whichever script,

16:37.880 --> 16:39.880
that we found, but I think we've got this assistance,

16:39.880 --> 16:41.880
and we're all quite keen.

16:41.880 --> 16:43.880
I'm hopeful of the future.

16:43.880 --> 16:46.880
Some lessons, probably not a simple one,

16:46.880 --> 16:48.880
but build and resilience from the ground up.

16:48.880 --> 16:53.880
It's way too fragile to have things depend on a single person.

16:53.880 --> 16:56.880
It's also way too fragile to have things depend on individuals,

16:56.880 --> 17:00.880
really, you need to be thinking about what teams of people

17:00.880 --> 17:03.880
are going to manage things, how do you organize the teams?

17:03.880 --> 17:05.880
Folks on what you can do,

17:05.880 --> 17:08.880
and I think the flip side of the other way to say that is,

17:08.880 --> 17:10.880
don't worry too much about the things you can't do,

17:10.880 --> 17:12.880
so there might be something that's very important,

17:12.880 --> 17:13.880
if you're not unable to do it,

17:13.880 --> 17:17.880
it's maybe not worth worrying about it too much.

17:17.880 --> 17:21.880
I'm not sure if this is one of these might be a bit controversial.

17:21.880 --> 17:25.880
I think in bases going slow, community project like this,

17:25.880 --> 17:30.880
you've got to protect your own sanity and happiness,

17:30.880 --> 17:34.880
it's not going to end of the world if things take a bit longer

17:34.880 --> 17:36.880
than you would have hoped.

17:36.880 --> 17:38.880
Mason's before features,

17:38.880 --> 17:40.880
like I was explaining earlier,

17:40.880 --> 17:42.880
like having the site there,

17:42.880 --> 17:44.880
the fact that it's there, it works, it's usable.

17:44.880 --> 17:45.880
Okay, that's fine.

17:45.880 --> 17:48.880
There's loads of stuff we'd like to do and could do to make it better,

17:48.880 --> 17:50.880
but there's that baseline of,

17:50.880 --> 17:51.880
can you use this,

17:51.880 --> 17:53.880
and I think that's the critical thing,

17:53.880 --> 17:56.880
I'm sorry, I didn't remove all that.

17:56.880 --> 17:59.880
But the world bike shedding,

17:59.880 --> 18:01.880
bike shedding is spending,

18:01.880 --> 18:03.880
as of unreasonable amount of time,

18:03.880 --> 18:05.880
fussing or worrying,

18:05.880 --> 18:07.880
or taking action on things which are not important.

18:07.880 --> 18:09.880
And I do,

18:09.880 --> 18:11.880
I think I'm reflecting a bit on,

18:11.880 --> 18:13.880
maybe it was that,

18:13.880 --> 18:15.880
was the doc's reading most important thing,

18:15.880 --> 18:17.880
maybe we should have,

18:17.880 --> 18:19.880
with the jumped into the cell I first.

18:19.880 --> 18:21.880
I'm not too sure.

18:23.880 --> 18:25.880
And another sort of reflection,

18:25.880 --> 18:27.880
when as I was digging a bit into the

18:27.880 --> 18:29.880
doing it with get archaeology,

18:29.880 --> 18:31.880
looking at the contribution history,

18:31.880 --> 18:34.880
it's sort of hard to figure out exactly what happened,

18:34.880 --> 18:35.880
but I think,

18:35.880 --> 18:37.880
it's probably isn't the first time this has happened,

18:37.880 --> 18:38.880
you can sort of see,

18:38.880 --> 18:40.880
looking through who's contributing,

18:40.880 --> 18:42.880
looking at the frequency of commits,

18:42.880 --> 18:44.880
there are definite peaks and troughs.

18:44.880 --> 18:45.880
There's an initial flurry,

18:45.880 --> 18:47.880
whether the spec and the CLI are written,

18:47.880 --> 18:49.880
there's a bit of a quiet period,

18:49.880 --> 18:51.880
there's another flurry where someone's working on the,

18:51.880 --> 18:53.880
the GitHub bot,

18:53.880 --> 18:55.880
and there's another quiet period of maintenance,

18:55.880 --> 18:57.880
and I think there's a different thing,

18:57.880 --> 18:58.880
most recently,

18:58.880 --> 19:00.880
was that it wasn't a quiet period of maintenance,

19:00.880 --> 19:01.880
it was sort of a flat line,

19:01.880 --> 19:03.880
nothing was happening at all.

