IPFS Team Structures - Working & Research Groups (Was: "Proposal: IPFS Working Groups") #285
Conversation
ghost
assigned 
|
Thanks! This looks pretty good to me. Should some of these rather be product teams/groups than working groups? Not all of the groups seem like verticals:
The internationalization/localization group could probably be marked as #future. I guess the next step after this document is for the working groups to constitute, make a repo, and refine+approve their goals? |
|
Love the focus on themes. +1 on captains (or max 2 co-captains) for each of these which will help scale.
|
|
I would structure this differently. If you don't group these they get confusing, especially when you start listing products like ipfs-cluster alongside ongoing efforts like supporting giant files. Here's how I think of it, though I'm not bound to the names for each grouping -- we might find better names.
|
|
We should also identify which are the key working groups who should be defining at least 1 KR, which ones should define their own set of OKRs, etc. |
WORKING_GROUPS.md
Outdated
|
|
||
| ## Description | ||
|
|
||
| The IPFS Project Working Groups are teams of people that are appointed to research, develop and deploy new work under the scope that the working group is focused. This structure is designed to provide clarity and direction to the project, enabling individual contributors to focus their time and energy on the areas they are most interested. |
There was a problem hiding this comment.
why specifically mention new work? a lot of these WGs do ongoing work that needs perennial attention.
There was a problem hiding this comment.
Perhaps new was a poor word choice here. I meant new as in their are creating that in the scope. I'll remove it as it is not necessary.
|
Do you mind if I add those headings and generate a Table of Contents to make the doc more readable? |
WORKING_GROUPS.md
Outdated
|
|
||
| ## Current Working Groups | ||
|
|
||
| ### Release |
There was a problem hiding this comment.
Release Working Groups
- JavaScript Releases
- Go Releases
There was a problem hiding this comment.
I considered this and I'm confident that releasing JS and Go can be handled by the same team. In the end, the Release team should focus on auditing the release, ensuring that the quality bar didn't lower (check features added, CI, documentation and so on), that communication with the community and partners happens (changelog, post, twitter, reddit, emails and so on).
Yes, there are some people that have more context of what is happening in Go and others in JS, but in the end it is the same protocol and given that this is a Working Group and not a Working individual, we can have 2~4 people that share enough context to be capable of launching the rocket with the same level of quality.
WORKING_GROUPS.md
Outdated
|
|
||
| ### Release | ||
|
|
||
| The Release Working Group manages the releases for go-ipfs, js-ipfs and its internal dependencies. |
There was a problem hiding this comment.
This should address relationship between IPFS and libp2p, or the document should give an overarching statement about the relationship between IPFS work and libp2p work, and how that impacts the scope of each working group.
WORKING_GROUPS.md
Outdated
|
|
||
| ### Specifications | ||
|
|
||
| The Specifications Working Group coordinates and participates on the specification writting process. |
There was a problem hiding this comment.
There will also be WGs for each spec or group of specs -- ie. who is on the IPLD Spec WG vs the libp2p spec WG.
There was a problem hiding this comment.
In the same way that the Internationalization Working Group participants are not the ones writing the translations to every language but rather create the platform and coordinate multiple contributors and possible vendors for those translations to exist, the Specifications working group should be more focused on ensuring that specs are written, clear, and that individuals that participate in other working groups and/or projects can effectively contribute to specs (through training, mentoring and so on).
The Specifications working needs to have more of a role of an Editor + Reviewer, create a Peer Review system and structure the spec writing process (templates, coordination, timelines and objectives).
WORKING_GROUPS.md
Outdated
| - Collaborate with with Browser vendors. | ||
| - Identify blockers and design constraints of each integration, figuring out new ways to go around those limitations. | ||
| - Increase the adoption of the Distributed Web by making it easy for users to access it. | ||
|
|
There was a problem hiding this comment.
- Define specifications for address schemes and advocate for those schemes to be adopted
WORKING_GROUPS.md
Outdated
|
|
||
| - Design Importers and Exporters for different data sets | ||
| - Maintain Unixfs | ||
| - Improve the APIs that let users interact with Files on IPFS |
There was a problem hiding this comment.
- inform and guide users who want to define their own IPLD-based data formats
WORKING_GROUPS.md
Outdated
| - Write and make sure that docs are up to date. | ||
| - Create Tutorials and Workshops to teach concepts about IPFS and the Distributed Web. | ||
| - Develop the front facing Webpages of each project (ipfs.io, libp2p.io, multiformats.io and ipld.io) | ||
|
|
There was a problem hiding this comment.
- ensure that public forums (discuss.ipfs.io, irc, github, etc) have adequate moderation and responsive support
- ensure that the code of conduct is followed and provide appropriate support or response when violations occur
|
Excellent feedback everyone! Thank you so much for writing these notes, I'm incorporating some and will push in just a bit, but here are some follow-ups to your comments:
Good point, we might need to define a language to distinguish a team from a working group and have the two classes. I feel here the fundamental question is that there will be endeavors that require different cycles of feedback and lifetime. We can either use different words to distinguish those or we can use always the term Working Group and then tag them with different properties through labels.
