-
Notifications
You must be signed in to change notification settings - Fork 332
Meeting 2017 04 21
https://identi.ca/larjona/note/JiaSpssmSVaB_02AGdCdCg
Our monthly community meeting will be today Friday 2017/04/21 at 20:00 UTC on the #pump.io channel on the Freenode IRC network, which is also mirrored to the [email protected] jabber/XMPP MUC room.
You're all welcome to join us there =)
Feel free to add to this before the meeting!
- Updates from last meeting's tasks
- Community governance and sustainability
- Conservancy membership: We're approved. We need to design/elect a project leadership committee (several individuals) to sign the agreement with Conservancy.
- Node status
- Node adoption
- Funding
- Pump.io code development
-
connect-auth-pumpio: mostly wondering about the name/namespacing (can't use scoped packages because of old npm versions) - Perjury: wondering if there's any reason not to; if Evan says OK then I trust him
- CLA: need an OK from Evan since I'm not sure exactly how he wants the CLA to be done, or if this matters that much
- Node 0.10/0.12
- We basically only support these because Debian
- https://nodecompat.com/ shows most modern distros have 4+; upcoming Debian stable is very close (in feature freeze) and will ship with 4+, plus 4+ is now in jessie-backports
- We have technical debt because of these (#1234) and we cannot support Node 7 without dropping support for 0.10/0.12, because Zombie
- Proposal: next release (4.0) is the last to support these Node versions. That gives a bit under 2.5 months from today for admins to upgrade.
- bcrypt: need a response to the above issue - wondering why Evan wants to switch and if he has a comment on the perf impact
[22:04:13] <larjona> #############################################################
[22:04:13] <larjona> BEGIN LOG
[22:04:13] <larjona> #############################################################
[22:04:13] <larjona> Welcome to this month's Pump.io community meeting! Everyone
[22:04:13] <larjona> is welcome to participate.
[22:04:13] <larjona> This meeting is being logged and it will be posted on the
[22:04:13] <larjona> wiki at https://github.com/e14n/pump.io/wiki/Meeting-2017-04-21
[22:04:13] <larjona> If you would like your nick redacted, please say so, either now or after the meeting.
[22:04:26] <larjona> Let's start with roll call - who's here?
[22:04:29] -*- larjona is here
[22:05:26] -*- strugee is here!
[22:05:46] -*- jxself hovers
[22:07:19] <larjona> paroneayea? weren't you here? :)
[22:07:54] <larjona> let's wait a couple of minutes more...
[22:08:04] <evanpro_> Hey all
[22:08:10] <larjona> hey evanpro_! welcome!
[22:08:10] <evanpro_> Sorry I'm late
[22:08:17] <larjona> we're in roll call
[22:08:29] <paroneayea> i'm here
[22:08:30] <paroneayea> sorry
[22:08:35] <larjona> welcome!
[22:08:40] <-> evanpro_ es ahora conocido como evanpro
[22:08:47] <evanpro> I'm here
[22:09:22] <strugee> hey evanpro!
[22:09:27] <strugee> thanks for being here
[22:09:34] <larjona> ok, I remind the agenda
[22:09:35] <larjona> https://github.com/pump-io/pump.io/wiki/Meeting-2017-04-21#notes-from-strugee-in-case-internet-drops-out
[22:09:38] <larjona> pardon
[22:09:44] <larjona> https://github.com/pump-io/pump.io/wiki/Meeting-2017-04-21 << this one
[22:09:53] <evanpro> Oh no
[22:10:02] <evanpro> I hope the Internet doesn't disappear entirely during this meeting
[22:10:06] <strugee> lol
[22:10:08] <strugee> I'm on a train
[22:10:09] <strugee> (again)
[22:10:11] <evanpro> Ahhhhhh
[22:10:17] <evanpro> OK let's do this thing!