19:03.880 --> 19:06.880
But it's sort of isn't the first time this has happened,

19:06.880 --> 19:08.880
we're not the first teams to kind of

19:08.880 --> 19:10.880
breathe some life back to this project,

19:10.880 --> 19:12.880
we might not be the last,

19:12.880 --> 19:14.880
and so yeah,

19:14.880 --> 19:16.880
I think the most important lesson is,

19:16.880 --> 19:17.880
if there's a community,

19:17.880 --> 19:19.880
if there are people that really value something

19:19.880 --> 19:20.880
and really care,

19:20.880 --> 19:24.880
then our projects can come back.

19:24.880 --> 19:27.880
I hope the source comes out again.

19:27.880 --> 19:31.880
So thanks so much for listening,

19:31.880 --> 19:35.880
I'd especially like to thank your own

19:35.880 --> 19:38.880
Magnus who's the original Crater and maintainer,

19:38.880 --> 19:40.880
Kent Dodson, Jake Bellum,

19:40.880 --> 19:44.880
who are a table of very important role

19:44.880 --> 19:46.880
and maintenance,

19:46.880 --> 19:48.880
therefore having to get us together.

19:48.880 --> 19:51.880
Everyone who's contributed all contributors,

19:51.880 --> 19:53.880
and also the trainway,

19:53.880 --> 19:54.880
can you can describe area,

19:54.880 --> 19:56.880
some of the other stations,

19:56.880 --> 19:57.880
you can sing,

19:57.880 --> 19:59.880
have been made by them,

19:59.880 --> 20:01.880
so they talk would be a lot more boring without those.

20:01.880 --> 20:02.880
If you want to get involved,

20:02.880 --> 20:04.880
if the sound is interesting,

20:04.880 --> 20:07.880
use the button, your own projects,

20:07.880 --> 20:09.880
recognise all of your contributors

20:09.880 --> 20:11.880
and all the wonderful things they do,

20:11.880 --> 20:13.880
you can read the dots,

20:13.880 --> 20:15.880
if you're interested in joining,

20:15.880 --> 20:16.880
please reach out,

20:16.880 --> 20:18.880
we're looking for people to help.

20:18.880 --> 20:20.880
Yes, thanks very much.

20:21.880 --> 20:26.880
Time for some questions.

20:26.880 --> 20:28.880
Anybody have any questions?

20:32.880 --> 20:33.880
I'm just curious,

20:33.880 --> 20:38.880
how did you regain control over your project

20:38.880 --> 20:42.880
in terms of GitHub permissions and write?

20:42.880 --> 20:45.880
Or was it in your team already,

20:45.880 --> 20:49.880
someone who has admin rights?

20:50.880 --> 20:52.880
Did you say permissions to write?

20:52.880 --> 20:54.880
The permission of GitHub rights,

20:54.880 --> 20:55.880
like GitHub rights?

20:55.880 --> 20:56.880
Yeah.

20:56.880 --> 20:59.880
So the question was,

20:59.880 --> 21:02.880
how did we get the permissions we needed

21:02.880 --> 21:04.880
to start maintenance?

21:04.880 --> 21:09.880
Was someone in our team sort of already had those permissions?

21:09.880 --> 21:11.880
Sort of.

21:11.880 --> 21:14.880
So I suppose the initial thing was,

21:14.880 --> 21:16.880
we were opening issues,

21:16.880 --> 21:18.880
on repository saying,

21:18.880 --> 21:20.880
does anyone have access?

21:20.880 --> 21:21.880
What's going on?

21:21.880 --> 21:24.880
We would be willing to help out.

