<daviddias>
now it is karma that doesn't love me "Peer karma-webpack@1.7.0 wants webpack@^1.4.0"
<daviddias>
<dignifiedquire>
hahahah
corruptednode has joined #ipfs
<daviddias>
any insight from the Karma maintainer? Will the packages will be only updated when WebPack officially releases? Do you keep close contact with the WebPack team, are there any timelines?
corruptednode has quit [Ping timeout: 276 seconds]
<dignifiedquire>
are you trying to use webpack@2 ?
<dignifiedquire>
or where is the issue you are having coming from?
<daviddias>
it is when I try to install latest WebPack with my node-libs-browser update
<dignifiedquire>
as far as I know there are still releases happening on the 1.x branch as 2.x is still in beta
<dignifiedquire>
so the karma plugin will probably only be updated when it's out of beta
<richardlitt>
I have no idea how to use this thing
reit has joined #ipfs
<richardlitt>
Well, I think I figured it out. It seems super brittle, though
voxelot has joined #ipfs
reit has quit [Client Quit]
Encrypt has quit [Quit: Lunch time!]
zorglub27 has joined #ipfs
patcon has quit [Ping timeout: 245 seconds]
gaboose has joined #ipfs
Oatmeal has joined #ipfs
voxelot has quit [Ping timeout: 240 seconds]
yellowsir1 has quit [Quit: Leaving.]
<Kubuxu>
richardlitt: It is. It has multiple modes of operation but only two configuration slots.
edubai__ has joined #ipfs
<richardlitt>
Kubuxu: huh. Cool.
<Kubuxu>
richardlitt: So you can for example make it have U2F (for two factor) in first slot and PGP in second. If GitHub wants to do two factor just touch/tap the device. If you want to sign a message with GPG or login into SSH you touch for 2 seconds.
elima_ has quit [Ping timeout: 276 seconds]
<ipfsbot>
[js-ipfs] diasdavid pushed 1 new commit to feat/blocks: https://git.io/vzieV
<ipfsbot>
js-ipfs/feat/blocks 8a5d52d David Dias: check
m0ns00nfup has quit [Quit: undefined]
Encrypt has joined #ipfs
elima_ has joined #ipfs
<daviddias>
dignifiedquire: <3<3<3
<daviddias>
I was chasing getting brfs to work
<dignifiedquire>
:)
<daviddias>
but having brfs was crashing dependencies of fs-blob-store
<daviddias>
oh wait, still doesn't do the trick :(
<dignifiedquire>
damn :(
<daviddias>
oh wait
<daviddias>
got it :D
<daviddias>
YEES
<dignifiedquire>
lol
<dignifiedquire>
leer10 has quit [Remote host closed the connection]
jaboja has joined #ipfs
arpu has joined #ipfs
jaboja has quit [Remote host closed the connection]
gaboose has quit [Remote host closed the connection]
mildred has quit [Quit: Leaving.]
mildred has joined #ipfs
<grncdr>
question about hosting HTML pages on the gateway... would I get any performance benefit by always pointing to /ipfs/$hash instead of /ipfs/$parent_hash/some-name ?
<grncdr>
also: I'm finding it really hard to get larger pages (e.g. many resources) to load through gateway.ipfs.io
ulrichard has quit [Read error: Connection reset by peer]
ashark has joined #ipfs
<patagonicus>
There's a small benefit as there's one object less to fetch, but I doubt it'll make much of a difference.
ulrichard has joined #ipfs
<grncdr>
patagonicus: any idea how small? I've got paths up to 6-7 levels deep I think
ulrichard has quit [Read error: Connection reset by peer]
<patagonicus>
Well, that cost is only paid the first time any object is fetched, as the parent hash will be cached. But I don't know how much it is. I'd go for easier/better understandable directory structure over micro optimizations.
ulrichard has joined #ipfs
Aleks_ has joined #ipfs
ulrichard has quit [Read error: Connection reset by peer]
<ipfsbot>
[js-ipfs] diasdavid pushed 1 new commit to feat/blocks: https://git.io/vziZQ
<ipfsbot>
js-ipfs/feat/blocks 5e3ca3a David Dias: only missing local-storage-blob-store being able to store binary data
ulrichard has joined #ipfs
<Aleks_>
Hi everyone, we're a small hackerspace located in Strasbourg and in March we are planning a conference to present some cool projects/technologies/initiative regarding the future of internet and in particular decentralized/distributed technologies
<Aleks_>
And it would be awesome to have a (french) speaker for IPFS
<Aleks_>
Dyou know anyone I could contact ? :P
<achin>
Aleks_: i think you are in the right place!