[22:10:21] <larjona> ok, topic 1
[22:10:23] <larjona> Updates from last meeting's tasks
[22:10:40] <larjona> I have to say that I couldn't put much time in pump.io last month
[22:11:06] <jxself> That's OK - we still like you. :)
[22:11:08] <larjona> I was expecting to do some bug triaging and update the documentation license and maybe review some wiki pages, but didn't do anything
[22:11:09] <strugee> I don't think anyone had specific tasks? can't remember though
[22:11:20] <strugee> ah right. no worries larjona
[22:11:21] <larjona> there is the [email protected] mail alias pending
[22:11:37] <larjona> and the network monitoring website pending too
[22:12:10] <evanpro> [email protected] for me, right?
[22:12:12] <strugee> ah the uptime status was up for a bit and is now down along with the rest of my server
[22:12:13] <strugee> evanpro: yeah
[22:12:14] <larjona> yes evanpro
[22:12:36] <strugee> just have it forward to [email protected] and you if you want
[22:12:44] <strugee> larjona: can't remember if you asked to be cc'd on those?
[22:12:46] <evanpro> strugee: OK
[22:12:52] <larjona> I said yes
[22:13:09] <strugee> thx evanpro
[22:13:38] <larjona> I guess that the rest of the pending stuff just passes on our TODO for this month? :)
[22:14:04] <larjona> and move for next topic
[22:14:09] <strugee> yeah
[22:14:14] <larjona> topic 2. Community governance and sustainability
[22:14:29] <larjona> Our application to Conservancy has been approved
[22:14:34] <larjona> now it's official
[22:15:18] <evanpro> That's awesome!
[22:15:41] <paroneayea> WOW!
[22:15:44] <paroneayea> congrats :D
[22:15:52] <strugee> \o/
[22:15:53] <larjona> I've got a mail from Conservancy and we need to send to them the names of the persons that would sign the agreement
[22:16:09] <larjona> Details about "being a member" here: https://sfconservancy.org/projects/apply/
[22:16:35] <evanpro> @larjona, can I propose some names?
[22:16:41] <larjona> yes, of course
[22:16:58] <larjona> let me paste here for the record what they need
[22:17:16] <evanpro> I'd like to suggest strugee, @larjona, myself
[22:17:38] <larjona> fiscal sponsorship
[22:17:38] <larjona> agreement (FSA) for pump.io's review and signature. The agreement will
[22:17:38] <larjona> be between Conservancy and a group of signatories representing pump.io.
[22:17:38] <larjona> Traditionally, the signatories for the FSA also comprise the project
[22:17:38] <larjona> leadership committee (PLC) -- the body of founders, leaders, and/or
[22:17:38] <larjona> influential committers who will represent the project in its
[22:17:38] <larjona> relationship with Conservancy.
[22:18:50] <evanpro> So
[22:18:53] <evanpro> What next?
[22:19:04] <larjona> I guess we agree, if possible :)
[22:19:15] <evanpro> Can I propose we vote?
[22:19:34] <evanpro> Or something like that?
[22:19:48] <strugee> those names LGTM
[22:19:53] <larjona> same for me
[22:20:07] <evanpro> PROPOSED: first project leadership council for pump.io consists of evanpro, strugee and larjona
[22:20:21] <strugee> +1
[22:20:24] <evanpro> +1
[22:20:27] <larjona> I don't know if me not being US-based is a problem, I guess no
[22:20:49] <evanpro> I think they have other non-US members
[22:21:00] <larjona> +1 then
[22:21:27] <evanpro> Any objections?
[22:21:45] <larjona> paroneayea jxself what do you think?
[22:21:57] <jxself> I'll +1 it too
[22:22:32] <larjona> ok so I think we all agree
[22:22:55] <larjona> I'll send those names and mails to Conservancy, so they draft the FSA or ask the other needed things
[22:23:25] <larjona> meanwhile, I'd like to ask you to review the Conservancy webpage I posted before
[22:23:30] <larjona> https://sfconservancy.org/projects/apply/
[22:23:31] <larjona> and
[22:23:37] <larjona> https://sfconservancy.org/projects/services/
[22:23:55] <larjona> and you can send to me whatever doubts or comments you have, and I can transmit to them
[22:24:29] <larjona> They are in #conservancy here in Freenode, I joined the channel
[22:24:58] <strugee> cool.