21:24.880 --> 21:28.880
I suppose Maximilian was the person

21:28.880 --> 21:32.880
who did have permissions

21:32.880 --> 21:35.880
and was sort of paying attention to that

21:35.880 --> 21:37.880
and helped us sort of get on board a bit.

21:37.880 --> 21:38.880
But yeah, overall,

21:38.880 --> 21:41.880
we couldn't have done it without those people.

21:41.880 --> 21:44.880
There were a few places where we cut things out.

21:44.880 --> 21:47.880
So originally all the CI was on circle CI

21:47.880 --> 21:50.880
and we've cut that out and we're using GitHub actions

21:50.880 --> 21:53.880
and that was a place where we could do that.

21:53.880 --> 21:56.880
I think that the hostings on that defined,

21:56.880 --> 21:58.880
I think that was another place where we decided,

21:58.880 --> 22:01.880
okay, it's just going to be easier to set up a new team

22:01.880 --> 22:02.880
and start fresh.

22:02.880 --> 22:05.880
But some of the things like the DNS and GitHub,

22:05.880 --> 22:07.880
we really needed those people.

22:07.880 --> 22:10.880
So some of that was people noticed on GitHub when we asked,

22:10.880 --> 22:13.880
some of that was we did some digging to find emails

22:14.880 --> 22:17.880
and send emails and we're very polite and patient.

22:17.880 --> 22:19.880
But yeah, we couldn't have done it without those people.

22:19.880 --> 22:22.880
And there are the interventions, the helpful and enthusiastic.

22:22.880 --> 22:25.880
Any other questions?

22:36.880 --> 22:39.880
Let's take an example and have like,

22:39.880 --> 22:42.880
imagine you have a project that you expect

22:42.880 --> 22:44.880
to communicate to die for a few years

22:44.880 --> 22:46.880
and then resurrect at some point.

22:46.880 --> 22:50.880
Do you have any tips for the maintainers,

22:50.880 --> 22:55.880
I guess, but I'll get a big bug before everything shut down.

22:55.880 --> 22:57.880
Like is it documentation for,

22:57.880 --> 23:00.880
I mean, you don't try this documentation the same for tomorrow

23:00.880 --> 23:02.880
and for next year, I guess.

23:02.880 --> 23:04.880
So any tips?

23:04.880 --> 23:07.880
Yeah, the question was,

23:07.880 --> 23:11.880
if you expect a project to go through

23:11.880 --> 23:15.880
have some flows and maybe quite a down or down a dive for it,

23:15.880 --> 23:20.880
what can you do to make sure that people are able to pick it up in the future?

23:20.880 --> 23:24.880
I suppose what I can say is what we found difficult

23:24.880 --> 23:27.880
and I think yeah, we should be really careful

23:27.880 --> 23:30.880
to think on that and try and maybe make sure we do a better job.

23:30.880 --> 23:33.880
And I think you are right in suggesting documentation.

23:33.880 --> 23:38.880
And particularly, it's that kind of developer documentation.

23:38.880 --> 23:42.880
So we're finding bits of infrastructure which are kind of important

23:42.880 --> 23:45.880
for maybe running the GitHub boss or testing that.

23:45.880 --> 23:47.880
And we just didn't know it existed.

23:47.880 --> 23:49.880
It's not written down anywhere.

23:49.880 --> 23:52.880
Or it might be written down somewhere, it's very obscure

23:52.880 --> 23:55.880
and a kind of a bleak reference and it just really makes sense.

23:55.880 --> 24:00.880
So I think really documenting who has access to things,

24:00.880 --> 24:02.880
what different, what you're,

24:02.880 --> 24:05.880
a piece of the infrastructure are, what they're for.

24:06.880 --> 24:09.880
Even in really very simple terms,

24:09.880 --> 24:13.880
we're super, super helpful.

24:13.880 --> 24:15.880
I think we just had one more question, right?

24:15.880 --> 24:16.880
You're good?

24:16.880 --> 24:17.880
Okay.

24:17.880 --> 24:18.880
All righty.

24:18.880 --> 24:19.880
Anyway, thanks, Jim.

24:19.880 --> 24:20.880
Thank you.

24:20.880 --> 24:22.880
Thank you.