Yes, I agree that it is better for us to get really well defined working groups that currently have already people working on them rather than trying to boot all of them at the same time and put a lot of responsibility in few individuals.
Once we are happy with this document and get it merged, we can move to that step, yes :)
Thank you, I feel that it is a fundamental little detail that can improve a ton how we grow and rotate the folks working on this groups.
Yes, for js-ipfs, go-ipfs and the whole project as a whole. Currently, there is a lot of work that falls on the go and js IPFS leads that can be lead by working groups. Once these are formed, js-ipfs and go-ipfs will no longer need to be a team, but rather a collection from different individuals that contribute to the project, while the Releases working group ensures that quality gets merged and released.
I've received some feedback with uncertainty about the value of a Steering Committee in the long term. I still strongly believe that having such Committee will valuable for the long-term vision and facilitation between working groups, however, we can totally live without it for a bit more time and at the same time start formalizing the working groups around it.
Grouping sounds good to me, especially because this list will tend to increase overtime. Nevertheless, I would like everyone to consider a slow start for creating working groups. Let's focus on the ones that are already becoming a thing (Security, Releases, ipfs-cluster, IPFS in Web Browsers) and work with the folks on those working groups to polish our OKR setting process and then create the template for more working groups to start. We should avoid having one person being responsible for many (many === number that exceeds the capacity of an individual to actively manage and set goals for that working group). |
|
Updated the doc with the suggestions made :) |
WORKING_GROUPS.md
Outdated
|
|
||
| #### Release | ||
|
|
||
| The Release Working Group manages the releases for go-ipfs, js-ipfs and its internal dependencies. |
There was a problem hiding this comment.
Do you want this to be one working group or do you want it to be separate release working groups for each language?
There was a problem hiding this comment.
I provided an answer on your previous comment, perhaps it got hidden with the rearranging. Here you go #285 (comment)
There was a problem hiding this comment.
I considered this and I'm confident that releasing JS and Go can be handled by the same team. In the end, the Release team should focus on auditing the release, ensuring that the quality bar didn't lower (check features added, CI, documentation and so on), that communication with the community and partners happens (changelog, post, twitter, reddit, emails and so on).
Yes, there are some people that have more context of what is happening in Go and others in JS, but in the end it is the same protocol and given that this is a Working Group and not a Working individual, we can have 2~4 people that share enough context to be capable of launching the rocket with the same level of quality.
There was a problem hiding this comment.
Adding to the above, there should be some work on a release that involves checking spec compliance and interoperability.
There was a problem hiding this comment.
Defining the release roadmap and deciding what actually goes into a release feels like a responsibility of the dev team for that implementation, since they ultimately know when something is ready to be released or not.
With the rest of the points, I feel like we can fold that into QA, as it got to do with some sort of automated pipelines, testing and feedback cycles.
WORKING_GROUPS.md
Outdated
| - Test releases. | ||
| - Announce and gather feedback from the community on each release. | ||
|
|
||
| #### Specifications |
There was a problem hiding this comment.
Having a generic "Specifications" WG invites the same kind of confusion we have with a generic "javascript team". People work on the specs that are relevant to their work and don't necessarily need to be involved in other specs. Do you intend to have a small "specifications" WG that deals with all specs or is this a large group that includes everyone who has these kinds of responsibilities for any spec?
There was a problem hiding this comment.
Did you get to see my response to your last comment on the Specifications Working Group? Here #285 (comment)
There was a problem hiding this comment.
Reposting it here to make it easier to find:
In the same way that the Internationalization Working Group participants are not the ones writing the translations to every language but rather create the platform and coordinate multiple contributors and possible vendors for those translations to exist, the Specifications working group should be more focused on ensuring that specs are written, clear, and that individuals that participate in other working groups and/or projects can effectively contribute to specs (through training, mentoring and so on).
The Specifications working needs to have more of a role of an Editor + Reviewer, create a Peer Review system and structure the spec writing process (templates, coordination, timelines and objectives).
|
Thing to add to each of these: What are this WG's responsibilities for contributing to the quaterly & annual planning process? What are their contributions to setting each quarter's OKRs? |
|
Shall we schedule a call to identify what is left for us to do so that we can move confidently with Working Group structured planning for next quarter? |
WORKING_GROUPS.md
Outdated
| - Create and monitor continuous testing for the multiple supported platforms (Architecture, OS'es, Browsrs). | ||
| - Explore new runtimes and create Roadmaps towards supporting those. | ||
|
|
||
| #### QA, Testing and Dev Team Enablement |
There was a problem hiding this comment.
@victorbjelkholm did you had the chance to review this PR? Let's sync up on what this means for 2018 quarter OKRs on QA, Testing and Dev Team Enablement :)
|
thanks very much @flyingzumwalt
The IPLD work i have in mind is different in nature than what's listed here -- IPLD can include those pieces, but those pieces (files/dex) are much more on IPFS camp than IPLD. (definitely an interface between the two and could fall in either). The "direct IPLD work" is about:
more notes
I propose:
|
daviddias
added
P1
daviddias
changed the title
|
A lot has happened since these proposal was originally posted (November 2017). I've updated the doc to change the character of it from proposal to actually just
|
WORKING_GROUPS.md
Outdated
| > This Working Group hasn't been formed yet. | ||
|
|
||
| - **Coordination**: https://github.com/ipfs/community | ||
| - **Captain**: ??? |
There was a problem hiding this comment.