[22:25:02] <larjona> anybody is welcome to join too, but if you feel it as a burden to keep track of it, don't worry, I'll keep track
[22:25:12] <larjona> and serve as a bridge
[22:25:18] <evanpro> larjona, the main question I have is how we can start moving servers and domain names to the Conservancy
[22:25:24] <evanpro> Ones that haven't been adopted
[22:25:51] <larjona> I guess we need to sign the FSA before, but I'll ask
[22:25:57] <evanpro> OK, cool
[22:26:05] <strugee> evanpro: are you thinking they'd be paid for by project donations?
[22:26:15] <evanpro> Well, that would be a nice way to do it
[22:26:20] <evanpro> Or even a tip jar system
[22:26:26] <strugee> yeah
[22:26:33] <larjona> they get the 10% of funding revenues of projects
[22:26:40] <evanpro> Like, if you use this server, consider giving /month to support pump.io and public servers
[22:26:52] <strugee> makes sense
[22:26:54] <strugee> to both
[22:27:00] <evanpro> larjona: that sounds great
[22:27:18] <larjona> my "they" was meaning "Conservancy", I'm not sure what your "they" meant
[22:27:42] <strugee> "they'd be paid" = servers
[22:28:06] <strugee> and associated costs like HTTPS certs and domain renewals
[22:28:14] <larjona> ok, now it's clear
[22:28:14] <evanpro> Exactly
[22:28:22] <strugee> right on
[22:28:51] <evanpro> If the Conservancy owns the servers and all the domains and stuff, I think it'd make more sense to have volunteer admins too
[22:28:58] <strugee> right
[22:29:08] <evanpro> This is down the road of course
[22:29:30] <evanpro> larjona: should we move on?
[22:29:34] <larjona> yes
[22:29:56] <larjona> I'll report in a wiki page or however the things that I learn from this topic, and people can participate/ask/propose too, ok?
[22:30:06] <strugee> +1
[22:30:06] <evanpro> Perfect
[22:30:14] <larjona> I guess the "Funding" subtopic is covered too already
[22:30:23] <larjona> (I'll ask about how-to to Conservancy too)
[22:30:30] <larjona> then we have node status and node adoption
[22:30:35] <larjona> Any news on this?
[22:30:47] <evanpro> Well, let me start with node adoption and work backwards
[22:30:57] <larjona> go
[22:31:00] <paroneayea> +1 (late)
[22:31:06] <evanpro> I've been sending out node adoption paperwork to people who offered to adopt servers
[22:31:31] <larjona> great
[22:31:31] <evanpro> Even if we move the un-adopted servers under the Conservancy umbrella...
[22:31:52] <evanpro> ...I still think it's a good idea to have a variety of owners/admins/responsible parties on the network
[22:32:00] <strugee> agred
[22:32:03] <evanpro> So I'm going to proceed apace
[22:32:08] <strugee> s/agred/agreed/
[22:32:27] <thunfisch> evanpro: you've sent me an email about maintaining a server earlier. thought i'd drop by. :)
[22:32:32] <evanpro> Anyway, sharing out the paperwork, which I think is pretty readable
[22:32:35] <larjona> indeed, Conservancy only provides the "name ownership", the nodes need to have somebody actually working on them
[22:32:37] <evanpro> thunfisch: Oh awesome!
[22:32:44] <larjona> welcome thunfisch!
