This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
2016-01-06
Channels
- # admin-announcements (266)
- # alda (20)
- # announcements (1)
- # aws (16)
- # beginners (16)
- # boot (288)
- # brevis (7)
- # cljs-dev (40)
- # cljsjs (32)
- # cljsrn (5)
- # clojars (23)
- # clojure (169)
- # clojure-art (2)
- # clojure-czech (3)
- # clojure-finland (1)
- # clojure-italy (3)
- # clojure-norway (1)
- # clojure-russia (88)
- # clojure-sg (3)
- # clojurescript (300)
- # clojurewerkz (8)
- # community-development (14)
- # component (4)
- # core-matrix (1)
- # cursive (9)
- # datavis (26)
- # datomic (44)
- # devcards (3)
- # funcool (1)
- # hoplon (7)
- # jobs (4)
- # ldnclj (11)
- # lein-figwheel (1)
- # nyc (2)
- # off-topic (4)
- # om (149)
- # onyx (1)
- # overtone (1)
- # parinfer (15)
- # proton (3)
- # re-frame (9)
- # reagent (45)
- # yada (2)
@nogeek so I have to say http://www.newspaper.io is awesome now...one of my slideshares from a talk I gave in Dec just popped up in my Clojure feed.
@agile_geek: nice!
Anyone have experience using http://hackerrank.com to take a “test” (this is different from just solving the problems)? I timed out trying to get it to output ANYTHING and just have no idea what I could have been missing. Tried: println
, spit
, normal return, providing explicit *out*
, wrapping in doseq
, … nothing.
@nogeek: @agile_geek not had chance to check it out yet, tried to add it into here but we are over our integration limit again 😕
And the Slack crew still shows no interest in supporting open source communities? I still maintain that they should want to take advantage of this sort of thing and work with us.
Often just contributing to open source, the beneficence of it, is enough. People contribute to open source all the time without expecting much if anything in return.
my guess is that Slack is getting "profit pressures" from investors
As far as I understood it from conversations months ago (and @gjnoonan knows more about the details than I do), right now, there’s no tier for open source or 501c non-profits or anything like that.
and spending money on the infrastructure of large, free communities is eating away at profits
@bostonaholic, yeah, I’m not unsympathetic toward Slack. The fact that they let us exist at this capacity at no charge is pretty great as it is.
I wouldn't be quick to blame Slack the Team™, but perhaps Slack the Investors™
Not blaming anyone. I would just like it if we weren’t in this gray area. They could just unplug the whole thing if Slack the Investors™ wanted, I suppose.
right
the IRC #C03S1KBA2, that is
Yeah as @akiva eluded to, I have been trying since I started this community( nearly a year ago now, eek) to negotiate with slack, they won't budge
Hence why I tried to kick off a clojure version, free and open source for the community, by the community showcasing all our tools etc (#community-development)
However for both personal and work reasons I have not been able to dedicate as much time (or any) to it as I wanted to, hopefully that will change as it is something that really interests me
@gjnoonan: Discord has worked out really well for React, but I don’t think it has quite the same integration level Slack has.
@gerred: I think I played with that when @cfleming and I were exploring other options
Personally, I have no desire to return to IRC. I remember maintaining an account on EFNet being a pain in the ass.
If/when we were to move, it is a big ask for the community we have built to be fair, so it would have to be worth it.. I really do like the idea of developing our own solution together, building the community to build the community if that makes sense
I don't blame slack par se, they've got their business plan etc, but I do think they could be a little bit more flexible for OSS communities, even if it was just removing the 10k message limit
I like the idea of building out a custom solution in 100% Clojure. Slack seem to be adamant that large-ish OSS communities are not a use-case they’re interested in supporting, so I wonder at what point will we exceed the limits of what Slack can do to support this community.
While I like the idea of building our own tool, I question how realistic it is to get enough work done on it to make it a useful tool.
If there were another OSS project that does what we want, I’d vote for moving to that rather than building our own. Ideally you want web interfaces, platform native clients, mobile clients… it’s a lot of work.
I don’t think any of it is rocket science, but the time to do it has to come from somewhere.
I’m definitely sympathetic to the “it’s another tool” argument, but Slack realistically is never going to be what we want long term. The archive restrictions are already a PITA.
http://www.mattermost.org/ is pretty great apparently (and it’s OSS)
@shanekilkelly: Yeah, that looked like one of the strong contenders when I looked at this a while back.