I wasn't sure if I would leave this one around or not. The truth is that there is a lot of work done and planned, but we don't have an official Captain for this front.
@mikeal given that you are the person leading most of the new charges, would you like to take this role and plan OKRs for next quarter? Knowing that this is a role that you are also actively looking to fill/replace yourself?
WORKING_GROUPS.md
Outdated
| > This Working Group hasn't been formed yet. | ||
|
|
||
| - **Coordination**: https://github.com/ipfs/docs | ||
| - **Captain**: ??? |
There was a problem hiding this comment.
I was also unsure if I would leave this one around. Again, truth is that there has been work and even structure, but since Q3, we no longer have OKRs for Documentation. @Mr0grog was there a change of direction? Is the Docs front finished or going to be pursued in a completely different way?
There was a problem hiding this comment.
Ugh, sorry for missing this and not replying.
was there a change of direction? Is the Docs front finished or going to be pursued in a completely different way?
I don’t know. The specific projects I was contracted to do are done and I’ll be gone in a week (after, obviously, an absurdly late delivery). I did not do Q3 OKRs because I’m not here for all of Q3 and nobody asked me to, so it seemed like it would be setting up expectations I could not fulfill. I think/hope the thread of this work is going to be picked up by @meiqimichelle and/or @mikeal, but I really can’t speak for future directions.
There was a problem hiding this comment.
Hi all! I just finished writing up my thoughts regarding the docs work here: https://github.com/ipfs/user-research/tree/master/distributed-kit . Long story short, setting up and staffing a structure to make sure @Mr0grog 's work continues would address our biggest usability problem, and our biggest blocker to user success, in the IPFS ecosystem. @mikeal and IPFS leadership (not sure exactly who, for this convo) should connect and align on how to staff next steps on this work.
I know @diasdavid and @flyingzumwalt are short on time right now. Perhaps I'll reach out to @mikeal first to get his thoughts, and then broaden the convo with a proposal rather than waiting for a time when all are available.
daviddias
force-pushed
the
working-groups
branch
3 times, most recently
from
23cd7ec to
2659183
Compare
daviddias
force-pushed
the
working-groups
branch
from
2659183 to
ff37fd8
Compare
|
|
||
| Each Working Group is free to experiment with setting their own pace, tracking work and defining priorities. The only requirements are that the Working Group exposes its focus through OKRs to the rest of the org (common interface), that it assigns a Captain, creates an entry point repo and has at least 2 full time contributors dedicatted to it. | ||
|
|
||
| Each contributor shouldn't carry responsibilities accross multiple working groups. This is not forbidden by any means but it is greatly disencouraged as it will disable the contributor from achieving full focus. |
There was a problem hiding this comment.
This is a bit confusing as what seems to be outlined as Working Groups here, is what I believed "Teams" was before.
One of the OKRs for the Testing WG this quarter is "Have at least one member from each team be a part of the Working Group" which is based on the fact that I thought Working Groups are supposed to have members who can also be a part of other working groups. I think that's the "common" definition of a working group as well, as set of people who can participate in many working groups.
It's also unclear where the "IPFS Core" Working Group would fit into this here, since people activate in that, needs to be active in other working groups as well.
There was a problem hiding this comment.
This is a bit confusing as what seems to be outlined as Working Groups here, is what I believed "Teams" was before.
There hasn't been a Teams definition before.
One of the OKRs for the Testing WG this quarter is "Have at least one member from each team be a part of the Working Group"
In reality, it is still OKRs of the Dev Team Working Group and not complete different contexts
TEAM_STRUCTURES.md
Outdated
| > This Working Group hasn't been formed yet. | ||
|
|
||
| - **Coordination**: https://github.com/ipfs/community | ||
| - **Captain**: ??? |
There was a problem hiding this comment.
Put me down as the Captain here. Is there a doc of best practices for getting a WG up and running?
TEAM_STRUCTURES.md
Outdated
|
|
||
| **Responsibilities include**: | ||
|
|
||
| - Organize Meetups in multiple cities. |
There was a problem hiding this comment.
Can we remove this? The next point on helping community members organize meetups in their own communities covers this.
Alternatively, we could alter this to focus on larger events. "Support IPFS related conferences."
|
I'm going to move ahead and merge this doc as is. As described in #285 (comment), this document changed from proposal to just capture what got developed almost organically over the last 9 months. |
This is my first stab at designing solid working groups that can tackle different verticals of the IPFS project independently. This comes as a follow up of many informal conversations that many of us had.
The wording still has ways to improve and we should expand on the responsibilities list, specially after checking in with whom has been taking more or less those responsibilities for the last year or so.
There is also an open question on what is the life cycle of a working group and if we are going to have captains for each of these.
UPDATE: See #285 (comment)