As in many flocks and fudcons before them, I’d like us to have a Infrastructure and/or Releng hackfest at flock later this year, but I thought it might be good to discuss what kind of format or topics folks would like to cover.
The last year or two we did a ‘barcamp’ style thing. This is basically where we spend the first 30min or whatever listing down topics/areas, and then everyone votes and we do them in the order of votes. This is good in that we can cover things people there want to, but bad in that we can’t really prepare too much ahead of time and we also can easily drift if we aren’t being careful.
We have also tried at various times focusing on things like:
Helping people with applications/issues/ideas to deploy
Onboarding new folks (This is great, but we often don’t have too many people available in person that want to onboard while so much else is going on at flock).
focused on a particular area we want to get something done in (docs, deploy some hard to deploy thing, convert something, etc)
planning the future. This is usually great, but often just has us having a nice discussion without too much actionable things coming out of it.
The CPF has 4 hour and 8 hour hackfest requests.
IMHO we should prioritize:
High bandwith discussion type things
Things that require stakeholders that are all available
No decisions on things, but rather plans we can bring back to everyone (including those folks who cannot be there, or join remotely)
There’s also a number of BIG things looming later this year, although we may be pretty far along on them by the time flock hits, but:
fate of dist-git
fate of bug tracking
etc
There’s a ton of things we could do… but I would really love to hear from all of you:
What topics/areas would you be most interested in?
Would you prefer something more planned, or a ‘barcamp’ / wingit type of workshop?
Would you like more discussion or more ‘doing things’?
How much time would you think you would want to spend? 4 hours? 8? less?
Of the list of things tried in the past, this is my favorite of the four things.
This is not dependent on the CFP and could happen on an ongoing basis after the CFP closes. But is it possible to come up with a shortlist of potential projects that could be worked on during the hackfest? Especially the high-bandwidth things that are easier to do when people are together in a room.
This was my second favorite idea.
I think four hours is an ideal length. Admittedly, I am not sure we should even offer the option for full-day hackfests. It can ask for a lot, and usually people want to hop around at some point too.
The four hours could be divided into thematic areas too, to maximize participation of people in topics that are most interesting. Perhaps two hours could be a general planning/discussion topic, and the second half could be on some focused project(s).
My only request might be some kind of OpenShift / Ansible intro of Fedora Infra. Especially with Communishift now being available, I could see it being useful to support skill development of community members in these tools. Perhaps it could even be as simple as a walkthrough / guided setup of a Communishift instance for a simple app, or a real Fedora Infra app.
Of the list of things tried in the past, this is my favorite of the four things.
This is not dependent on the CFP and could happen on an ongoing basis after the CFP closes. But is it possible to come up with a shortlist of potential projects that could be worked on during the hackfest? Especially the high-bandwidth things that are easier to do when people are together in a room.
Yeah, we totally could!
Is there likey to be other sessions on dist-git/forge things?
If so, we could leave that to those, but if not, that might be a good
thing to discuss…
This was my second favorite idea.
I think four hours is an ideal length. Admittedly, I am not sure we should even offer the option for full-day hackfests. It can ask for a lot, and usually people want to hop around at some point too.
Yeah, 4 hours seems reasonable to me too.
The four hours could be divided into thematic areas too, to maximize participation of people in topics that are most interesting. Perhaps two hours could be a general planning/discussion topic, and the second half could be on some focused project(s).
good thought.
My only request might be some kind of OpenShift / Ansible intro of Fedora Infra. Especially with Communishift now being available, I could see it being useful to support skill development of community members in these tools. Perhaps it could even be as simple as a walkthrough / guided setup of a Communishift instance for a simple app, or a real Fedora Infra app.
yeah, we could totally do that.
or would that be good as a seperate talk perhaps? Or not generally
interesting?
My only request might be some kind of OpenShift / Ansible intro of Fedora Infra. Especially with Communishift now being available, I could see it being useful to support skill development of community members in these tools. Perhaps it could even be as simple as a walkthrough / guided setup of a Communishift instance for a simple app, or a real Fedora Infra app.
yeah, we could totally do that.
or would that be good as a seperate talk perhaps? Or not generally
interesting?
I think this is a great idea but it also may be covered at Devconf as well. If it does get covered in Devconf I have an idea that might work. A group workshop where folks get together, guided by one or more people who are confident in Communishift, write a simple demonstration application, deployment config, and deploy it together as a group.
