Logs: freenode/#haskell
| 2021-04-29 18:01:51 | → | vicfred joins (~vicfred@unaffiliated/vicfred) |
| 2021-04-29 18:01:52 | × | nut quits (~gtk@roc37-h01-176-170-197-243.dsl.sta.abo.bbox.fr) (Ping timeout: 240 seconds) |
| 2021-04-29 18:03:36 | × | rond_ quits (5940206b@89-64-32-107.dynamic.chello.pl) (Quit: Connection closed) |
| 2021-04-29 18:03:44 | <DigitalKiwi> | maerwald: nix shell nixpkgs#dateutils nixpkgs#jq -c bash -c 'ddiff "$(curl -s https://api.github.com/repos/input-output-hk/haskell.nix |jq -r '.created_at' )" "now" -f "%y"' |
| 2021-04-29 18:03:46 | <DigitalKiwi> | 3 |
| 2021-04-29 18:04:24 | <DigitalKiwi> | a lot's happened with just that project in the last year |
| 2021-04-29 18:04:47 | → | nut joins (~user@roc37-h01-176-170-197-243.dsl.sta.abo.bbox.fr) |
| 2021-04-29 18:05:45 | × | nut quits (~user@roc37-h01-176-170-197-243.dsl.sta.abo.bbox.fr) (Remote host closed the connection) |
| 2021-04-29 18:06:37 | × | ddellac__ quits (ddellacost@gateway/vpn/mullvad/ddellacosta) (Ping timeout: 265 seconds) |
| 2021-04-29 18:06:42 | <maerwald> | yep, 3 years sounds like the amount of time you need to invest to do anything non-trivial with it :p |
| 2021-04-29 18:07:57 | → | Qwerky joins (~qwerky@37.170.132.154) |
| 2021-04-29 18:08:00 | <DigitalKiwi> | https://input-output-hk.github.io/haskell.nix/tutorials/getting-started/ |
| 2021-04-29 18:08:33 | <DigitalKiwi> | https://haskell4nix.readthedocs.io/ |
| 2021-04-29 18:08:46 | <maerwald> | ? |
| 2021-04-29 18:09:19 | <maerwald> | You're suggesting one can make anything work with *just* reading the docs? |
| 2021-04-29 18:09:48 | <maerwald> | Usually you end up going on an issue-hunt through IRC, reddit, blog posts and obscure github gists |
| 2021-04-29 18:10:33 | <DigitalKiwi> | yeah that's just tech lol |
| 2021-04-29 18:10:37 | <maerwald> | no |
| 2021-04-29 18:10:54 | <maerwald> | that's poor ergonomics |
| 2021-04-29 18:11:03 | × | DavidEichmann quits (~david@156.59.147.147.dyn.plus.net) (Remote host closed the connection) |
| 2021-04-29 18:11:06 | <DigitalKiwi> | that so much has happened in the last few years is actually one of the problems, finding old stack overflows ;p |
| 2021-04-29 18:11:59 | <maerwald> | I've used ansible mostly by just reading the docs. It's not a great piece of tech, but it has better ergonomics. |
| 2021-04-29 18:12:26 | <DigitalKiwi> | and what i was suggesting was that at some point it was 'project nobody else could use' and 'project that is increasingly easy for a wide variety of other people to use' |
| 2021-04-29 18:12:51 | × | geekosaur quits (930099da@rrcs-147-0-153-218.central.biz.rr.com) (Quit: Connection closed) |
| 2021-04-29 18:12:52 | <DigitalKiwi> | how much nix docs have you read and related to what? |
| 2021-04-29 18:13:01 | <maerwald> | I think by now the only interesting use case that remains of nix is cross-compiling static binaries. |
| 2021-04-29 18:13:24 | <maerwald> | That can get a little hairy even with containerized alpine |
| 2021-04-29 18:13:29 | <DigitalKiwi> | there's actually quite a bit of nix docs, but sometimes it's hard to know which of the several manuals to look in |
| 2021-04-29 18:13:52 | <maerwald> | yes, poorly structured docs aren't very useful... hence why I asked: how would one teach nix? |
| 2021-04-29 18:14:11 | <maerwald> | I think packaging systems are too messy in general, some more than others |
| 2021-04-29 18:14:49 | <DigitalKiwi> | because nix the language, nix the program, nixos the distro, nixpkgs the expression set, are different things... |
| 2021-04-29 18:15:37 | <DigitalKiwi> | and then there's all of the things that have been built with it |
| 2021-04-29 18:16:42 | <maerwald> | yeah, too much complexity for the gains |
| 2021-04-29 18:16:49 | → | ADG1089 joins (~aditya@171.76.29.233) |
| 2021-04-29 18:17:37 | <DigitalKiwi> | like nixops, hydra, node2nix, home-manager, ... https://github.com/nix-community |