<grncdr>
is anybody running a public pin service these days?
<Aleks_>
achin: you mean I could do a PR to ask for a speaker volunteer ? :P
<grncdr>
gateway.ipfs.io is routinely taking 2 minutes to perform 13 requests and transfer 500kb
<Aleks_>
(or an issue)
<achin>
Aleks_: i think that opening an issue to discuss your conference (and also ask for speakers) seems like the right thing to do
<Aleks_>
okay, that sound cool, will do this
<achin>
ipfs doesn't really have a mailing list
<achin>
Aleks_: the core team will all be in this IRC channel later today for a weekly sprint syncup. that would be a good time to catch them to ask for their advice
<Aleks_>
souds great
neurosis12 has joined #ipfs
<ipfsbot>
[js-ipfs] diasdavid pushed 1 new commit to feat/blocks: https://git.io/vziWN
<ipfsbot>
js-ipfs/feat/blocks 57f309a David Dias: update local-storage-blob-store
compleatang has quit [Quit: Leaving.]
compleatang has joined #ipfs
computerfreak has quit [Quit: Leaving.]
voxelot has joined #ipfs
voxelot has joined #ipfs
voxelot has quit [Changing host]
boxxa has joined #ipfs
<richardlitt>
dignifiedquire: where is dist currently hosted?
<richardlitt>
Aleks_: achin is right! :D
<richardlitt>
Opening in ipfs/community is definitely the thing to do
<Aleks_>
Doing it right now :P
ulrichard has quit [Quit: Ex-Chat]
jamie_k_ has joined #ipfs
adamc1999 has joined #ipfs
<dignifiedquire>
richardlitt: on ipfs
<richardlitt>
dignifiedquire: dist.ipfs.io doesn't seem to be working for me :/
Tv` has joined #ipfs
_ylp_ has quit [Quit: Leaving.]
<dignifiedquire>
richardlitt: there were some issues with ipfs.io this morning, maybe related ::cc ping jbenet lgierth
<dignifiedquire>
yep down for me as well
Aleks_ has quit [Remote host closed the connection]
dlight has quit [Ping timeout: 250 seconds]
<ipfsbot>
[js-ipfs] diasdavid pushed 1 new commit to feat/blocks: https://git.io/vziuY
<ipfsbot>
js-ipfs/feat/blocks b575bec David Dias: skip browser for now
<richardlitt>
So... anyone else going to drop their To Do's in there?
<lgierth>
yep me typing right now
<noffle>
richardlitt: working on it :)
ashark has quit [Ping timeout: 276 seconds]
<richardlitt>
ok
<richardlitt>
Well, I think we should probably start
<dignifiedquire>
alright typing now
<whyrusleeping>
woohoo! i made it up before jbenet did!
<dignifiedquire>
whyrusleeping: nice job
<jbenet>
whyrusleeping Nope I'm here
<dignifiedquire>
whyrusleeping: jbenet is probably still trying to figure out his mic :D
<richardlitt>
haha
<whyrusleeping>
lol
<vaelys>
mics are pretty bad.
<richardlitt>
ok
<richardlitt>
Well, shall I start this, guys?
<noffle>
yeah!
<noffle>
folks*
<richardlitt>
(Btw: at a cafe in Reykjavík right now, talking to a guy in the Pirate Party)
<whyrusleeping>
lol, nobody is in the same timezone as last sprint
<richardlitt>
(He instantly got what IPFS is for)
<richardlitt>
Ok. Here comes my update
<richardlitt>
### This Week
<richardlitt>
- [x] Sprint Management
<richardlitt>
- [~] Write the weekly roundup
<richardlitt>
- [~] IPFS API
<richardlitt>
- [x] Split into current functions `master` branch and `expected` spec branch.
<richardlitt>
- [x] Add a link to the api from the specs repo. PR specs/pull/69
<richardlitt>
- [ ] Add `curl` commands to all method descriptions
<richardlitt>
- [~] Check all responses have maximal entries (done for `ipfs add`)
<richardlitt>
- [ ] Log any errors in go-ipfs
<richardlitt>
- [ ] Standardize formatting for `ipfs --help` entries, basing on 0.4
<richardlitt>
- [~] PR docs-styleguide.md links where needed.
