From 06940a40b7eb4b10ccefffd82f0ca0bf82926373 Mon Sep 17 00:00:00 2001 From: Volker Mische Date: Tue, 13 Aug 2019 00:52:53 +0200 Subject: [PATCH] Meeting notes 2019-08-12 This meeting was about going over all issues and PRs of the spec repo. --- meeting-notes/2019/2019-08-12--ipld-sync.md | 138 ++++++++++++++++++++ 1 file changed, 138 insertions(+) create mode 100644 meeting-notes/2019/2019-08-12--ipld-sync.md diff --git a/meeting-notes/2019/2019-08-12--ipld-sync.md b/meeting-notes/2019/2019-08-12--ipld-sync.md new file mode 100644 index 0000000..e81832d --- /dev/null +++ b/meeting-notes/2019/2019-08-12--ipld-sync.md @@ -0,0 +1,138 @@ +# 🖧 IPLD Weekly Sync 🙌🏽 2019-08-12 + +- **Lead:** @vmx +- **Notetaker:** @rvagg +- **Attendees:** + - @vmx + - @warpfork + - @rvagg + - @mikeal + - @stebalien +- **Recording:** https://youtu.be/LC1oLYL9k1k + + +## Agenda + +- General + - Start recording + - Start live stream + - Find a notetaker + - Ask everyone to put their name into the list of attendees + - Ask everyone to put their items that they've been working on the past two weeks (should be done prior to the meeting) + - Ask for last minute agenda items +- This meeting + - Go through open PRs in https://github.com/ipld/specs/ repo + - _add your agenda item_ + + +## Notes + + + + +### Weekly Update + +@vmx + - Getting js-ipfs upgraded to latest js-ipld which uses the latest ipfs-repo release which is async/await based + +@warpfork + - most of the selector stuff from hannah has landed on go-ipld-prime master :tada: + - i wrote a small book about performance: https://github.com/ipld/go-ipld-prime/commit/d0ce3ded4aaa49b4d047b79a5c2ddc03f06d7f80 + - clearing out a bunch of naming refactors that have been buffered until an opportune moment... + - https://github.com/ipld/go-ipld-prime/pull/27 + - https://github.com/ipld/go-ipld-prime/pull/28 + - merged all of the specs PRs that were uncontentious (which was many -- yayy) + +@rvagg + - More Go, plus some Go/JS compat work https://github.com/rvagg/js-ds-zipcar and https://github.com/rvagg/go-ds-zipcar for test fixture data. + - Fixing serialization bug for dag-pb https://github.com/ipld/js-ipld-dag-pb/pull/156 + - Minor adventures in simplified browser testing + +@mikeal + - Preparing for and participating in meetings: + - UnixFSv1 and UnixFSv2 + - First project advisory meeting, updated on current status and projects, new forum for getting organizational things dealt with - bubble up questions to this meeting via Mikeal (what we're struggling with, how to message things, things where we would benefit from external input) + +### Spec repo pruning + +Dove head-first into the spec repo to try and clean up issues & PRs. + +Issues: + - #49 - and earlier we are closing due to staleness. + - #55 - dag-pb linking, @vmx to pull it in to dag-pb spec work and close + - #56 - primitives - close + - #57 - type encoding - close + - #58 - non-string keys - close this one with reference to #59. + - #59 - need to write up that we only support string keys in map kinds and _why_. + - #65 - closed + - #67 - path encoding - archive this to design history @vmx + - #71 - type convensions - archive to design-history @vmx + - #76 - compression - archive to design-history - @vmx to reach out to user at IPFS camp that had an interesting use-case for this + - #77 - dag docs - @mikeal to fix this one by putting in a description of a DAG, and close it out + - #80 - cbor numbers - @vmx working on cbor spec so will handle this + - #81 - lowercase filenames - close + - #83 - links - leave open, we have a links doc TODO + - #84 - slice paths - needs to go into path spec + - #86 - action items from a meeting - can be closed + - #94 - selectors exploration - archive to design-history @vmx + - #98 - ipld ietf name - leave open + - #101 - replication protocol - to design history @vmx + - #102 - mime type iana registration - leave open but @mikeal to add some notes to it with current thoughts + - #103 - selectors spec naming - close as things have move on + - #105 - recursive selector types - close + - #106 - nested recursive traversal - close + - #107 - sequential selector - close + - #108 - dag-json whitespace - @mikeal to work on this + - #118 - schemas & advanced layout - @vmx to archive + - #119 - design rules - closing + - #120 - graphsync query dedupe - @vmx to archive + - #130 - advanced data structure "criticial decisions" - @vmx to archive + - #140 - schema questions - answered, closed + - #144 - reconcilling terminology - close but @mikeal to put a note on current status and future progress (via his work on UnixFSv2 coming up - working toward specs describing what these things are) + - #158 - drop "kind" from adv layout - active discussion, leaving open + +PRs: + - #69 - go-car links - close - we might benefit from a "related projects" link pileup though + - #104 - selectors - close with prejudice - it's already been folded into other things + - #109 - why's hamt - probably close, probably move to historical report document + - #111 - graphsync metadata - ping hannah to check -- decision unclear + - #114 & #115 - dag-pb spec - @vmx to extract/merge the important pieces and close (also pull in discussion in #55) + - #128 - cardinality doc move - close - moved reorg to action items + - #129 - new dag-json format - create an archive of the discussion and put in design-history (@vmx to archive) + - #131 - HashMap spec - merged + - #137 - AMT spec for sparse arrays for Filecoin, no known implementation - put "experimental" at the top and merge it into design/ appended with a letter. + - #138 - Vector spec - @rvagg to take care of outstanding comment(s) and merge + - #153 - linking in foundations doc - ended up being more controversial than anticipated - @warpfork to close and split it into multiple future PRs attacking the same things but separatelyy + - #159 - capture histor of graphsync A spec proposal and discussion, merged + - #160 - capture histor of graphsync C spec proposal and discussion, merged + +### Other Actions + +- Move 'concepts' into 'design/concepts' +- Up-one everything in the 'schema-layer/' directory + +### Discussion later... + +When we park something into a design history document... + +- do we want to immortalize all the comments on github? +- or is it better to write a terse summary, and then sufficient to just link to original PR/issues? +- (in other words, are we archiving in defense of github going down? or just building summaries and link indexes in-repo?) + +Graphsync... + +- is it possible to move graphsync specs to graphsync repo? + - warpfork: istm graphsync is "ipld ecosystem" moreso than core... and it's a repo under the ipld org, which is close enough. + - weakly held opinion. + +Organization overall... + +- let's bat around other dir organizations. (let's do this as the *whole* thing at once, not single moves.) + - here's one go: https://gist.github.com/warpfork/104c8bca42b87704c3b9a41cf1c690f2 + +Internal messaging vs external messaging... + +- there's a ["book" repo](https://github.com/ipld/book) that might be a place for external messaging. + - this might be useful so when we *do* get something to the point we want external messaging, that external facing doc gets a clean start +- warpfork: thinks that sounds nice but is slightly dubious we'll have the energy to maintain two sets of messaging... +