XMPP Council - 2017-07-19

Mhhh. Looks like s2s between this conference server and my server hangs again

daniel', have you notified the iteam?

Link Mauve: I just now noticed that. So no

Ok.

Also probably tricky if s2s hangs...

alright

seems it's about time

1) Roll call

o/

Here

Here

great..guess dave is still missing

Dave already said he wouldn’t.

right

2) Minute taker

I can do it

great...thanks

3) Suggestions for possible agenda topics?

There are some in trello, but they kind of all depend of XMPP Editor working again

I don't have anything

Me neither.

what about SamWhited?

have you send the mail about asking for help with the editor tasksß

have you send the mail about asking for help with the editor tasks?

I did not; I'm still worried about how we'd onboard someone.

I don't know how things work right now to teach them how things work.

how do we find out how things work right now?

I'm not sure where that information lives; it might be another thing that only Kev knows and I'm sure he doesn't have extra time given all his commitments

great

SamWhited, is there anything we as a council could do to obtain or create that information?

I can probably figure it out and update the readme if Kev can walk me through publishing a container again.

I feel like we are completely stuck currently, none of the decisions we make can have any use without an editor process.

So I’m willing to help you and Kev with whatever needs to be done to resume that.

Although I still think it won't be that useful since even if I recruit more people they won't have admin access which is currently required to publish anything.

But at least it would be progress to have docs again.

better than nothing I suppose

SamWhited, Link Mauve, how about opening a thread on Members list about this issue

?

maybe a lot members are unaware that there's still need for help

Alright; I'll go ahead and ask.

ta

4) Date of next

same day of week, same time of day?

In still not sure what to tell any volunteers we get though, but at least it would break the holding pattern

WFM

great

WFM.

5) AOB

seems there is none

thanks everybody

SamWhited: /docker/xmpp.org_extensions/create.sh

Well, `sudo /docker/xmpp.org_extensions/create.sh` actually.

Kev: thanks

that's step one anyways, I'll try to update the docs to reflect that

email going out shortly

If someone fancied writing a script that fetches the latest version of the image, checks if the version has changed, and if so restarts the container, I'm happy enough to cron it.

And then there's no manual step between merging PR on GitHub and it appearing on the website (albeit an hour later).

Multiple hours later since it takes about an hour for the build to even start and about an hour for the build to finish + whatever time a cronjob runs in (assuming the worst case scenario where the docker build finishes immediately after a cron run and you have to wait for the next one)

We should also review and do something about that PR that uses a base image for faster building.

eos:/docker/xmpp.org_extensions/> sudo docker pull xmppxsf/xeps 3:28PM Using default tag: latest latest: Pulling from xmppxsf/xeps 9f0706ba7422: Pull complete df331fee6036: Pull complete 318ea148ed48: Pull complete bf15fa51cc96: Pull complete f845ac678d97: Pull complete 916f6d8945ae: Pull complete b37315aac5fe: Pull complete c5de6dcf82c2: Pull complete 96bf79d78251: Pull complete d2ccf6702fc4: Pull complete a40013e9db37: Pull complete a1555b712484: Pull complete Digest: sha256:924f05c655272792051d4cdb5c0de6457be96a538cb0f9e504dfd4a7e3416320 Status: Downloaded newer image for xmppxsf/xeps:latest eos:/docker/xmpp.org_extensions/> sudo docker pull xmppxsf/xeps 3:31PM Using default tag: latest latest: Pulling from xmppxsf/xeps Digest: sha256:924f05c655272792051d4cdb5c0de6457be96a538cb0f9e504dfd4a7e3416320 Status: Image is up to date for xmppxsf/xeps:latest

Are the two outputs, if someone did want to script to check that pull has done anything.

Does restarting break requests (eg. is the container exposed to the internet directly, or is it behind a proxy like nginx)? Can't it just always restart?

It's behind a proxy, but you're still talking about the time to stop and start a container, which will cause requests to fail.

I'd rather not do that a few times an hour, if avoidable.

makes sense

Kev: how does the attic work in docker land? Does it still exist?

Currently, it doesn't, and we're just falling back to old style 'keep it on disk with no versioning whatsoever and hope nothing ever goes wrong'.

Which is as bad an idea now as it's always been :)

I'm inclined to suggest that we just store the generated files in Git for the attic.

(Be it the same repo or a different repo)

Kev: Okay, thanks; going to remove the stuff about generating the attic from the readme then

Well, we still need to do it somehow

I might look into how difficult it would be (I think it would be trivial) to ping the Internet Archive when a container starts up and just have it index everything

So better to put something less daft there, instead of what was done in the past, rather than nuking it, IM(NS)HO.

If you have suggestions I'm all ears, but right now I'm just trying to make sure these instructions are something a new editor could actually follow

Oh hey, the Internet Archive offers a service where you can create collections and ask them to crawl it every so often. I might sign up for this and make some collections for eg. the wiki and the XEP series. That's just useful in general.