<richardlitt>
- [x] Open an issue about sharness testing. http-api-spec/issues/23
<richardlitt>
- [x] Added docs-styleguide.md to ifps/community.
<richardlitt>
- [ ] Open a feature issue about having multiple language implementations for curl requests
<richardlitt>
## Extra
<richardlitt>
- [x] CR+M project-repos/pull/36
<richardlitt>
- [x] CR+M js-ipfs-api/pull/191
<richardlitt>
- [x] CR+M project-repos/pull/37
<richardlitt>
- [x] CR go-ipfs/pull/2231
<richardlitt>
- [x] Added 'needs review', 'difficulty:*' labels to every repo. community/issues/94, PR community/pulls/95, community/pulls/96
<richardlitt>
- [x] Created RichardLitt/gh-description to help with IPFS standardization, eventually.
<richardlitt>
- [x] Added log to go-ipfs about C-c. PR go-ipfs/pull/2243
<richardlitt>
- [x] Capitalized and period-appended logs in go-ipfs. PR go-ipfs/pull/2244
<richardlitt>
EOF
<richardlitt>
KNow what I'm doing for the IPFS API, but haven't found much time recently due to travel. Working on it.
<daviddias>
ahaha good point, if there was a sprint where everyone could be confused with timezones, this is the one
<richardlitt>
jbenet: I'm not sure what to write about for Distributions, but will work on a few paragraphs in the next hour.
<jbenet>
richardlitt: the weekly is prob higher prio
<richardlitt>
jbenet: so, that is really hard.
<richardlitt>
It's really difficult for me as one person to understand the best highlights
<richardlitt>
I'm trying, though.
<daviddias>
richardlitt: let me know if you stumble upon any other obstacle during API documentation, it was great to get some of that work done and pave the path at Svalbard :)
<richardlitt>
daviddias: agreed! And will do.
<whyrusleeping>
richardlitt: lets me and you sit down and talk about highlights
<whyrusleeping>
either today or tomorrow
<kyledrake>
richardlitt if it's Smari tell him I said hi
<richardlitt>
whyrusleeping: sounds good. Want to do that directly after this sync, but before the other talks?
<whyrusleeping>
uhm, sure. that works
<daviddias>
kyledrake is ALIVE :D
<whyrusleeping>
whoa
<daviddias>
we missed you :D
<richardlitt>
kyledrake: it isn't, but the guy said he'd pass that along
<richardlitt>
ok
<richardlitt>
noffle: next?
<noffle>
yep. incoming...
<noffle>
As expected, ArcticJS and being sick kept me pretty busy! No blockers.
<noffle>
richardlitt: I was sick the whole time we were in svalbard ;)
<richardlitt>
noffle: :(
<richardlitt>
I got sick after leaving. Yesterday was just curled up in a ball all day.
<noffle>
maybe my fault
<noffle>
anyways: just doing a full build of distributions now /w ipget included
<richardlitt>
cool.
<richardlitt>
Maybe ping whyrusleeping on that arm build issue?
<richardlitt>
As he used a Chromebook with ARM processor for IPFS for a while
<whyrusleeping>
<3 arm
<noffle>
whyrusleeping: whew, that sounds painful :)
<whyrusleeping>
it was wonderful
<whyrusleeping>
i would actually do it again
Nickfost has quit [Ping timeout: 252 seconds]
<noffle>
richardlitt: that's all for me
<whyrusleeping>
although the new macbooks would provide a similar experience i beleive
<richardlitt>
Cool.
<whyrusleeping>
i have a cat on my hands on my keyboard
<jbenet>
richardlitt: i can suggest things to highlight but i cant keep doing that much of it every week, so want to cause you or "the maintainer(s) of the weekly" to develop a strong sense for it
<jbenet>
(but yeah i can post some)
<richardlitt>
next!
<richardlitt>
jbenet: yep I understand. It's a learning process, and this week I learned a few things _not_ to do.
<richardlitt>
Could've been avoided with better planning.
<richardlitt>
I think having a PR that lasts throughout the week is a smart way; anyone who has anything can add it to the issue or PR, and we can move forward that way.
<richardlitt>
whyrusleeping: You're next.
<richardlitt>
noffle: Anything you want in the weekly?
<daviddias>
daviddias@biham:~$ ipfs files ls / --api=/ip4/127.0.0.1/tcp/6000 | wc -l
<daviddias>
1
<dignifiedquire>
1??