I see the argument about our own project being a lot of work, especially for the mobile and native clients
Another option would be a Clojure backend to an existing protocol with existing clients.
Both of those add a lot of complexity probably and might not be exactly what we want.
I think it’s a given that we need something better than IRC, and also clear that Slack is not going to do anything to accommodate FOSS communities. So realistically it’s either a) our own project, or b) one of the OS Slack-a-likes
See, there already is a clojure IRC channel, and while it’s reasonably active, it’s nothing like what we’ve got going on here
I guess if we build our own thing we can at least re-use the web UI from one of the existing OSS solutions to get started.
I think for something to work the concept would have to be a leapfrog over what is already out there as an option. And I don't know what that would look like. Something like slack + some-really-compelling-out-of-left-field feature that would do two things: motivate users to overcome the pain of switching, and two - motivate the development team to go through all the trouble and effort.
Google Summer of Code planning is coming up sooner than you think - maybe this would be a good project for one or more students
pieces of it that is
While I like the idea of using a slack-like-tool that was written in Clojure, unless this new tool was really amazing every developer in every other language camp is going to find the fact that it is written in Clojure to be a turn off. Just being realistic. And I'm hardly realistic in general.
One alternative is to make it somehow friendly to other languages. Kind of like how Cursive makes me love IntelliJ IDEA.
Ok, thought experiment: tomorrow morning Slack announces they’re changing their free tier in a way which makes this community untennable. What do we do?
@shanekilkelly: switch to the best alternative regardless of what it is written in.
Bingo. Which at this moment looks like either Discord or Mattermost.
I think the idea of writing our own tool was floating around because it could be a good community-building project, not necessarily because the implementation language of the tool actually matters
Discord is pretty cool.
clojure
(defn [a b] (+ a b))
works there with syntax colouring for example.So that is one issue - how to respond to the loss of slack, if and when it happens. The other is to be inspired by slack and create some kind of leapfrog that takes the best of say Om Next and Devcards and other ClojureScript tools and create an alternative, perhaps focusing on web REPL integration or concurrent code editing or something.
huh, voice channels could actually be a really cool and useful feature, especially for one-on-one beginner help and brain-storming
@meow: given that Slack is investor-driven startup, I’m hedging on “When” rather than “If"
I don't think discord has robust enough management for those voice channels for 4k-ish of users.
@jaen: yeah, on demand voice channels where you could splinter off conversations from the main channels.
wouldn’t want a single 4k-member voice lobby
I’m going to download the client and see how it fares
I mean, you can have multiple voice-channels and such, but as far as I understand the unit of invitation, so to speak, is a whole server, not a channel.
I think it makes sense to separate the concerns into disaster-prevention/alternate-short-term-switch and then some new project for those (and I'm one of them) who would like to build a community-supportive tool built as a community effort showcasing the best of Clojure/Script.
Or I might be wrong now that I look at it, huh. I seem to get different invitation links for different voice channel. Maybe it could be manageable? Who knows.
yeah, it’s not clear how the channels work.
I wonder if you can do both public lobbies and private voice
Slack was a great move and it has been a lot of fun to watch it grow and thrive. So I think we should be risk-averse when it comes to planning for how to deal with the plug being pulled on us. I'd hate to lose what has been built up so far. Unless it were trivial to create a slack-alike I'm in favor of folks evaluating alternatives that are available right now.
@meow and @shanekilkelly are both right, it is more about community building and showcasing what we have, rather than creating another slack service for others to use all mattermost, although being oss that would be possible
And I'm just throwing out my opinion. Nothing more, nothing less. I'm happy to defer to folks like @gjnoonan who got the ball rolling and did such an awesome job of it.
This is like working with Borland products back on Delphi before there was even America Online or the Internet. The Clojure community is such a good one and I just hope it stays that way.
If people want to jump in the #community-development channel, maybe we should form a "task force" to evaluate alternatives, and discuss exactly want we want and need.. And whether building our own is viable etc
Or am I confusing it with the Borland Object Pascal language thing? I programmed in that too.
Oops, Delphi was the Borland language. What was the big dialup service that I would go to for Borland tech support and such?
I just did a quick google search and apparently there are people still using AOL dialup service. Say what?!?!?!?!!
