Fork me on GitHub
#admin-announcements
<
2015-06-16
>
micha13:06:21

remote for standup today, ito afetr

micha13:06:28

s/afetr/after/

micha13:06:50

derp disregard that

devn17:06:48

It's kind of a bummer we're locked into 10,000 messages in the archive

devn17:06:55

Has anyone spoken with the Slack folks to see if they'd be willing to give us a paid account for free?

erichmond17:06:13

The downside to moving to a paid account would be having to pay $6*15000 /mo right?

devn17:06:21

@erichmond: yes, and for obvious reasons, that ain't gonna happen

devn17:06:49

i messaged their support and politely request they consider giving us access the ability to search the archive

devn17:06:24

being capped at 10k is a pretty big bummer

devn17:06:39

unless someone is capturing all of the messages somewhere and making them searchable...

devn17:06:53

(which is an option)

akiva17:06:20

I do believe that’s a violation of the ToS.

meow17:06:31

@devn: this is a known issue and there is a bot, or there was one. The admins are dealing with it so we probably shouldn't all bug the slack folks.

devn17:06:38

I wonder what the language there is...

akiva17:06:55

It’s actually kind of vague. I think that @gjnoonan was looking into it.

devn17:06:13

some integrations by their very nature force you to see every message anyway

devn17:06:29

people persist all sorts of bits already

devn17:06:57

seems more or less unenforceable

meow17:06:35

@devn: I agree, but I think we're trying be polite at this point. Anyhow, I haven't bothered to look because I've been on here a while, but I'm pretty sure someone added a bot to all the channels and was archiving this. @gjnoonan is probably the best person to ask.

devn17:06:15

@meow: gotcha. thanks for the info. given Slack's recent valuation, you'd think they could find it in their hearts to give an open source community some love.

meow17:06:47

@devn: I know, right! simple_smile And welcome aboard.

akiva17:06:06

Looks pretty clear that we should be able to create our own searchable archives: > Slack is the custodian of data on behalf of the teams that use Slack. We don’t own team communication data. Teams own their data. They like it that way and so do we. and > Administrators of all teams, both free and premium, can export their team’s message content that is shared in channels. Message history for private groups and direct messages is not included in this.

meow17:06:49

IRC + cute colorscheme == $$$

devn17:06:00

@meow: thanks man! I'm kind of amazed to find out about this so late. I've been in IRC since 2010

devn17:06:01

@meow: yeah, and the pricing model just doesn't make sense for a community like this. the number of active users for a particular day is nowhere near the full number of users who have access

meow17:06:12

@devn: the big adoption of slack by clojurians is quite recent.

devn17:06:13

so charging by user is kind of insane

meow17:06:46

On the bright side, the clojure community is really amazing and I'm so glad I decided to start using clojure.

martinklepsch18:06:27

How does map-vals (https://github.com/metosin/potpuri/blob/master/src/potpuri/core.cljx#L220) differ from something like this:

(into {}
  (for [[k v] m]
    {k (f v)}))

juhoteperi18:06:16

martinklepsch: It uses transients

martinklepsch18:06:32

@juhoteperi: and that’s the only difference?

martinklepsch18:06:24

@juhoteperi: is that actually a relevant property in the context of https://github.com/hashobject/perun/pull/11?

juhoteperi18:06:17

@martinklepsch: Well, it's a function so it's easier to compose etc.

martinklepsch18:06:30

@juhoteperi: but that could equally easy be achieved with into & for right? (not trying to rule out using the map-vals from potpuri, just curious for reasons)

juhoteperi18:06:44

martinklepsch: Yes. Or (reduce-kv (fn [acc k v] (assoc acc k (f v))) {} m)

juhoteperi18:06:17

Btw. I don't like slacks nick autocomplete. I always forgot to write @, the nicks aren't ordered by who has spoken the last and ctrl-w doesn't remove previous word but closes the tab.

johnmendonca19:06:38

@juhoteperi: you can start a message with /feedback to send to slack. some good tips there

juhoteperi19:06:58

That was easy. Though I guess bringing readline commands to browser would be hard and upset many, and I can do that myself with cVim.

meow19:06:15

logbot is the bot

meow19:06:16

I can't remember who gets credit for setting it up.

devn19:06:24

@ul did it based on the whois info simple_smile

devn19:06:34

thanks for the link @meow

devn19:06:50

@martinklepsch: i was really resistant to ideas like map-vals, filter-vals, etc. once upon a time

devn19:06:11

but now i have to say i view them as a obvious hole in the stdlib

devn19:06:22

along with things like deep-merge

devn19:06:51

i voted for (into {} (for ...)) on numerous occasions

alexmiller20:06:20

deep-merge has been batted around in the past and Rich didn't like the fuzziness of certain parts of the proposed version

alexmiller20:06:31

but I don't think he was against the idea

alexmiller20:06:53

I personally would love to see a few things like map-keys and map-vals in the box

alexmiller20:06:27

but of course there is nothing stopping you from using the fine versions available in your favorite utility library

alexmiller20:06:23

it has 1 vote :)

arrdem20:06:44

It'd be a fun hack to try to walk the crossclj dataset and find the most used library symbols... maybe even most duplicated functions and then propose those to core.

alexmiller21:06:20

arrdem: that sounds like a terrible idea :)

alexmiller21:06:12

I think weavejester/medley has most of the things in it that I wish were in core

alexmiller21:06:06

fyi, I created #C06E3HYPR

arrdem22:06:27

alexmiller: I mean if your goal is to include in core the things that people otherwise reinvent for themselves 😛

alexmiller22:06:14

Well it's a matter of judgement of course- I don't think core can or should contain everything. But obviously some things are widely used and useful.

alexmiller22:06:35

And I think it's particularly compelling when it affects portability across platforms or greater perf can be achieved

alexmiller22:06:56

Some of the map-keys etc have many possible impls that vary a lot

danielcompton23:06:13

@alexmiller: dissoc-in would be a prime candidate for graduation from Clojure Incubator

gjnoonan23:06:49

@meow: @devn @akiva Sorry for the late reply, I have been out to the cinema and for a meal with friends (Yay Jurassic World). at our current number of users it will cost $125,856 per year to upgrade to the standard plan; Far more than I can afford! I have contacted slack about adding an open source tier to their pricing structure, for the myriad of communities that have started to use it (I myself am a member of 11 so I can attest how popular it is in that space) and how much good standing it would put them in in said communities, heck I couldn’t care if they didn’t give us unlimited integrations .. just unlimited message history. Initially they were very receptive and said they had been throwing around the idea themselves, then after further discussion I was told “not yet”. I contacted them about their rather ambiguous TOS and our logbot but did not hear back— I will Chase that up now though. However as you have discovered we own the data so I don’t see the harm, a few channels have asked for the logbot to be taken out (clojureladies) which we have done.

gjnoonan23:06:04

I will chance them now, and let you know when/if I get a response

alexmiller23:06:51

@danielcompton: yeah, I think the main open question on that is what to do when a submap is empty - remove or leave?

danielcompton23:06:21

Yeah it’s tricky, I’m personally in favour of removing submaps, though I don’t really mind one way or the other though. It would just be nice to have one of them in there, and document it’s behaviour wrt submaps.

dmich23:06:43

@gjnoonan: @devn @erichmond @meow @akiva I know it's probably too late but why not https://gitter.im/ it has unlimited chat history and seems more geared toward developers (not that I've used it). Also has explore option. I find it hard (impossible?) to search for slack teams without somehow just stumbling across them.