Skip to content
This repository has been archived by the owner on Feb 8, 2023. It is now read-only.

Sprint: June 20 → July 3 #114

Closed
RichardLitt opened this issue Jun 20, 2016 · 13 comments
Closed

Sprint: June 20 → July 3 #114

RichardLitt opened this issue Jun 20, 2016 · 13 comments

Comments

@RichardLitt
Copy link
Member

RichardLitt commented Jun 20, 2016 •
edited
Loading

Sprint June 20th -- July 3rd

This sprint, instead of having the long 30-minute discussions, we are having another short video sprint. This is because of travel issues and vacation for many endeavors, and because of a tight schedule at the moment for a few people working on IPFS.

For this sprint please do the following:

  • Put your completed tasks in the old sprint issue.
  • Put your To Do items for the upcoming sprint in the Etherpad for the video sprint before the sprint. (I'll ping you about this, if you're in the active contributor list).
  • Finalize your To Do items during the hangout and put them in this issue by Tuesday, June 21st, the Solstice. ☀️

Sprint Discussions

Schedule

Endeavour Lead Time (PDT - UTC/Z - CEST) Pad
Video sync @RichardLitt 11:00PDT 17:00Z 19:00CEST https://public.etherpad-mozilla.org/p/ipfs-june-20-video-sync

@RichardLitt will post the Hangout link on the IRC channel, as normal.

Participants

@whyrusleeping
@diasdavid
@Kubuxu
@lgierth
@nicola
@em-ly
@RichardLitt

Sprint Deliverables

  • Add your tasks below in a comment, that way we only have people listed who are really in the sprint
  • Add links to issues down here. Only add things you can finish this sprint.
  • Add your To Dos in Markdown format. That means using - [ ], which GitHub automatically converts to a checkbox.
  • As the sprint progresses, refer back to this issue and check of any tasks you have done, and subdivide, delete, or add tasks accordingly. That way, the list at the end gives an accurate depiction of what was done this sprint.
  • If anyone seems to be doing something you could do, approach them and see if you can take any of their tasks. Likewise, if you have too much.
@RichardLitt
Copy link
Member Author

RichardLitt commented Jun 20, 2016 •
edited
Loading

@RichardLitt To Do

Adding my preliminary list here, because I basically have it already filled out for the video hangout.

go-ipfs

Blog

name-your-contributors

github-task-status

  • [~] Build badge to check issue lists Tool Request: Issue Status Badges #111
    • Started work at Roadmap RichardLitt/github-task-status#1
    • Enable grabbing items from a list of URLs to check
    • Set up as a server-side task that runs on a cron-job
    • Make a GitHub bot to add issues to a list
    • Enable for sublists
      • Like this one
    • Enable for comments that aren't the first in the issue
    • Check completion of tasks from issues linked in one issue

ipfs-http-api

  • Go through issues and open PRS on ipfs-http-api with fine comb
  • [~] Standardize returns (errors + formats) on API
  • [~] Ship ipfs-http-api (on it)
    • [~] Put up publicly on Apiary
      • Publish on Apiary
      • Talk about it
        • Draft a blog post
        • Link on the website
      • Check in with @diasdavid about using for the clientbase
      • Link @jbenet and @daviddias to talk about automating generation of the CURL requests in the spec

READMEs

@Kubuxu
Copy link
Member

Kubuxu commented Jun 20, 2016 •
edited
Loading

@Kubuxu - To Do

go-ipfs:

Misc

  • - clean up hasteIPFS
  • - refresh Ethereum knowledge
  • - setup HTTP redirect server to global gateways: http://bin.ipfs.ovh -> https://ipfs.io/ipns/bin.ipfs.ovh

@daviddias
Copy link
Member

daviddias commented Jun 20, 2016 •
edited
Loading

David Dias Task List 📝

From the js-ipfs Captain Log

  • ⏫ Finish the remaining items on libp2p Transports + Connection section
  • ⏫ Finish the remaining items on go and js interop
  • ⬆️ interface-ipfs-core
  • ⬆️ Finish the remaining items in libp2p UX
  • ↕️ Isolate tests on js-ipfs - available through npm scripts