[22:32:46] <evanpro> larjona: right
[22:32:54] <evanpro> So, that takes me to node status
[22:33:15] <evanpro> A lot of the current pumps are down -- I think 3-4 at least
[22:33:26] <evanpro> There are also a few of the third-party apps down
[22:33:36] <evanpro> So I'm gradually trying to revive them and transfer them
[22:34:08] <strugee> great
[22:34:12] <evanpro> strugee, I *think* the most reasonable structure for this is to use a docker configuration with 4.0.0
[22:34:21] <jxself> https://sjoberg.fi/pumpcheck.txt
[22:34:50] <evanpro> That is, something like an Ubuntu 16.04 server with Docker running, and a Docker instance of pump.io and MongoDB running using Docker Compose
[22:34:51] <strugee> evanpro: did you ever get back openfarmgame.com back?
[22:34:57] <evanpro> strugee: no, not yet
[22:35:03] <strugee> evanpro: we have an open PR for Docker support, OP got a fair way through and then didn't come back
[22:35:03] <evanpro> It's on my list
[22:35:08] <strugee> I think I'll just finish it up myself soon
[22:35:24] <evanpro> strugee: I've been dockerizing a lot of stuff lately, so if you want to assign it to me I'd love to take a shot
[22:35:33] <thunfisch> I'm in the process of writing a Dockerfile already right now
[22:35:39] <thunfisch> Should I target 4.0.0 already?
[22:35:48] <evanpro> I think so!
[22:35:51] <thunfisch> Alright.
[22:36:19] <thunfisch> I'd be running this on my swarm behind an nginx for tls termination.
[22:36:31] <evanpro> jxself: yikes!
[22:36:41] <evanpro> Oh awesome
[22:36:41] <thunfisch> So using Docker would be far easier, especially as I'm not touching ubuntu with a ten-foot pole. :)
[22:36:52] <evanpro> thunfisch: OK, cool
[22:36:59] <xmpp-pump> [strugee] evanpro: will do
[22:37:04] <thunfisch> Can you give me an estimate on how big the datadir would usually get?
[22:37:19] <evanpro> Uh...
[22:37:21] <thunfisch> Just so I can be sure to provision an appropriate size for the block-volume
[22:37:21] <evanpro> Lemme see
[22:37:28] <strugee> both of you should check out the existing PR if you want to work on this
[22:37:30] <larjona> Pump,io issue about Docker image, FTR: https://github.com/pump-io/pump.io/issues/789
[22:38:05] <evanpro> microca.st uses ~20Gb of data
[22:38:13] <larjona> and the related PR https://github.com/pump-io/pump.io/pull/1307
[22:38:14] <evanpro> That's been up for a few years
[22:38:29] <evanpro> thunfisch: so that's probably a good comparison
[22:38:42] <thunfisch> ouch, both those Dockerfiles can be improved.
[22:38:57] <evanpro> thunfisch: I'd probably like it to be 12-factor
[22:39:07] <thunfisch> evanpro: alright, that's next to nothing. I'll just reserve 100gb for now, can increase if necessary
[22:39:09] <evanpro> OK, let's go on to next items
[22:39:17] <larjona> ok
[22:39:21] <larjona> topic 3: Pump.io code development
[22:39:42] <larjona> subtopic: ActivityPub implementation
[22:39:59] <evanpro> Yeah!
[22:40:18] <strugee> not much to report here
[22:40:22] <larjona> I think there was a kind of deadline for this, that I don't know if it passed or not
[22:40:25] <larjona> paroneayea?
[22:40:47] <xmpp-pump> [strugee] End of month
[22:40:50] <evanpro> So, the big thing is that we need to get an implementation done to help get ActivityPub published
[22:41:02] <evanpro> strugee: are you bridged AND on IRC?
[22:41:10] <xmpp-pump> [strugee] I poked at this a little this week, didn't really get anywhere. Mostly because I'm bad at implementing specs
[22:41:17] <xmpp-pump> [strugee] evanpro: yeah
[22:41:25] <xmpp-pump> [strugee] I use XMPP on my phone
[22:41:29] <jxself> strugee is everywhere.
[22:41:33] <evanpro> Ah, that makes a lot of sense
[22:41:50] <evanpro> strugee: would it be nuts for us to just do a weekend sprint sometime in May?
[22:42:02] <thunfisch> I guess mongodb is the usual approach?