<daviddias>
it seems that it lost all the other 35000++ modules that it already had added
<lgierth>
1 == /npm-registry ?
<daviddias>
nvm
<lgierth>
anyhow, next?
<whyrusleeping>
yeah, 1 == npm-registry
<dignifiedquire>
lol
<whyrusleeping>
hodor
<richardlitt>
ok
<daviddias>
daviddias@biham:~$ ipfs files ls /npm-registry --api=/ip4/127.0.0.1/tcp/6000 | wc -l
<daviddias>
19845
<richardlitt>
lgierth: you're next
<daviddias>
did you had to delete the old repo?
<lgierth>
c&p inc
<lgierth>
Really messy sprint on my end, but the dev040 work is largely done, and provsn works nicely. jbenet had to clean up after me today, and wrote a post mortem.
<lgierth>
- dev040
<lgierth>
- [x] deploy dist.ipfs.io
<lgierth>
- [x] dev040 gateways/bootstrappers
<lgierth>
- :star: that means all of ipfs.io is now served by both v03x and v04x, whatever responds faster. v03x.ipfs.io and v04x.ipfs.io themselves are exempt from this.
<lgierth>
- [ ] dev040 pinbot
<lgierth>
- [x] CR ipfs/go-ipfs#2191
<lgierth>
- infrastructure
<lgierth>
- [x] use provsn for deploying ipfs
<lgierth>
- more about provsn in the coming days
<lgierth>
- [ ] ipv6
<lgierth>
- [ ] wildcard cert
<lgierth>
- [ ] wget benchmark
<lgierth>
about dev040 gateways/bootstrappers, there's a tiny bug in multireq that prevents this right now, will fix tonight
<lgierth>
thanks for multireq whyrusleeping <3
<whyrusleeping>
lgierth: i was too tired to comprehend how that bug was affecting us
<whyrusleeping>
post mortem?
<richardlitt>
lgierth: where's the post mortem?
<lgierth>
the redirect thing? foo needs to be redirected to foo/ if it's a unixfs-dir, so that the dir-index can build correct backlinks
<richardlitt>
we should be doing more of those! +1 to that
<lgierth>
post mortem in ops-private.git, let me know if you don't have access
<kyledrake>
lgierth was that you I saw on cjdns #ipfs?
<lgierth>
yeah i'm larsg there
<parkan>
am I correct in thinking that gateway.ipfs.io will not serve objects published from 0.4.0-dev?
nicolagreco has joined #ipfs
<lgierth>
parkan: it will
ashark has joined #ipfs
<lgierth>
the v03x/v04x multiplexing is buggy right now, but we'll get that fixed tonight
<jbenet>
lgierth: good stuff with getting that dual gateway thing to work. the biggest thing holding back 0.4.0
<daviddias>
- [x] Deliver a talk on "JavaScript for Higher Education" at ArcticJS
<daviddias>
EOF
<dignifiedquire>
daviddias: do you know about the issue that there is a fixed limit of localstorage in chrome?
<parkan>
there are some hashes that I can repin from cubepin but not access through the gateway, was wondering if version API difference was the reason
<dignifiedquire>
with no way to increase it
<lgierth>
yay libutp
<daviddias>
I know, localStorage gives different quotas depending on the browser
<lgierth>
haha so much goodness
<lgierth>
s/libutp/libp2p/
<daviddias>
and depending on the machine storage capacity
<lgierth>
npm! go-ipfs interop!
<dignifiedquire>
so why not use indexdb? it allows you to increase the storage by asking the user
<daviddias>
lgierth: yeah, it was pretty great to get the MerkleDAG objs from go-ipfs up the way in js-ipfs
<daviddias>
dignifiedquire: we can try both, we hit some issues with indexedDB because the writes were not being flushed, even after the end callback was called
<daviddias>
which created a bunch of annoying race conditions
<jbenet>
parkan: not sure about that, can you pm me with examples? i can try taking a look
<dignifiedquire>
daviddias: interesting, can you point me to the issues where this is documented, will try to take a loo
<dignifiedquire>
*k
<jbenet>
daviddias: that's great to hear! :)
<daviddias>
dignifiedquire: the way IPFS Repo is implemented, it supports both with a swap :)
<dignifiedquire>
<jbenet>
daviddias: yay for repo compat, that's great
<daviddias>
dignifiedquire: the discussion happened half here https://github.com/ipfs/js-ipfs-repo/pull/26 half in person during ArticJS, to get the tests failing due to racing conditions, swap local-storage-blob-store with idb-blob-storage and you will see that sometimes 'it works', sometimes it doesn't
<richardlitt>
Cool. Thanks daviddias
<daviddias>
jbenet: :D
<dignifiedquire>
daviddias: okay will see what I can do
<daviddias>
also, big props to xicombd_ and vijayee, who started investing more time to contribute to js-ipfs stravaganza :D
<daviddias>
dignifiedquire: thank you! :)
<dignifiedquire>
daviddias: was this on any specific browsers or in all of them?
