Sprint F

[jbenet]

Sprint Goals

1. Wire protocol + interop with node
2. FIx all these bugs!

Sprint Participants:

• @whyrusleeping @krl @cryptix @wking @dylanPowers @jbenet

Sprint Deliverables

• Bugs to fix:
  • ipfs object stat should print out the hash kubo#1082 @wking
  • ipfs refs times out kubo#1050 @cryptix
  • disable 'ipfs name publish' while /ipns is mounted kubo#964 @whyrusleeping
  • Gateway broken for IPNS entries kubo#1150 @cryptix
  • Better resolution tests (@jbenet, @whyrusleeping)
  • merge @chriscool's PR (@jbenet)
  • http://godoc.org/net/http#CloseNotifier (@cryptix)
  • make accept use limited concurrency jbenet/go-peerstream#5 (@jbenet)
• webui
  • "raw" button breaks when using ipns paths ipfs-webui#42 (@krl)
  • https://github.com/ipfs/webui/issues/43 (@krl)
• interop with node
  • Protocol wire spec (@jbenet)
  • swapping out the stream muxer (or having multiplex in go) (@jbenet)
  • option (or build) to disable encryption for interop with node (@whyrusleeping)
• perf/bugs:
  • change the memory profiling (@whyrusleeping)
  • save memory profile + binary to ipfs (@whyrusleeping)
• Specs / scoping out
  • DHT spec (@jbenet)
  • Keystore spec (@jbenet, @whyrusleeping)
• Blog
  • make blog (@jbenet)
  • prep hello world post (@jbenet)
  • add blog to website (@jbenet)

backlog

• github.com/krl/ipfs-geoip - we need to add attribution for the GeoLite data in the webui as well.
• /files with thousands of files added makes my browser hang ipfs-webui#20
• Update readme assets + tests kubo#1179 (waiting for archive)
• add a 'local' tab ipfs-webui#41 (needs better definition on what we want to do with files (without names) in the webui)
• Test Suite ipfs-webui#38 (involved)
• make it easy to work with gateways
  • remove docker, init script, auto-rebooting (@jbenet)
  • cpu + mem profiling on gateways (@whyrusleeping)
  • change nginx to do fallback (not round robin) (@jbenet)
• make all adds add to an object representing local state
• scope out init + daemon sprint (@jbenet)
• reimplement goprocess without mem leak (@whyrusleeping)
• Address memory leak (likely culprit is goprocess)
• json encoding for add (@travisperson)
• specs:
  • shell / core api
  • scoping out how signed binaries + autoupdate will work (@jbenet)
  • scope out how testing will be done on both webui and api (@travisperson, @jbenet)

[cryptix]

Fixed as well:

• godeps: update metrics (freebsd build fix) kubo#1202
• godeps: rewrite import path kubo#1209

[jbenet]

Merged many PRs this week, so far:

• Improve parse_test.go kubo#1193
• increase iptb test stability kubo#1194
• make comments match the code in secure channel implementation kubo#1198
• mild refactor of bitswap kubo#1199
• correct the blocksize calculation for link blocks kubo#1200
• godeps: update metrics (freebsd build fix) kubo#1202
• ipnsfs/system: Fix 'retreive' -> 'retrieve' typo kubo#1205
• add link to community contributing #1088 kubo#1206
• godeps: rewrite import path kubo#1209
• core: add context.Context param to core.Resolve() kubo#1189
• path/resolver: Fix recursive path resolution kubo#1212 (important hotfix, thanks @wking!)
• linting for jsx files ipfs-webui#47
• make full path default in object view (fixes #42) ipfs-webui#45
• Make geoip requests over ipfs ipfs-webui#44
• add dht/findprovs and name/resolve ipfs-inactive/js-ipfs-http-client#15

[cryptix]

The putHandler of the writable gateway ate a lot of time.. In oder to fix ipfs/kubo#1150 now, we decided to disable it and it's sharness tests (t0111-writeable-gateway.sh) and fix the PUT method next week.

[jbenet]

I also cut the 0.3.4 release, cause it was time pressing the get these bugs fixed for users in a way they could feel sure about the fix (ipfs version). 

Will be updating changelog later today (unless someone beats me to it).

[jbenet]

hey all -- would be great if you could:

• list all the thins that got fixed / worked on this week
• take a look at the assigned things and see where they are (maybe write any needed notes here)
• comment about the things you'd like to see pulled into the next sprint(s).

we can use this as a way to pre-discuss the status, and the things desired things for the next sprint.

cc @whyrusleeping @cryptix @tv42 @krl @travisperson @dylanPowers

[cryptix]

I got hung up on the PUT handler while working on ipfs/kubo#1191 and didn't make any meaningful progress on the others I wanted to work on.

I commented on ipfs/kubo#1050 about my findings and suggestions for proceeding with it.

The http.CloseNotifier addition is fairly easy and I'd like to pull it over to next week.

edit: links and stuff.

[whyrusleeping]

I don't recall everything I got done, but I diagnosed a bitswap perf issue and am working on a significant refactor to address the problem, low bandwidth this week with moving out.

[jbenet]

• published blog: http://ipfs.io/blog (maybe it should be blog.ipfs.io ?? and use CNAME trick?)
• Made several updates to website (added examples, added blog, ...) and published new version

[tv42]

I've been redoing the pinning so that it stores data in IPFS objects. I have the first version of that about working, but need to do a series of commits first to clean up pinning bugs and API warts, and then make the relevant sharness tests be less explicit as they used to dictate behavior not results. I don't expect the actual code to change much for this iteration, anymore.

[jbenet]

Participants of Sprint F: could you look over the unchecked boxes above and point out anything that should be checked? (i'll go through these with people when we check in, too). I'll close the issue after that.

[jbenet]

Addressed the handshake / listener issue in ipfs/kubo#1227