[22:42:04] <xmpp-pump> [strugee] Sketchy internet right now mean's I'm switching :)
[22:42:09] <evanpro> thunfisch: almost always
[22:42:11] <xmpp-pump> [strugee] That would work
[22:43:08] <larjona> ok so you two evanpro and strugee agree later on date and so? (Sorry I'm being hurry but I guess the following items need our attention)
[22:43:12] <evanpro> Like 5/20-21?
[22:43:20] <evanpro> larjona: that's fair here
[22:43:23] <xmpp-pump> [strugee] Gonna try to push through next week though. We'll see how it goes
[22:43:36] <xmpp-pump> [strugee] Yeah let's move on
[22:43:38] <larjona> fine
[22:43:38] <xmpp-pump> [strugee] I'll email
[22:43:44] <larjona> subtopic: AJ needs access to connect-databank
[22:43:50] <larjona> I guess AJ is strugee
[22:43:59] <xmpp-pump> [strugee] Yep
[22:44:00] <evanpro> Yes
[22:44:12] <evanpro> so I just did the command listed in the agenda item
[22:44:20] <xmpp-pump> [strugee] Thx
[22:44:24] <larjona> ok
[22:44:25] <xmpp-pump> [strugee] Let's move on
[22:44:27] <larjona> subtopic: OK to publish connect-auth-pumpio?
[22:44:56] <evanpro> So, this is the forked version of connect-auth?
[22:44:57] <xmpp-pump> [strugee] Mostly wondering about naming
[22:45:05] <xmpp-pump> [strugee] Yeah
[22:45:07] <evanpro> Let's go with it for now
[22:45:10] <xmpp-pump> [strugee] Upstream is dead
[22:45:17] <evanpro> No need to make up a better name
[22:45:26] <evanpro> Are we going to move to passport at some point?
[22:45:47] <xmpp-pump> [strugee] We're gonna move to SOMETHING
[22:45:59] <xmpp-pump> [strugee] passport seems like a good choice
[22:46:35] <evanpro> OK, so let's just publish what works now and put passport on the roadmap for the future
[22:46:49] <evanpro> It's widely used, so we get some herd immunity for security issues
[22:47:09] <xmpp-pump> [strugee] There's an issue open
[22:47:13] <xmpp-pump> [strugee] I'll put a comment in it that passport will be the replacement
[22:47:18] <evanpro> Cool!
[22:47:26] <evanpro> Unless there's some new hotness I haven't heard of
[22:47:31] <evanpro> Which is entirely possible
[22:47:36] <xmpp-pump> [strugee] I'll look
[22:47:39] <xmpp-pump> [strugee] Later
[22:47:40] <larjona> https://github.com/pump-io/pump.io/issues/1321
[22:47:41] <evanpro> Let's keep going!
[22:48:02] <larjona> Switch to Perjury for unit tests?
[22:48:09] <larjona> (next subtopic)
[22:48:15] <evanpro> I'm +1
[22:48:31] <evanpro> So, we currently use vowsjs
[22:48:36] <evanpro> Which is unmaintained
[22:48:59] <evanpro> I had to write perjury (false vows) to run all my vows tests
[22:49:05] <xmpp-pump> [strugee] evanpro: if you're +1 I'm +1. I just wanted to confirm there weren't any major issues
[22:49:13] <evanpro> It should be close to a drop-in replacement
[22:49:25] <larjona> Is there an issue about this? I cannot find it
[22:49:30] <xmpp-pump> [strugee] Vows in theory has a new maintainer although there hasn't been that much activity
[22:49:35] <xmpp-pump> [strugee] larjona: no
[22:49:45] <larjona> ok so strugee, will you open it?