<lgierth>
sloow sync is slooow
<dignifiedquire>
lgierth: long sync is loooong
<daviddias>
dignifiedquire: we tested chrome
<dignifiedquire>
daviddias: k
<jbenet>
dignifiedquire: any 0.4.0 weirdnesses are pretty high prio to figure out before merging. repros to any issues huge.
<dignifiedquire>
(will no stop spamming the sync so lgierth is happier)
<parkan>
jbenet: ok of course it immediately started working
<richardlitt>
dignifiedquire: you're next
<jbenet>
whyrusleeping ^ looks like we need to spend some time looking over and testing things more thoroughly. may be a good time to do those "long running tests" we've been wanting to do (like keep a node constantly doing work and halt + issue report the moment something goes wrong/unexpected.
<whyrusleeping>
yeap, i've felt the same
* dignifiedquire
prepare incoming
<whyrusleeping>
thats why i'm working on the simulated network test stuff
<jbenet>
parkan: hm :/ if you find it again just record an asciinema of it sned it
<jbenet>
dignifiedquire: awesome, super productive --
<richardlitt>
dignifiedquire: K-O
<richardlitt>
yeah. Dig, you're awesome.
<jbenet>
(we'll fix that)
<lgierth>
:):)
<dignifiedquire>
thanks
<richardlitt>
:D
<dignifiedquire>
(just trying to keep up with daviddias production levels :D)
<daviddias>
ahahah xD
<jbenet>
you two <3
<richardlitt>
OK
<richardlitt>
jbenet: your turn
<richardlitt>
My cafe is closing, so I need to head off back home (2 mins away)
<richardlitt>
I'll be back in 5~
* whyrusleeping
is out of coffee
<dignifiedquire>
run richardlitt run
* whyrusleeping
is pinned to his chair by a cat
<daviddias>
I still miss the days of Seattle where it was hacking all the time with everyone, dimsum and coffee, it was top for productivity. That being said, we should do it again :D
<jbenet>
incoming
<jbenet>
This week i've been in SF meeting with people. so reviewing things on remaining time. same this week.
<jbenet>
- [ ] IPLD
<jbenet>
- [ ] finalize IPLD pathing issue -- didnt finish it but made examples repo which helps. i think i've settled on one pathing option.
<jbenet>
- [x] Review + comment on dist.ipfs.io things
<jbenet>
- [x] Review + comment on firefox gateway things
<jbenet>
(sorry the links to PRs without titles are so unhelpful
<richardlitt>
jbenet: I can make a tool.
<richardlitt>
Like, a chrome extension which pulls down titles
<richardlitt>
Or something that inteprolates titles given issue or PR links
<jbenet>
there were other things i've reviewed here and there
<richardlitt>
seeing as how we all do this, all of the time.
<whyrusleeping>
richardlitt: how about a 'what did i do on github this week' tool?
<richardlitt>
whyrusleeping: I can do that.
<lgierth>
(github puts the title in title="")
<jbenet>
richardlitt: well, i actually want to increase the SNR of posting this
<richardlitt>
jbenet: sure. But not having titles is kind of... not too helpful.
<lgierth>
thanks for cleaning after me
<noffle>
whyrusleeping: can't you already view all commits/prs/issues on your github profile page for the last week?
<richardlitt>
But thanks for all of the CRM!
<richardlitt>
:)
<jbenet>
richardlitt: it doesnt really add much to list every tiny thing, so i dont want to encourage every single github entry / notif to be reposted to this
<dignifiedquire>
richardlitt: that would be awesome
<richardlitt>
jbenet: ok. So, not a 'What did I do', but certainly a 'grab titles'
<richardlitt>
Just for the bigger stuff. Maybe it's bikeshedding, though.
