Fork me on GitHub
#off-topic
<
2019-01-09
>
flyboarder07:01:08

What should I put on a T-shirt? Topic: clojure or functional programming

Eric Ervin23:01:50

Clojure is fn

Eric Ervin23:01:12

Simple not Easy

pavlosmelissinos11:01:19

Hey, any idris users here?

Conor11:01:09

Recreationally

pavlosmelissinos11:01:41

What's your opinion as a clojure user? I learned just recently that dependent types are a thing so I'm thinking about giving Idris a go but I'm not sure whether it's worth the time investment.

Conor12:01:59

It is certainly interesting

Conor12:01:20

Dependent types let you assert certain things that will always be true about a program

Conor12:01:30

Not sure if I would use Idris in production though

pavlosmelissinos13:01:55

That's what I've heard as well, that it's not ready for production. Some people (most coming from haskell) mention Agda as a better alternative. I'm still torn between no types vs a very strict type system (with strong type inference), i.e. clojure vs one of the above 3, not sure which way to go!

Conor13:01:26

Agda is more of a proof system as I understand it

Conor13:01:58

Church types and Curry types suit different problem domains to different degrees IMO

pavlosmelissinos13:01:48

Ah, TIL, had no idea that there's a formalization for the absense of types as well

pavlosmelissinos13:01:57

I should probably take a closer look at type theory... Thanks

pablore16:01:16

Could I use Agda to formally prove code on other langs?

Conor16:01:08

Maybe? Can you express in well-defined types?

Conor16:01:50

I found that actually writing proofs in Idris can be a right pain, as sometimes the compiler takes a lot of convincing that something is the case

pablore16:01:32

I dont know much about formal verification, but want to try to verify some code we use at work that's difficult to test

Conor16:01:33

If I wanted to formally verify a system, I'd probably look at TLA+ as a starting point

Conor12:01:17

Developed at my alma mater though, big up St Andrews 🙂

seancorfield17:01:29

Also where SASL came from -- and then Prof Turner went on to create KRC and Miranda.

pablore19:01:27

any nixos users here? How do I install an specific version of leiningen? nix-env -i nixos.leiningen installs 2.8.1 but I need 2.8.3.

mattly22:01:51

... just gonna assume this is spam

👍 8
mattly22:01:07

Mobile doesn’t load images automatically

boomboom22:01:56

@billydarkholme this isn't grindr. please stop

fellshard22:01:13

It's a spammer causing issues on several channels. Hopefully will be dealt with shortly.

Mario C.22:01:41

Yo what is this... At work right now

gjnoonan22:01:43

It's being dealt with

cfleming22:01:45

Last time this happened a new mod was added for this timezone.

robert.spurrier22:01:59

@mario.cordova.862 enough to get you fired lol

quang22:01:19

thought it wasn an off-topic emergency of some sorts

Mario C.22:01:34

It looked like it was about to autoplay too... I was like "/collapse"

micha22:01:46

"off-topic emergency" my old band

👍 1
fellshard22:01:52

Yeah, I've got inline images turned off on my preferences

eggsyntax22:01:28

User deleted. In a meeting, can't do more now, please feel free to PM me if it starts again.

👏 5
timgilbert22:01:47

Nice, thank you

Mario C.22:01:29

Maybe it was a bot written in clojure

😆 4
dpsutton22:01:47

there's another account posting porn in #clojure

timgilbert22:01:48

Actually, another user is trolling up #clojure

mattly22:01:02

Or an act of war from a Scala user :p

😂 2
duncanmak22:01:05

now it’s Conor

Mario C.22:01:09

yep... nsfw images

fellshard22:01:34

Keep in mind it's another impersonation

alice22:01:56

who is so sad and pathetic they hav eto try and disrupt a programming slack, like wtf

bballant22:01:18

That went bad pretty quickly, wow

fellshard22:01:41

I've seen it happen once before, it was a wave like this. Probably one person using an automated tool to create impersonations. Dunno who'd have beef, but they seem to use NSFW content and racially charged language to stir the pot.

mattly22:01:31

That slack isn’t really meant for “public” communities and doesn’t provide good tools for managing them doesn’t help anything either on that front

☝️ 2
micha22:01:42

naughty ascii art code blocks comenext probably

fellshard22:01:12

Wonder if the migration to Zulip would help any.

dpsutton22:01:42

zulip is immutable. which means you can't delete them 🙂

fellshard22:01:26

Any ability to mute by user on an individual basis, at least?

mattly22:01:28

But ... an admin could hide them? I’d hope?

fellshard22:01:32

Still not a good solution, but

mattly22:01:53

That seems like a very bad and potentially legally troublesome design decision on their part

dpsutton22:01:00

don't know. i stopped going there because it seemed dead

quang22:01:02

if it keeps happening, pretty sure @everyone can be disabled / only for admins

fellshard22:01:18

Hmm, that doesn't square with what I just saw; a couple messages just got deleted from the announcements Slack cross-post thread, there

dpsutton22:01:18

^ yeah. i thought it already was disabled too

fellshard22:01:24

It is, they're bypassing it

gjnoonan22:01:44

it is disabled, only admins and owners can

eggsyntax22:01:45

Been deleting the ones I see. If there are other porn/disruptive ones, please PM me, I'll happily delete. I just left the meeting I was in to try to keep this under control 😕

❤️ 4
😓 4
eval202022:01:31

> zulip is immutable. which means you can’t delete them As admin we can delete messages on Zulip 👍

eggsyntax22:01:32

It's tricky because they're impersonating real users, and their profile doesn't show their email address, so it's hard to distinguish.