I thought at devconf folks were mostly going to talk about communishift
side of things. I think we do need more docs/understanding of how to
move to staging/prod after that. I could be wrong on the scope of the
devconf talk tho.
A demo app might be interesting. We could just delete it once we are
done…
My heart wants to believe there will be a git forge workshop. My brain knows that the timing of the recent announcement with the CFP close date later this month might mean there will not be a dist-git session. Unless someone else takes ownership of this topic for a Flock workshop, it would be wise to plan for spill-over of this topic to come up in the workshop.
A workshop specifically focused on Communishift would be interesting to me. My hunch is that a workshop targeting traditional sysadmin folks who are trying to orient themselves on OpenShift would be useful. My hunch could be wrong.
I like this idea. Even walking people through a test deployment of some test application could be useful. Alternatively, it might be fun to try walking through how an existing application is deployed, although I would not want to lose focus on the self-service end of Communishift.
My aim is giving people more practical skills and experience on how they can dive into this stuff themselves.
My heart wants to believe there will be a git forge workshop. My brain knows that the timing of the recent announcement with the CFP close date later this month might mean there will not be a dist-git session. Unless someone else takes ownership of this topic for a Flock workshop, it would be wise to plan for spill-over of this topic to come up in the workshop.
Yep. Thats what I was wondering… discussing pros and cons is a very
good thing to do in a high bandwith in person discussion. I just wasn’t
sure how much will be left to discuss at that point. But we can never
know the future.
A workshop specifically focused on Communishift would be interesting to me. My hunch is that a workshop targeting traditional sysadmin folks who are trying to orient themselves on OpenShift would be useful. My hunch could be wrong.
Yeah, could be indeed…
I like this idea. Even walking people through a test deployment of some test application could be useful. Alternatively, it might be fun to try walking through how an existing application is deployed, although I would not want to lose focus on the self-service end of Communishift.
My aim is giving people more practical skills and experience on how they can dive into this stuff themselves.
Yeah, communishift is really quite open. Once you have the project you
can deploy/manage/tweak your app all you like.
When/if something goes to staging, it’s more constrained. Deployment is
done via ansible from git, although you can tweak things manually a
little bit.
Going to production after that is more locked down still. You have to
use the git repo/ansible to make any changes, you can’t manually tweak
anything.
But I am not sure how much all of that is interesting to people outside
folks wanting to get their application running in infrastructure.
The communishift is not a bad idea for topic, same for ansible basics in Fedora Infrastructure. It would also be interesting to talk about state of apps we are currently hosting in Fedora Infra and where we need help maintaining.
I would also like to submit some talk about state of Fedora Infrastructure what we did and what is planned.
The communishift is not a bad idea for topic, same for ansible basics in Fedora Infrastructure. It would also be interesting to talk about state of apps we are currently hosting in Fedora Infra and where we need help maintaining.
All good ideas.
I can write up a hackfest proposal perhaps here soon and we can adjust
as we get closer too.
Do keep the ideas coming!
I would also like to submit some talk about state of Fedora Infrastructure what we did and what is planned.
maybe this isnt infra related but i think a hackfest surrounding fixing the wiki would probably be the most impactful thing the community could help with and also is something that is very easy to massively parallelize
i could basically grab every wikipage ive ever seen (as basically all my interactions with the wiki other than creating my userpage have led me to frustration) and bad/old information and start a list of pages to edit, or we could have people randomly pick maybe
From an infrastructure standpoint we are trying to move off the wiki
onto docs.fedoraproject.org pages. But yeah, that could possibly be a
good thing to get folks working on at the workshop!
So, the infra and releng hackfest was accepted for flock.
It’s on friday aug 9th starting at 9am and running until 1pm.
(4 hours)
I want to try and see if we can at least hash out some
things we definitely want to get to and then some nice to haves.
I really think we want to prioritize things where we need to discuss
or come up with a plan over things that we could as easily do
anytime. We should also be mindful to not actually decide things,
but instead plan things and present those plans in the way
we normally would for coment from people not able to attend.
That said, he’s a first cut of what I think we should try for:
(This is culled from replies here, with some things I added):
Come up with some proposal for standards for openshift apps
Discuss infra-sig packages maint: Propose list to add/remove
Should we try and figure out order of things now?
Or just wait until day of and vote on them based on the people that are
there? I like that in that it would take into account people who are
there, but I dislike it because we have 4 hours and if we spend the
first 30min deciding what we should talk about we can talk about less
things.