<whyrusleeping>
i want a nice list of everything i did on github, so i can then trim unimportant things down
<richardlitt>
Not a high priority
<jbenet>
lgierth: np at all-- that was super hard to check and revealed a bigger issue: we need end to end tests on the running services
<richardlitt>
human storytelling = the best storytelling
<dignifiedquire>
yeah as whyrusleeping said a nice list with everything, and then I can remove what is too much
Not_ has joined #ipfs
<jbenet>
yeah that may be the sweet spot (ok richardlitt wins)
<jbenet>
;)
<whyrusleeping>
richardlitt always wins
<whyrusleeping>
unless its a drinking game
<richardlitt>
ok
<richardlitt>
So, thanks all!
<whyrusleeping>
lol
<richardlitt>
End of Sprint.
<richardlitt>
Please be online later.
<dignifiedquire>
thanks everonye
<richardlitt>
I ahve to run, but <3 to all of our contribs
<dignifiedquire>
sync starts in 15min
<dignifiedquire>
(video sync)
<whyrusleeping>
richardlitt: chat when you get home?
<richardlitt>
yep
<richardlitt>
whyrusleeping: yep
<daviddias>
15 mins to hangout?
<richardlitt>
talk in 5
<whyrusleeping>
kay
<richardlitt>
yep, apps on IPFS at 18:30GMT
<dignifiedquire>
daviddias: yes or did the plan change?
ygrek has quit [Ping timeout: 240 seconds]
<daviddias>
almost no time to get coffee
<dignifiedquire>
lol
<dignifiedquire>
make it a quickie then
devbug has quit [Ping timeout: 240 seconds]
<daviddias>
I have a grinder now
* whyrusleeping
is enjoying having all his coffee making supplies back
<daviddias>
bought it while I was with whyrusleeping in a very nice coffee shop
<kyledrake>
I'll be on most/all the sprints today
<dignifiedquire>
whyrusleeping: we need to talk coffee at some point, I wanna up my coffee creation process
<dignifiedquire>
richardlitt: are you setting up the hangouts?
<jbenet>
kyledrake: you're alive! <3 yay and great, see you there
<grncdr>
richardlitt: are the hangouts public?
<jbenet>
grncdr: yep, everyone welcome.
<dignifiedquire>
grncdr: hangouts are public, but we provide a youtube stream if you only want to listen for conserving bandwith
<dignifiedquire>
*bandwidth
<jbenet>
grncdr: there are limits so if dont want to participate just use the "live" link to watch to avoid taking up space (we've hit limits often in the past)
<kyledrake>
jbenet :)
<grncdr>
youtube stream is good enoug for me
<grncdr>
I got it right that "Apps on IPFS" is the topic?
<richardlitt>
jbenet: ok, thanks. I am going to merge last week and the week before into one weekly
<jbenet>
richardlitt: sgtm
<richardlitt>
jbenet: I know about those logs, but it is very difficult to search them for highlights; the only way to get all of the logs into one file is to constantly scroll up and down for pagination
<richardlitt>
jbenet: It takes ages to get enough to actually put into a file, so much so that I don't think it is really helpful for this purpose. I can't just export logs.
<jbenet>
richardlitt: irccloud logs for you too. i sometimes go back and skim through to keep up to date with tings.
<richardlitt>
jbenet: oh?! here?
<jbenet>
just scroll up. anyway, the reason i mentioned irc is that often a bunch of cool stuff happens through here that's worth highlighting.
<jbenet>
yes woudl b great if the community helped surface it and add it to the weekly. maybe your idea is the right thing: open a PR for the next weekly ahead of time and have people post things there that they find interesting to fill it out over the week
<jbenet>
you can then curate from that
<jbenet>
or whatevr
hashcore has joined #ipfs
<achin>
skimming irc logs for interesting stuff seems capital-h hard to me
<jbenet>
hey sorry i have to skip out on infra -- lgierth do you need me here? (only thing i'd say is lets make those service tests for testing after deploys
<bigbluehat>
still finding all the things in all the repos :)
<noffle>
there's a lot to sort through :)
<noffle>
every week there's a new issue in ipfs/pm for that sprint, which'll include schedules and individuals' progress
<jbenet>
(Someone suggested we move to biweekly sprints. May be less sync overhead. Worth considering. But the weekly heartbeat is really nice too)
<noffle>
jbenet: I wonder if weekly irc sync + biweekly video sync would work. video sync seems to take up a lot of time
<jbenet>
It does. Also don't have to go to all of them
mrdomino has quit [Quit: Changing server]
<jbenet>
Better agendas up front can lead to better allocation of time
<whyrusleeping>
biweekly is okay in my book
<whyrusleeping>
or maybe have the morning irc sync every week
<whyrusleeping>
but have the chats every other
<whyrusleeping>
or something to that extent
<richardlitt>
daviddias: almost time for libp2p
<daviddias>
:)
<dignifiedquire>
jbenet: -1 for biweekly it takes time, but I feel it's pretty important
<dignifiedquire>
(also it improved in organization level in the past I'd say)
<Kubuxu>
If you guys spend 30mins talking every time, it is needed.