I agree that a Slack replacement in clj(s) would be a nice showcase, but I think that getting off Slack is more urgent than that.
What if we created something like shared Devcards that were like Wiki pages that anyone could edit and the code snippets could be executed in a REPL and there was chat built into it? Just trying to think of a best of all things combo.
Does it allow pasting code snippets etc? Seems like it’s aimed more at gamers who wouldn’t need all that.
Possibly, but then aren't we moving the community twice? I suppose it's self hosted so we could do it transparently though so not as bad as first though
I'm happy to get mattermost installed on http://clojurians.net tonight to trial, @cfleming you up for helping test again? And a few others
@meow: I like the idea of a devcards like integration down the road :-) would be a good extra for helping in the #beginners
I’m in nine Slacks now and we use it extensively at work. I stopped using IRC altogether a while back because Slack was better (and has a much bigger community for each of the topics I participate in). I can see moving to Gitter (which I have to use for one project) but I wouldn’t want yet another chat app open on my system 😞
@gjnoonan: I’m pretty busy right now, no promises sorry - let me know when you have it set up and I’ll try to take a look.
@seancorfield: But what do we do about The Slack Problem then?
What are folks’ objections to Gitter? (just curious since that rarely seems to get mentioned)
As for The Slack Problem… just live with it? There’s http://slackarchive.io for the archives (Slack are working out some sort of arrangement with them to satisfy their T&Cs).
http://matrix.org looks interesting, it’s an open standard for Slack-like things
@seancorfield: I think you are representative of the big challenge that we face - who wants to add yet another tool? I'm only on 3 slacks, but even there I would still have 2 I can't leave behind.
Gitter can have arbitrary chat rooms but mostly they’re associated with repos or organizations.
It’s still a private company so who knows what they’ll do in the future, but I can live with that I think.
all of the gitters I'm on are specific to a repo - if someone created a general Clojure one we could test it out.
doesn’t gitter pose the same challenge as irc? Not very user friendly to start with? You need to know what github is, etc….. how many users does it support? History? etc...
I think the archive problem is significant for Slack - I’ve lost a lot of DMs, for example, which aren’t archived anywhere.
https://gitter.im/clojure/general?utm_source=share-link&utm_medium=link&utm_campaign=share-link if folks want to try it out
@roberto: gitter is pretty easy - I have a couple of gitters for a couple of my repos - very easy to set up
I just created that room under the clojure organization.
But, yes, you need a GitHub account to use it so that is a barrier for some people.
My issue with gitter is that we can’t tie together disparate repos into a single… well, that it’s not Slack. Hah.
I'm not sure Gitter has a multi-billion $$$ color palette, but you can't have everything.
@seancorfield: So to use Gitter we’d set up clojure/<whatever> rooms?
I guess the nice thing about that is that the Cursive one doesn’t have to be tied to clojure.
with this slack, to discover all clojure related rooms, there is one place to do that
@akiva: Sure, we can have all the Clojure-related rooms under the clojure organization.
Some tied to repos, some just "free" channels.
i've been working on a slack-alternative written in cljs and using datomic, that is designed around 'tags' (instead of rooms) and conversations (this really helps with the '3 simultaneous conversations in a room' problem), have a look: https://www.youtube.com/watch?v=pa2bUsChFqM
it's more mature than in the video (that was 5 months ago), but I was planning on open sourcing it (and having a paid hosted version)
I'd be willing to give the project over to the community as a base to work off of, and lead if necessary
@rafd: That's very generous of you and deserves taking a look. I will just warn you in advance that people like me might be brutally honest about any perceived shortcomings, just like we were with Gitter. We just need to look out for the greater community and allow its needs to take priority.
@cfleming: Just a thought for you. I think you definitely should have a Gitter room for Cursive so that you have archives and it is tied to Github and such. You have a tremendously popular product and you need archiving and you need to offer technical support. So it is a good fit for you. Then challenge there would be simply how to divide you attention between Gitter and the #C0744GXCJ channel here on Slack. And figure out how we can easily send people from here to your gitter room, if you decide to go that route.
For the record, gitter got vetoed because: No image pasting, no linking to other gitter channels, slightly less user friendly, bad doc hurts feature discoverability.
Having said that, I think part of the success of slack has been the ease with which anyone can create a new channel - and then popular channels flourish and less popular ones just sit there stagnating, but don't interfere with anything.