☝️ 2
fellshard22:01:00

Time zone and unique user ID are available from their profile and can differentiate

eggsyntax22:01:21

Nice, thanks, I didn't see the user id.

eggsyntax22:01:11

Heh, can't see user id from the admin page. Sigh.

fellshard22:01:28

I'm not sure how useful it is directly, unfortunately, but it can still differentiate users from their impersonators, at least...

dpsutton22:01:17

ah ok. thanks for clarifying. I thought there was some issue about not being able to remove threads or something. sorry for the FUD 🙂

1
seancorfield22:01:06

Yeah, admins can download the entire list and find these folks near the bottom (most recently joined).

👍 3
fellshard22:01:58

Maybe a possibility of automated impersonator skimming?

mattly22:01:43

I really wish we had like IRC2 or something

mattly22:01:54

Yes I’m there

eval202022:01:17

ah, ok. It’s indeed less busy than Slack…

mattly22:01:32

“Group Chat for Work” and “Public Chat for a topic” have entirely different needs

mattly22:01:52

Slack aims for the former

eval202022:01:50

yeah, highly synchronous IMO

devn22:01:52

>"off-topic emergency" my old band

devn22:01:06

how you doin'? haven't seen you in awhile

micha22:01:44

@devn the dude abides! how's the life?

micha23:01:08

so what's all this i hear about prepls and taps and all kinds of awesome new clojure stuff?

micha23:01:34

just catching up, so many interesting things

seancorfield23:01:47

For the curious, although we have permissions locked down on @-ing groups, there's a known bug Slack will not fix that allows you to override this client-side if you tinker with the JS in your debugger 😞 Anyways, follow-up to #slack-help and/or #community-development at this point.

seancorfield23:01:36

@micha I'm using REBL in my everyday workflow now, it's nice to be able to visualize and drill-down into arbitrary data structures alongside my editor.

micha23:01:13

oh cool, i haven't tried it yet

micha23:01:59

by binary you mean like hexdump?

micha23:01:28

sorry disregard, my eyes read "binary" instead of "arbitrary", that was weird

seancorfield23:01:14

REBL uses the prepl internal (and will soon have taps connected up I expect).

micha23:01:49

> The UI is written in JavaFX

micha23:01:07

nice! i'm just working on a javafx thing, so far really enjoyable

duncanmak23:01:17

i’m looking to learn more about javafx too

duncanmak23:01:27

@micha are you running on JDK 11?

micha23:01:57

i'm on 1.8 currently, but i was thinking about making the upgrade soon

micha23:01:16

i don't think it will be a huge deal, do you?

duncanmak23:01:20

which JDK are people using here?

duncanmak23:01:40

@micha JDK 11 removed JavaFX from the base distribution, so you’ll have to download separate JARs

micha23:01:55

but the API is not substantially changed, right?

duncanmak23:01:03

afaik, that’s correct

micha23:01:12

it looked to be pretty compatible, but it's hard to tell, a lot of things going on in there

duncanmak23:01:28

yeah, with tools.deps, i guess it shouldn’t be too painful to add new jars

duncanmak23:01:41

i’ve been playing around with Kawa Scheme

micha23:01:54

haha i tried that too

micha23:01:01

they have the fx stuff in there

micha23:01:27

but i found it really easy to use from clojure

micha23:01:18

still just a toy but my experimentation is so far pretty fun: https://github.com/hoplon/hoplonfx

Robert Randolph23:01:46

darnit, I got a ping and thought someone cared about me... but no.

1
1
seancorfield23:01:32

@duncanmak We're on 8 -- we have some legacy software that won't run on 11 yet.

micha23:01:11

what kind of issues are you having with moving to 11?

seancorfield23:01:35

With the latest clj/`tools.deps` you can specify dependencies for JavaFX components and it will handle it properly. There was a bug preventing that before.

seancorfield23:01:37

@micha We rely on a few libraries that have not yet been made compatible with JDK 11 but the main problem for our legacy (non-Clojure) apps is that we're stuck on an old version that doesn't run on JDK 11 and newer versions break things in our app that we can't change easily. Luckily, we're retiring most of them within a month or two 🙂

micha23:01:22

sounds like a lot of work, i hope this isn't the new normal for java

seancorfield23:01:38

For Java 8 JavaFX a pain because OracleJDK includes it but OpenJDK doesn't and I haven't found a compatible download of it...

seancorfield23:01:02

@micha It was Java 9 that caused nearly all the breakage. 10 and 11 weren't too bad after that.