| 2021-04-29 18:17:49 | <monochrom> | I imagine that "ActiveX" would suffer the same problem. It really is a consortium of several not really related things that are grouped together only because some marketting people said to. |
| 2021-04-29 18:18:12 | <DigitalKiwi> | that's like saying you don't want to use haskell because edwardk already wrote a library you need |
| 2021-04-29 18:18:16 | <monochrom> | But this is where human nature is utterly broken. |
| 2021-04-29 18:18:29 | × | hypercube quits (~hypercube@2603-6011-f901-9e5b-0000-0000-0000-08cf.res6.spectrum.com) (Quit: WeeChat 3.1) |
| 2021-04-29 18:18:40 | <maerwald> | DigitalKiwi: abstractions in haskell are usually no that brittle |
| 2021-04-29 18:18:44 | <monochrom> | Human nature is such that they learn a mess and then they think it's a bragging right. |
| 2021-04-29 18:19:33 | <Clint> | that explains corporate culture |
| 2021-04-29 18:19:42 | <maerwald> | DigitalKiwi: it starts with nix the language being poor (module system anyone?) |
| 2021-04-29 18:20:03 | <monochrom> | People fall in love with both what they create and what they learn. These are specially cases of the sunk cost policy. |
| 2021-04-29 18:20:09 | × | ddellacosta quits (~ddellacos@ool-44c73afa.dyn.optonline.net) (Remote host closed the connection) |
| 2021-04-29 18:20:13 | <maerwald> | reasoning about 2 lines of nix code is already a challenge |
| 2021-04-29 18:20:18 | <DigitalKiwi> | i'm not even good at nix lol i just hit shit with hammers until it works |
| 2021-04-29 18:20:39 | <maerwald> | exactly, you don't reason about nix code, you just play with it until it does what you want |
| 2021-04-29 18:20:40 | <monochrom> | If you have wasted time, you can either admit it's a waste or rebrand it as an investment. Which one feels better? |
| 2021-04-29 18:20:46 | <Cale> | Yeah, nix-the-language not having a module system goes a fair way to explaining all of my own frustrations with it. Of course, I wouldn't be frustrated if nix wasn't actually irreplaceable in many other ways. |
| 2021-04-29 18:20:58 | <DigitalKiwi> | https://github.com/tazjin/nix-1p |
| 2021-04-29 18:21:00 | <monochrom> | And yes people have made this a policy, I didn't have a typo. |
| 2021-04-29 18:21:49 | → | z0 joins (~zzz@2a03:b0c0:3:d0::3095:3001) |
| 2021-04-29 18:21:53 | → | geekosaur joins (930099da@rrcs-147-0-153-218.central.biz.rr.com) |
| 2021-04-29 18:21:58 | <DigitalKiwi> | https://nixos.org/guides/nix-pills/index.html |
| 2021-04-29 18:22:06 | <maerwald> | but I don't see an alternative that will target the same audience |
| 2021-04-29 18:22:10 | <Cale> | It's really a style thing that is influenced by that lack of module system though: people make gigantic attrsets of things that behave a bit like modules, then they tweak those collections of definitions recursively, repeatedly, in ways that affect how things inside them behave deeply. |
| 2021-04-29 18:22:48 | <maerwald> | That's basically how OOP works |
| 2021-04-29 18:22:50 | <maerwald> | :) |
| 2021-04-29 18:23:11 | <Cale> | The end result is that even though the nix language is basically a pure functional programming language, you struggle to extract the benefit of that, because equational reasoning is only helpful if you can actually find the equations that define the things you're working with. |
| 2021-04-29 18:23:30 | → | malumore_ joins (~malumore@151.62.116.76) |
| 2021-04-29 18:23:34 | × | xff0x quits (~xff0x@2001:1a81:524d:ea00:6e93:d8:4f43:8c54) (Remote host closed the connection) |
| 2021-04-29 18:23:52 | → | xff0x joins (~xff0x@2001:1a81:524d:ea00:4bc5:b0ea:c9c9:30a8) |
| 2021-04-29 18:23:56 | <maerwald> | right, so what remains useful is nix the package manager istelf (the approach, that was used by stack, cabal, ...) |
| 2021-04-29 18:23:58 | → | ddellacosta joins (~ddellacos@ool-44c73afa.dyn.optonline.net) |
| 2021-04-29 18:24:23 | <maerwald> | you could just reboot it with a better language, but probably not gonna happen soon |