@rafd: Looks interesting, so it’s structured around conversations rather than rooms?
There’s also https://zulip.org/
I'll admit that the chat thing I posted isn't near slack levels of polish, but if it's a concept that interests the community, I and my partner have the capacity to work full time to bring it up to speed over the next few months
@cfleming: yes, conversations that can be tagged (and pretty much belong to multiple rooms)
Perhaps we should make a list of candidates, and make a time to get together to try them in a co-ordinated fashion.
we have 4500 people in here, if even just 20% of us can contribute a small monthly payment I think it'd be better than forcing everybody who's already switched from irc to slack to move somewhere else again
@bronsa: It’s very expensive - because it’s per user, for a community this size it ended up being multiple hundreds of thousands of dollars a year
It’s been suggested that Slack allow users to individually pay monthly to access archives, I’d love that - I’d definitely pay the $6 a month for that.
AFAIK there’s been no interest though, and I can imagine their auth system not being designed for that.
That’s why I want Slack to step up and offer some solutions for FOSS and non-profit groups so we could opt in to certain features while accepting other limitations.
Ironically, a tool like Zulip OSSed by an established company is probably safer in terms of long-term viability.
In my experience, when you’re in a gray area like this, they ALWAYS pull the plug eventually. It’s just a matter of time even if that timespan is years.
Matrix actually looks pretty interesting. Decentralised protocol for this sort of thing, and their next-gen web UI is a bunch of reusable React components: http://matrix.org/blog/project/matrix-react-sdk/
I think that’d be the way to go. Find something with a robust foundation and then build tools on top of it. I think the community would rally around that far more than starting from lein new
.
@cfleming: (Raf is in transit right now, I’m the fellow working with him) not public yet, although it functions on invites
the UI needs a bit of polish, b/c it's been 'good enough' for us, but could definitely be improved
should work at: http://chat.leanpixel.com
So I created a Cursive group - how are the group conversations distinguished? Only by their tags?
@jamesnvc: Can you shoot me an invite too? (address in my profile)
thanks, we weren't really planning on having a bunch of new users tonight; we've mostly been using it internally and haven't polished the on-boarding process
@cfleming: our idea is that you would just subscribe to the tags that interest you and people would have more focused conversations
to be fair, it may be a better model for teams rather than a community.... but they 'many topics at once in a room' is definitely an annoyance that i've experienced with IRC-like systems
I’m subscribed to a couple of channels that I mostly ignore, but have mention alerts on.
Including #C03S1KBA2 and #C03S1L9DN, which I can’t keep up with and get work done at the same time.
@gjnoonan: In your discussions with Slack, have you discussed the possibility of individuals paying for access to the archives?
@rafd: If you are serious about this I think you should come back with a demo when you are ready. If you look at our previous discussions you should have a sense of the requirements because everything is being discussed out in the open, including what we don't like with, for example, gitter.
I personally don't think we need to be terribly formal about any of this, unless someone wants to go through the trouble. We need something very much like slack, which I guess is to slack's credit that we feel this way. Slack has made a lot of things very, very easy. We can post text, code snippets, multi-line code, pictures, links, etc. People use all of these ways of posting. All of them.
@rafd: @jamesnvc: I agree - if you’re interested in developing it I think the community would definitely be interested. I think the interaction model will need work for a large community though, and I worry that it’s sufficiently different from existing solutions to hamper adoption.
We create lots of channels without consensus or control. Popular ones thrive, less popular ones don't get in the way. This is working. I wasn't sure early on but I think this is important. (I also am just one loud voice, no more or less important than anyone else's).
@gjnoonan: I’d be interested to know what they say. If I could pay for access to the archives in Clojurians, I’d have no need to switch and would prefer to stay on Slack.
@cfleming: I will bring it up, as I need to contact them again anyway. Will report back
It seems pretty clear they won’t do anything for the community as a whole, but if they allow it per-user I think that would be an acceptable compromise.
there are some archives recorded here: http://clojurians-log.mantike.pro/
I think someone could setup a bot to do complete recoding, not sure if it is not against Slack TOS tough
@darwin: Well, that’s the conversation being had between Slack and SlackArchive right now, since logging conversation is technically against the T&C as I understand it.
If they can work that out, then SlackArchive becomes a really good option.