[22:49:53] <xmpp-pump> [strugee] I added it to the agenda right before the meeting since I happened to see it while poking npm
[22:49:58] <larjona> ok
[22:49:59] <xmpp-pump> [strugee] Yeah will do
[22:49:59] <evanpro> ha
[22:50:02] <larjona> fine
[22:50:12] <larjona> subtopic: Where to put the CLA? (#1283)
[22:50:19] <larjona> https://github.com/pump-io/pump.io/pull/1283
[22:50:46] <strugee> evanpro: just need your OK on this rename
[22:50:50] -*- jxself points out that the DCO isn't really a CLA...
[22:50:57] <strugee> not sure if there was any significance to what was there before
[22:51:13] <strugee> hm
[22:51:20] <strugee> yeah, idk how this is done usually
[22:51:25] <evanpro> So I agree
[22:51:35] <evanpro> DCO's usually live in a HACKING or CONTRIBUTING file
[22:51:38] <evanpro> So this is good
[22:52:03] <evanpro> I think there's a trick to putting them in a git hook so they appear when you try to push, but I don't remember exactly how that's done
[22:52:25] <strugee> evanpro: you mean with putting it in CONTRIBUTING.md?
[22:52:44] <strugee> CLAHub is a status check for GitHub PRs
[22:52:58] <strugee> that's what we used to use but it's kinda busted now
[22:53:04] <evanpro> Right
[22:53:05] <evanpro> Forget it
[22:53:09] <strugee> ok
[22:53:15] <evanpro> strugee: yes, I think sticking it in the contributing file is fine
[22:53:17] <evanpro> It's pretty short
[22:53:24] <paroneayea> oh
[22:53:25] <paroneayea> hi
[22:53:30] <paroneayea> sorry, I kept stepping afk
[22:53:32] <evanpro> paroneayea: hi
[22:53:38] <strugee> should we move find something similar but not busted? or not bother
[22:53:43] <strugee> evanpro: ok
[22:53:45] <evanpro> not bother
[22:53:49] <strugee> cool
[22:53:52] <paroneayea> I'm back in case you want me around for AP stuff
[22:53:57] <evanpro> paroneayea: OK
[22:54:08] <strugee> we were discussing a weekend sprint at some point
[22:54:11] <evanpro> We are going to do a sprint this month to get support into pump.io
[22:54:18] <paroneayea> \o/
[22:54:19] <strugee> I'll cc you on the email paroneayea
[22:54:27] <paroneayea> strugee: please do, thanks
[22:54:33] <strugee> sure
[22:54:36] <larjona> not this month, in May :)
[22:54:51] <paroneayea> strugee: I wonder if it should be a meta-sprint, encourage other projects to "bring their own project to add AP to"
[22:54:54] <evanpro> larjona: I stand corrected
[22:54:57] <strugee> yeah sounds good
[22:55:07] <evanpro> paroneayea: happy to be part of that
[22:55:09] <strugee> we can coordinate this over email
[22:55:25] <strugee> let's move on; we've only got 5 minutes left
[22:55:26] <larjona> fine
[22:55:26] <thunfisch> how's the support for node>=7?
[22:55:38] <strugee> thunfisch: that's what I want to talk about
[22:56:00] <larjona> are we done with the CLA?
[22:56:12] <strugee> yeah I think so
[22:56:20] <thunfisch> okay :) I'm just dropping in some questions while I'm writing the Dockerfile right now. I'll base off of alpine:3.4 for now, which uses nodejs=6.7
[22:56:27] <larjona> wait thunfisch
[22:56:31] <larjona> a sec
[22:56:33] <thunfisch> okay
[22:56:33] <evanpro> thunfisch: I don't think it'll work with alpine right now
[22:56:40] <thunfisch> How come?
[22:56:50] <evanpro> We use one binary extension
[22:56:58] -*- larjona waits, better :)
[22:57:32] <evanpro> And the way that alpine does dynamic libraries makes big explosions when you dockerize Node with binary extensions
[22:57:34] <thunfisch> Which one is it? Should'nt be too hard to get running with musl
[22:57:42] <evanpro> thunfisch: bcrypt
[22:57:49] <evanpro> We're going to discuss it in a second
[22:57:51] <larjona> ah! you came to the topic!