Other goals

  • npm on ipfs
    • ↕️ rename registry-mirror to ip-npm
    • ↕️ update the CLI interface (following the discussion, design a more elegant and convenient one) ipfs-npm = registry-mirror.next() ipfs-shipyard/npm-on-ipfs#42
    • ↕️ implement (and test) the new interface
    • ↕️ dogfood my npm needs with registry-mirror (set up a node at home with whole npm on RPI)
  • ⬆️ write a blog post about libp2p and Decentralized Web Summit
  • ⬆️ chat with @whyrusleeping about closing connections in go-libp2p

I'm experimenting a new thing, adding priority to my tasks, so that others now if I'm still super busy with really important priorities or if I'm just doing 'stuff'. This way, if something comes up in the middle (which always comes), I can also tag it with priority and that way, I can evaluate if I did well bumping the priority of something else face of what were the primary needs.

Legend

  • ⏫ Top Priority
  • ⬆️ Next
  • ↕️ Regular

@ghost
Copy link

ghost commented Jun 20, 2016 •
edited by ghost
Loading

@lgierth todo

@nicola
Copy link
Member

nicola commented Jun 20, 2016 •
edited
Loading

nicola's task

TODO

  • do further work on IPLD selectors / IPLD transforms (see roadmap here: Roadmap towards IPLD specs#115)
  • read on pub/sub (possibly clean up gossip-cylcon)
  • write down a structure for a possible IPLD paper
  • write down about family URI (to be discussed)
  • further readings on BFT
  • join the browser conversation

@Kubuxu
Copy link
Member

Kubuxu commented Jun 27, 2016 •
edited
Loading

@Kubuxu - Mid-sprint updates

This week I worked mainly on blockstore performance, bloom filtering and ARC caching of Has calls which should reduce idle hard drive calls by about 95% to 99%. I also updated go-log across the codebase (not without problems) and fixed some misc bugs and added smaller features.

go-ipfs:

Misc

  • - clean up hasteIPFS
  • - refresh Ethereum knowledge
    • - learn new tools for development of contracts
  • - setup HTTP redirect server to global gateways: http://bin.ipfs.ovh -> https://ipfs.io/ipns/bin.ipfs.ovh

@RichardLitt
Copy link
Member Author

Mid-week updates

This week, I wrote around 100 pull requests, standardizing READMEs and licenses across the organization. That sums it up for me.

@whyrusleeping
Copy link
Member

Whyrusleeping mid-week update

the biggest thing i worked on last week was getting the path clean bug resolved. We've decided to base32 encode all binary keys before creating datastore keys with them. This will unblock a bunch of other PRs and fix a few fairly major issues we have.

Beyond that, i've been working on writing providers to disk so as to not keep them in memory. Fixing this will fix the largest cause of memory bloat in long running ipfs daemons.

@daviddias
Copy link
Member

During last week, I've been focused on refactoring our libp2p transports, stream muxer, swarm and libp2p builds, introducing the new Identify protocol (new as in: "working like go-ipfs"). This work was a dependency of getting go-ipfs and js-ipfs interop. I'm currently checking (hoping to finish today) if everything is ok now. Will try to tackle and check everything with @whyrusleeping while we have a overlap work time across timezones.

There were many PR's to achieve this, here is the list of the major ones. They can also be checked at the Captain.log

image

@nginnever
Copy link
Member

voxelot mid-sprint update

This week I worked a bit on js-ipfs add http and reviewed some first steps on js-ipfs-api get requests. Finished setting up a basic UI for a react/redux based distributed collaborative text editor running on orbit-db as well.

  • js-ipfs
    • add http
    • get http
  • js-ipfs-api
    • get
  • ipfs-pad

@daviddias
Copy link
Member

daviddias commented Jul 3, 2016 •
edited
Loading

David Dias Sprint Check In

I'm posting this on Sunday, as I'm not 100% of my availability to be present this Monday. Will be out till the 12th of July.

