OKRs - 2019 Q2 GUI

[olizilla]

It's planning time. Well, actually, we jumped the gun, met up, and sketched out a proposal plan already, but you get they idea. This PR is where we can surface our thinking and get feedback from others before commiting. Please comment if you have questions or suggestions.

TODO:

• Score your 2019 Q1 OKRs with End of the Quarter Scores
• Q1 Retrospective
• Sync Brainstorming
• Move 2019 Q2 OKRs to 2019 Q2 OKRs Spreadsheet
  • draft ideas for Q2 projects, using the central priorities agreed on here as reference

Parent issue: #902

[olizilla]

The current plan is to make IPFS Camp our main focus for Q2. All working groups need to create content for it, but we'll take responsibility for back filling missing foundational material, and ensuring that the content is framed in a useful, comprehensible way. We'd also like to work on improvements to proto.school and deploying static sites to IPFS.

The proposed OKRs are

Bring new contributors into the IPFS Community and ecosystem

• We create 2 new tutorials for protoschool before IPFS Camp
• Protoschool gets landing page for each tutorial and open ux issues are addressed before IPFS camp - see protoschool OKRs
• docs.ipfs.io and ipfs.io has clear install instructions with one-step installation before IPFS camp - Install Section on IPFS Homepage ipfs-inactive/website#299
• IPFS Desktop provides ipfs command on user PATH - Provide ipfs command line tools on install ipfs-desktop#727
• IPFS Web UI and desktop get in app help that links out to docs and proto.school - Unobtrusive help system to onboard new users ipfs-webui#979
• We ship a cli based service for pinning and updating websites on IPFS and https - Publishing websites to IPFS as a service and github app ipfs-gui#77
• We ship a github app for continuous deloyment of websites from github - Publishing websites to IPFS as a service and github app ipfs-gui#77

We know who our audience is

• Define the metrics that we need to evaluate and improve our decisions
• ipfs.io updated with up to 6 specific use cases and analytics flows that tell us which are popular

IPFS and its apps are a joy to use

• webui and desktop get end-to-end tests for adding files
• style.ipfs.io is set up as a functional style resource, and populated with a kit of basic elements
• All PL sites are auto-deployed when a PR is merged to master - https://github.com/protocol/infra/issues/454

[olizilla]

There are items in the IPFS top level objectives that @momack2 suggests would be a good fit for the gui team

• KR: Ship ipfs.green. A single dashboard that presents the user a table with things that IPFS does verifiably (benchmarks, tests, etc) and what it doesn’t
• KR: ipfs filestore should support easy interfaces for syncing/updating directories

[olizilla]

ipfs.green would be fun (i'm assuming a reference to https://node.green/) we should unpack who it's for, and what the desired outcome is, so we can evaluate whether it's the best way to achieve that goal right now. E.g. Would improving docs.ipfs.io more generally be a better use of time than creating a new thing?

ipfs filestore should support easy interfaces for syncing/updating directories

I dont think we have the right skills to tackle this as a team right now. We can review api proposals, but I imagine the work would be done by the the core js-ipfs and go-ipfs teams.

[momack2]

Fair - sounds like the filestore work is more core-protocol instead of UI-layer.

My suggestions come from trying to reconcile working group priorities with project priorities. While not all groups have a giant backlog to support the package managers priority (and will compensate for those spending 95% of their resources just focused on improving usability and performance for that group), I was surprised that there aren't any items to support the P0 objective for the project. Did you get a chance to brainstorm with @achingbrain and @andrew at all about potential opportunities.

The ipfs.green suggestion came from that motivation - a consumable UI to help package managers (and others evaluating IPFS) answer questions about performance, capabilities, and scalability. It would also give us a way to visualize improvements and track future work. It's just one suggestion though - there are likely others that are better!

[andrew]

The other potential gui+package manager project we talked about when I went over to London a couple weeks ago was a desktop app and UI for the npm-on-ipfs client but after the brainstorming in Lisbon it seemed like the GUI team had their hands full with IPFS camp so I suggested pushing that back to Q3, might be worth having another chat about that to see if it can be re-prioritized?

[olizilla]

@momack2 totally understand that! As @andrew mentions, we discussed an npm-on-ipfs desktop app. It's a fun idea, but at the time felt like it might be somewhat frivoulous / a demo rather than really pushing things forward... but we should unpack it some more, perhaps it could really help npm-on-ipfs by encouraging all users to share back what they fetch... I've created an issue for us to figure it out here: ipfs-shipyard/npm-on-ipfs#94

[olizilla]

I've tweaked our OKRs to lower the priority of the "service for publishing websites" and add npm-on-ipfs as a P0... see the sheet for priorities

Bring new contributors into the IPFS Community and ecosystem

• Get 1000 users actively re-hosting modules via an npm-on-ipfs app - npm-on-ipfs: the desktop app! A long lived client to encourage co-hosting ipfs-shipyard/npm-on-ipfs#94
• We create 2 new tutorials for protoschool before IPFS Camp
• Protoschool gets landing page for each tutorial and open ux issues are addressed before IPFS camp - see protoschool OKRs
• docs.ipfs.io and ipfs.io has clear install instructions with one-step installation before IPFS camp - Install Section on IPFS Homepage ipfs-inactive/website#299
• IPFS Desktop provides ipfs command on user PATH - Provide ipfs command line tools on install ipfs-desktop#727
• IPFS Web UI and desktop get in app help that links out to docs and proto.school - Unobtrusive help system to onboard new users ipfs-webui#979

IPFS and its apps are a joy to use

• webui and desktop get end-to-end tests for adding files
• style.ipfs.io is set up as a functional style resource, and populated with a kit of basic elements
• All PL sites are auto-deployed when a PR is merged to master - https://github.com/protocol/infra/issues/454
• We ship a cli based service for pinning and updating websites on IPFS and https - Publishing websites to IPFS as a service and github app ipfs-gui#77
• We ship a github app for continuous deloyment of websites from github - Publishing websites to IPFS as a service and github app ipfs-gui#77

We know who our audience is

• ipfs.io updated with up to 6 specific use cases and analytics flows that tell us which are popular
• Define the metrics that we need to evaluate and improve our decisions

[achingbrain]

Get 1000 users actively re-hosting modules via an npm-on-ipfs app

Sounds good!

[meiqimichelle]

ipfs.io updated with up to 6 specific use cases and analytics flows that tell us which are popular

Team Cluster has a solid start on this for use cases that are built on top of IPFS + IPFS Cluster. Could we work together to get those on ipfs.io ?

[hsanjuan]

hey, what happened to the cluster tab in the GUI effort? It would be really nice to get that. We're writing a Swagger Spec for the API and we expect most of the JS client code to be simply generated from there...

[olizilla]

@hsanjuan still up for working on it! Do you have an issue for collecting thoughts about what info and actions would be useful to offer the user? We're unlikely to get to it this quarter given the current list, but if it's a high priority for cluster, we could drop something in favour of that work.

[olizilla]

@meiqimichelle for sure!

[hsanjuan]

@hsanjuan still up for working on it! Do you have an issue for collecting thoughts about what info and actions would be useful to offer the user? We're unlikely to get to it this quarter given the current list, but if it's a high priority for cluster, we could drop something in favour of that work.

Maybe @pkafei or @meiqimichelle can work on this?