<jbenet>
Yeah I think so too.
<richardlitt>
Don't eat fermented shark from iceland, people
<daviddias>
it will be a quick one as I've been more focused on js-ipfs things, trying to get a bunch done so that then I can go heads down on libp2p :)
<dignifiedquire>
also 30min wouldn't be enough anymore for biweekly, we would have to cut a lot of thigns out
<jbenet>
richardlitt: dude ooooof the smell
<richardlitt>
I'm really glad I am sick right now
<daviddias>
richardlitt: whyyyyy
<richardlitt>
Because I honestly couldn't smell it. I just ate one and it tasted like piss. And that was that.
<kyledrake>
richardlitt what are you up to in Iceland?
* grncdr
missed the live chat about apps, watching it now and... oh no the globe! :(
<jbenet>
Pls remember that people are watching these hangouts to learn something. Hanging out and catching up is important but probably not the best use of the Live portion of the hangout.
<patagonicus>
Hmm. Using go-ipfs-api how can I get the size of a unixfs file? Or even just with go-ipfs, neither ipfs ls nor ipfs object stat report that.
<grncdr>
patagonicus: you can do `ipfs object get $hash` and sum the sizes of the links
<grncdr>
but that's probably not the recommended way
<patagonicus>
I feel like the API should offer a way to do that, but I can use that for now. But that means I'll have to do the parsing myself. :/
zorglub27 has joined #ipfs
<grncdr>
patagonicus: you could try doing a HEAD request to the gateway and checking content-length
<grncdr>
which is the more normal way to check the size of any HTTP resource
disgusting_wall has joined #ipfs
<grncdr>
but I'm not sure if the gateway actually supports that
<patagonicus>
Huh, I get {"Links":[],"Data":"\u0008\u0002\u0012\u0004bar\n\u0018\u0004"} for a file that only contains "bar\n".
<grncdr>
yeah, small chunks of data are inlined into the object
<patagonicus>
I don't even see how I'd go about getting the size of the file (without cat'ing it). What are all these characters in the data?
<grncdr>
jbenet, dignifiedquire: listening to the discussion about the web UI and polling, I don't see how API-over-websockets solves the problem? AFAICT there's still a need to implement the subscription semantics
patcon has quit [Ping timeout: 276 seconds]
boxxa has quit [Quit: Connection closed for inactivity]
<dignifiedquire>
yes but that was only part of the problem, constantly polling the local node is also very inefficent
<daviddias>
anyone else joining js-ipfs?
<richardlitt>
Philz. I miss Philz.
<richardlitt>
I used to live in Philz on Mission.
<dignifiedquire>
coming 5min late
<grncdr>
daviddias: I would just listen
<dignifiedquire>
start wo me
<daviddias>
jbenet: are you joining?
<jbenet>
Yes
devbug has quit [Read error: Connection reset by peer]
rendar has joined #ipfs
devbug has joined #ipfs
montagsoup has joined #ipfs
joshbuddy has joined #ipfs
Encrypt has quit [Quit: Quitte]
Peer3Peer has joined #ipfs
<Kubuxu>
daviddias: For FF add-on a js lib that performs detection if URI is possible IPFS location would be awesome
<Kubuxu>
daviddias: just saying that it would be great to export something like that that would be state of the art detection (currently simple is /ipfs/Qmbase58 and /ipns/Qmbase58 /ipns/correctDomain)
joshbuddy has quit [Quit: joshbuddy]
<Kubuxu>
daviddias: FF add-on is there, we need way to differentiate from /ipfs/go-ipfs and /ipfs/Qmaaaaa..
<Kubuxu>
daviddias: currently we are using our own regex but something that would be kept updated when IPFS changes would be better.