| 2021-04-29 18:24:49 | <Cale> | and all the descriptions of how to build every package, and the fact that you can set up reasonably consistent environments between developer machines and deployed machines |
| 2021-04-29 18:24:52 | × | merijn quits (~merijn@83-160-49-249.ip.xs4all.nl) (Ping timeout: 240 seconds) |
| 2021-04-29 18:25:00 | <Cale> | all that stuff is great |
| 2021-04-29 18:25:24 | × | zzz quits (~zzz@2a03:b0c0:3:d0::3095:3001) (Ping timeout: 258 seconds) |
| 2021-04-29 18:25:25 | z0 | is now known as zzz |
| 2021-04-29 18:25:53 | <Cale> | Nix the desktop OS, not as great imo -- it takes a lot of fiddling to get things into a usable configuration, so I don't even bother with it. Once you have a configuration you like, deploying it to more machines is easy though, which I have to admit is cool. |
| 2021-04-29 18:26:23 | <Cale> | But it feels a lot like fucking around trying to get Slackware to work in 1999. |
| 2021-04-29 18:26:27 | <DigitalKiwi> | https://logs.nix.samueldr.com/nixos/2021-04-28#1619579644-1619575351; |
| 2021-04-29 18:26:58 | <monochrom> | slackware :) |
| 2021-04-29 18:27:06 | <maerwald> | Yep, nixOS feels like gentoo, except with less QA |
| 2021-04-29 18:27:27 | → | timCF joins (~i.tkachuk@m91-129-104-226.cust.tele2.ee) |
| 2021-04-29 18:28:09 | <maerwald> | I think in the beginning, because there was too much community contribution to establish strict QA policies |
| 2021-04-29 18:28:11 | <Cale> | But yeah, I love the ability to quickly set up a default.nix for my projects such that I can nix-shell and be in an environment with whatever dependencies (both Haskell and system) that I'd need to work on or build the project. |
| 2021-04-29 18:28:25 | <maerwald> | And later, because ppl stopped using it as a desktop OS |
| 2021-04-29 18:28:30 | <Cale> | That's cool, and worth learning nix for. |
| 2021-04-29 18:28:38 | <DigitalKiwi> | well i am https://twitter.com/ArchKiwi |
| 2021-04-29 18:31:31 | <DigitalKiwi> | https://logs.nix.samueldr.com/nixos-chat/2021-03-14#1615715873-1615710746 |
| 2021-04-29 18:32:09 | <carbolymer> | >using twatter |
| 2021-04-29 18:32:45 | × | timCF quits (~i.tkachuk@m91-129-104-226.cust.tele2.ee) (Quit: leaving) |
| 2021-04-29 18:33:09 | × | rj quits (~x@gateway/tor-sasl/rj) (Ping timeout: 240 seconds) |
| 2021-04-29 18:34:27 | <DigitalKiwi> | https://logs.nix.samueldr.com/nixos/2021-04-25#1619383503-1619376013; maerwald i complain a lot too ;) |
| 2021-04-29 18:39:30 | → | dmytrish joins (~mitra@2a02:8084:a82:d900:61fb:ae9:64fb:f0ec) |
| 2021-04-29 18:39:58 | × | ram19890 quits (~ram@49.205.84.15) (Quit: Konversation terminated!) |
| 2021-04-29 18:40:03 | → | gnumonic joins (~gnumonic@c-73-170-91-210.hsd1.ca.comcast.net) |
| 2021-04-29 18:42:12 | → | mikoto-chan joins (~mikoto-ch@gateway/tor-sasl/mikoto-chan) |
| 2021-04-29 18:42:34 | → | ddellac__ joins (~ddellacos@86.106.143.203) |
| 2021-04-29 18:43:03 | → | nut joins (~nut@roc37-h01-176-170-197-243.dsl.sta.abo.bbox.fr) |
| 2021-04-29 18:44:58 | → | hypercube joins (~hypercube@2603-6011-f901-9e5b-0000-0000-0000-08cf.res6.spectrum.com) |
| 2021-04-29 18:45:26 | <DigitalKiwi> | monochrom: well the thing is that in spite of all of the sunk cost (and there is lots) there're still too many benefits that it does bring that nothing else solves as easily/at all (yet, if something better does come along, which i hope it does, because it does have problems, someday) |
| 2021-04-29 18:45:43 | × | ukari quits (~ukari@unaffiliated/ukari) (Remote host closed the connection) |
| 2021-04-29 18:46:34 | → | ukari joins (~ukari@unaffiliated/ukari) |
| 2021-04-29 18:46:52 | × | ddellac__ quits (~ddellacos@86.106.143.203) (Ping timeout: 240 seconds) |
| 2021-04-29 18:47:07 | <DigitalKiwi> | i'm not married to nix..not even that good at it as in i haven't really tried to learn it that much... probably part of my prolem ;p |
| 2021-04-29 18:47:38 | <DigitalKiwi> | https://patrickmn.com/software/the-haskell-pyramid/ |
All times are in UTC.