This sprints was pretty exciting, we got js-ipfs and go-ipfs exchanging blocks with one another after all the wire battles, pretty energising moment. During the process, I've had the change to work a lot in the transport layer, improving our interfaces, implementations, integrations and (something that is very dear to me), adding more tests, enabling that code base to be more robust, without the feeling of 'cutting corners' to reach a quick goal. More on this can be seen on the Mid Week Check in

Nevertheless, it is important to note that in order to get interop, a custom build of go-ipfs was required, simply because of some network assumptions in go-libp2p, namely that secio is assumed by default and that go-libp2p doesn't extract the public keys from the Identify protocol. More on ipfs/kubo#2738 and https://github.com/libp2p/js-libp2p-swarm/issues/78.

Other things I did during this sprint were:

Blockers:

I've had also some very good conversations with members on the IPFS community and some volunteered their free time to contribute to completion of the last entry of the js-ipfs Captain.log, you rock! ❤️. These tasks are:

As soon as I get back, one of my focused goals will be to complete the ipfs-api next generation, it is a dependency of many of our projects and the first door for many people in the community to try out IPFS and currently, it is kind of hard to use due to lack of good documentation, examples and specially, the code is just overly complex. We can make this way much better with ipfs-inactive/js-ipfs-http-client#305. So, if someone wants to take part of it while I'm out, please do, your contributions are always super appreciated :)

@RichardLitt
Copy link
Member Author

@RichardLitt check in

Last week, I worked on the http-api-spec a fair amount, closing issues and opening a few PRs, making sure that it is a bit more up to date. I also wrote a blogpost about it, here that would <3 some review. On the blog side, I had a lot of discussions on how best to move forward on this, and set up Waffle.io to track our blogposts. I also tried to rework the ipfs.pics post, and have some more work to do on that. I worked on the CLI interop a tiny bit for js-ipfs and go-ipfs with @Kubuxu, but didn't have as much time for it as I wanted.

I may have a been a bit overambitious in what I expected to do.

Updated TODOs

go-ipfs

Blog

name-your-contributors

github-task-status

  • [~] Build badge to check issue lists Tool Request: Issue Status Badges #111
    • Started work at Roadmap RichardLitt/github-task-status#1
    • Enable grabbing items from a list of URLs to check
    • Set up as a server-side task that runs on a cron-job
    • Make a GitHub bot to add issues to a list
    • Enable for sublists
      • Like this one
    • Enable for comments that aren't the first in the issue
    • Check completion of tasks from issues linked in one issue

ipfs-http-api

  • Go through issues and open PRS on ipfs-http-api with fine comb
  • Standardize returns (errors + formats) on API
  • Ship ipfs-http-api (on it)
    • Put up publicly on Apiary
      • Publish on Apiary
      • Talk about it
        • Draft a blog post
        • Link on the website
      • Check in with @diasdavid about using for the clientbase
      • Link @jbenet and @daviddias to talk about automating generation of the CURL requests in the spec

READMEs

@Kubuxu
Copy link
Member

Kubuxu commented Jul 4, 2016

@Kubuxu - update
This week at fist glance didn't seem as productive as I wanted it to be, not PR and published code wise but at second glance I see that I've done quite a lot by gaining knowladge and planning ahead. I started working with @RichardLitt on ipfs-mans, I've extracted data form go-ipfs and structured it into YAML as can be seen here. I've helped reviewing many PRs and finalized the bloom filtering of requests to blockstore (it is waiting for final merge ipfs/kubo#2885). I also refreshed my Ethereum and Solidity knowledge, learned about new security advisements and tools.
I experimented with injected interface pattern for Golang, were one package defines interface and other inject implementations of it. I planned to use it for go-log but it is something we should think about more before implementing it but it seems great pattern for metrics reporting.

@jbenet jbenet removed the sprint label Jul 5, 2016
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
7 participants
@jbenet @RichardLitt @daviddias @whyrusleeping @nicola @Kubuxu @nginnever