<patagonicus>
It's kinda confusing that go uses a packages original name when printing errors and not the one you used. :/
mildred has joined #ipfs
Aleks__ has quit [Ping timeout: 240 seconds]
<whyrusleeping>
patagonicus: where are you seeing this?
<whyrusleeping>
in a stack dump or something?
<patagonicus>
whyrusleeping: I'm importing github.com/ipfs/go-ipfs-api as "ipfs" and if I use something like ipfs.DoesNotExist it will complain about shell.DoesNotExist being undefined.
<jbenet>
Thanks everyone!
<jbenet>
Thanks richardlitt for running the sprint
<richardlitt>
np. Thanks all!
<jbenet>
I'll be out for a few hrs, back later
<richardlitt>
kk
<jbenet>
daviddias: tell whyrusleeping about captain.log stuff
<whyrusleeping>
captain.log?
<whyrusleeping>
patagonicus: oh, thats annoying
<whyrusleeping>
i should probably change the package name of that one anyways
<whyrusleeping>
shell was the old name for it
<daviddias>
Captain.log is a weekly log of things that have been going on a part of the project, for example I do one for js-libp2p and one for js-ipfs
<daviddias>
it lives inside a issue on the main repo for each project, so that every watcher gets notified
<patagonicus>
Yeah, problem was that I have shell *ipfs.Shell, but I was using ipfs.List instead of shell.List, so go complained that shell.List does not exist. :D
<daviddias>
and I try to list things people can also help if they have time during that week
<patagonicus>
Hmm. I'm calling PatchLink(QmUNLLsPACCz1vLxQVkXqqLX5R1X345qqfHbsf67hvA3Nn, "filename", filehash, false) and I get "incorrectly formatted root hash: add-link". I think the problem might be that the root hash has to come before "add-link".
<whyrusleeping>
patagonicus: i think i fixed that in a recent version of the go-ipfs-api
<whyrusleeping>
have you updated lately?
wiedi has quit [Quit: ^C]
<patagonicus>
I just started working on this yesterday, so yeah.
ispeedtoo has joined #ipfs
<whyrusleeping>
ah, okay
<noffle>
richardlitt: whyrusleeping: fyi I was on livestream during go-ipfs; was doing some things in the bg
<whyrusleeping>
could you file an issue?
<noffle>
did you want to sync about s.t.?
<patagonicus>
Sure
<patagonicus>
whyrusleeping: Ah, I fixed it. I'll open a bug report + pull request. :)
<whyrusleeping>
patagonicus: even better!
<patagonicus>
Luckily it was very simple. :)
mildred has quit [Ping timeout: 256 seconds]
devbug has quit [Read error: Connection reset by peer]
<patagonicus>
I also noticed that I should probably use rm-link instead of starting with an empty dir in my code. Would PatchRmLink analogous to PatchLink be a useful addition to go-ipfs-api?
<whyrusleeping>
patagonicus: you can actually call `Patch(<hash>, "rm-link", linkname)`
<whyrusleeping>
Addlink was added just as a helper method
<patagonicus>
Also, shouldn't PatchLink just be a simple wrapper around Patch? Looks like it's almost doing the same thing.
<patagonicus>
Ok
<whyrusleeping>
yeap
<noffle>
whyrusleeping: ping. let me know if you wanted to sync about s.t. in particular?
joshbuddy has quit [Quit: joshbuddy]
<ipfsbot>
[webui] greenkeeperio-bot opened pull request #222: Update lodash to version 4.0.1
<whyrusleeping>
noffle: oh, yeah
patcon has quit [Ping timeout: 256 seconds]
<whyrusleeping>
what are you planning on working on this week?
<whyrusleeping>
since youre my other go-ipfs-ish person
<noffle>
whyrusleeping: I'm putting that together currently :) I'll update the sprint issue shortly
<noffle>
I may also do a bit of js-ipfs (got to chat /w daviddias still)
Encrypt has joined #ipfs
<daviddias>
yes! :D
computerfreak has joined #ipfs
wiedi has joined #ipfs
hellertime has quit [Quit: Leaving.]
locusf has quit [Ping timeout: 250 seconds]
devbug has quit [Ping timeout: 240 seconds]
codeforkjeff has joined #ipfs
<patagonicus>
I think my dir sync tool (compare dir in IPFS with dir in local fs and add/remove files so that you get an IPFS dir equal to the local dir) might already be usable. Less than a day of work and I've never really written go before. :D
<patagonicus>
But let's test it first, maybe it's completely broken. ;)