[22:57:52] <thunfisch> okay
[22:57:56] <larjona> subtopic: JS bcrypt implementation (see #1233)
[22:58:02] <evanpro> Right
[22:58:03] <larjona> https://github.com/pump-io/pump.io/issues/1233
[22:58:40] <strugee> evanpro: is that the only issue with the binary extension?
[22:58:46] <evanpro> Yes
[22:58:52] <strugee> distros with not-glibc libc?
[22:58:53] <strugee> ok
[22:59:05] <evanpro> Or on non-Intel systems
[22:59:10] <strugee> ah
[22:59:18] <evanpro> ISTR people having things blow up on RazPi e.g.
[22:59:33] <evanpro> The JS is slower
[22:59:35] <evanpro> But
[22:59:36] <strugee> are you sure it won't work?
[22:59:41] <evanpro> No
[22:59:45] <strugee> ok
[22:59:48] <evanpro> But the thing is
[22:59:49] <strugee> we should test that then
[22:59:56] <evanpro> bcryptjs is a drop-in replacement
[23:00:06] <strugee> except for perf
[23:00:27] <evanpro> https://www.npmjs.com/package/bcryptjs
[23:00:29] <evanpro> Right
[23:00:33] <strugee> I'm worried about existing hashes that are done with a higher iteration count
[23:00:46] <evanpro> OK you're talking above my paygrade
[23:00:49] <thunfisch> I'm just quickly trying the bcrypt module
[23:00:59] <thunfisch> but it should really not be an issue
[23:01:01] <strugee> ok
[23:01:13] <evanpro> All I know is that I use bcryptjs for all my systems now and it's great
[23:01:29] <strugee> evanpro: tl;dr there's no way to shave off that perf impact
[23:01:37] <evanpro> It only needs to run at login, registration and when changing your password
[23:01:47] <strugee> hmm
[23:01:48] <strugee> fair
[23:01:49] <evanpro> We're not grinding through millions of hashes
[23:02:09] <evanpro> Making rainbow tables or whatever
[23:02:53] <evanpro> strugee: I think that performance issue is a fair one, but the installation issue is one I've bumped into multiple times
[23:03:02] <strugee> ah ok
[23:03:08] <strugee> on non-Intel? or
[23:03:35] <evanpro> Distros and non-Intel
[23:03:38] <strugee> gotcha
[23:03:40] <xmpp-pump> [JMobile] o/
[23:03:41] <strugee> I'll make the change
[23:03:50] <strugee> too late for this release cycle but
[23:03:50] <evanpro> Cool
[23:03:53] <strugee> next one
[23:03:55] <evanpro> Yeah of course
[23:04:04] <larjona> We just arrived to the end of the agenda
[23:04:05] <strugee> t
[23:04:09] <strugee> no we didn't
[23:04:12] <strugee> 0.10/0.12 support
[23:04:14] <evanpro> We didn't do backwards compatibility
[23:04:15] <evanpro> Right
[23:04:17] <larjona> oh sorry
[23:04:23] <strugee> np!
[23:04:30] <evanpro> strugee, what's the matrix look like now?
[23:04:40] <evanpro> I keep losing that valuable page you have up
[23:04:53] <larjona> subtopic: Drop Node 0.10/0.12 support?
[23:05:44] <strugee> evanpro: nodecompat.com
[23:05:58] <strugee> the only problems are RHEL and Debian
[23:05:59] <thunfisch> I don't have any problems building the bcrypt module on alpine
[23:06:08] <strugee> RHEL we probably just don't care about
[23:06:38] <strugee> Debian has Node 4 in backports now and the upcoming stable, which is frozen, will have Node 4 in main
[23:06:39] <jxself> I wonder how updated that page is? Because Debian does jave 4.7.2 in Jessie backports.
[23:06:57] <larjona> I guess the page only tracks stable
[23:06:59] <strugee> jxself: ah yeah I forgot to update the backports entry
[23:07:06] <strugee> no it does backports too IIRC
[23:07:09] <strugee> I should fix that
[23:07:13] <strugee> it's mostly up-to-date though
[23:07:31] <evanpro> So, if we have an official Docker image...
[23:07:37] <evanpro> ...a lot of the pain here goes away.
[23:07:44] <evanpro> Or at least it goes to Docker.
[23:08:09] <evanpro> I'm not sure how many of these platforms Docker can run on.
[23:08:09] <strugee> anyway, I see no reason to keep supporting these. they're already second-class citizens, we have a bunch of tech. debt that's required to keep supporting them
[23:08:11] <strugee> evanpro: true
[23:08:37] <strugee> even without that though I see no reason to not drop support now
[23:08:43] <thunfisch> evanpro: most modern distros have no issues running Docker. you can even run it on raspberrypi/ARM
[23:08:49] <strugee> (and by now I mean "with reasonable notice")
[23:08:59] <thunfisch> I'll be happy to do a PR when I have my Dockerfile figured out
[23:09:06] <evanpro> strugee: would it be OK to suggest this version policy: stable and LTS versions of Node only, older versions on a best-effort basic, and older OSes should use the Docker image
[23:09:58] <evanpro> Is that too draconian?
[23:10:18] <strugee> I think so. Node 4+ is pretty easy
[23:10:38] <strugee> we have Travis coverage for all of that
[23:10:43] <evanpro> Right
[23:10:53] <strugee> I don't expect Node 4.x to be a problem for a while
[23:11:48] <strugee> well, if you meant *all* Node LTS versions, not just what's on the front page, that seems reasonable
[23:11:54] <evanpro> https://github.com/pump-io/pump.io/issues/1342 <- !!!!!!
[23:12:14] <evanpro> strugee: well after a while that's a hassle
[23:12:18] <thunfisch> Oh, now I see - pump.io still has bcrypt^0.8.x, current is 1.0.2. Do you know if there are any breaking changes for that? Otherwise I'd just sed-upgrade it in the Dockerfile.
[23:12:34] <paroneayea> needs more coroutines :)
[23:12:47] <strugee> thunfisch: there aren't breaking API changes but that upgrade caused problems on install so I reverted
[23:12:55] <evanpro> strugee: so, can you restate my version policy above?
[23:12:58] <strugee> try it and if it works, I say go for it
[23:13:02] <thunfisch> alright
[23:13:46] <evanpro> https://github.com/nodejs/LTS#lts-schedule1
[23:14:23] <strugee> evanpro: it seems very reasonable to me to just follow that schedule
[23:14:54] <evanpro> Where "LTS" means "Active LTS or Maintenance"
[23:14:57] <evanpro> I'm down with that
[23:15:10] <strugee> cool
[23:15:17] <evanpro> \o/
[23:15:18] <larjona> so, anything else?
[23:15:21] <evanpro> Another fine meeting!
[23:15:33] <larjona> I have no ASCII appetizers
[23:15:40] <larjona> but we can go out a bit and watch some nature
[23:15:47] <larjona> ___
[23:15:47] <larjona> /~~ ~~\
[23:15:47] <larjona> /~~ Pump.io ~~\
[23:15:47] <larjona> { Conservancy }
[23:15:47] <larjona> \ _- -_ /
[23:15:47] <larjona> ~ \\ // ~
[23:15:47] <larjona> _- - | | _- _
[23:15:47] <larjona> _ - | | -_
[23:15:47] <larjona> // \\
[23:16:08] <larjona> thanks everybody for coming!
[23:16:08] <strugee> \o/
[23:16:11] <evanpro> Nice!
[23:16:17] <evanpro> Have a good month everyone
[23:16:20] <strugee> evanpro: I'll put up that policy on the wiki
[23:16:22] <strugee> thanks all!
[23:16:26] <evanpro> strugee: +1
[23:16:31] <larjona> #############################################################
[23:16:33] <larjona> END LOG
[23:16:34] <larjona> #############################################################