Logs: liberachat/#haskell
| 2021-08-24 12:35:57 | maerwald | buys more crypto |
| 2021-08-24 12:36:00 | <merijn> | You plan a bridge to last actual decades. Which means planning how the fuck you will be able to fix/maintain things latr |
| 2021-08-24 12:36:04 | × | markpythonicbtc quits (~textual@2601:647:5a00:35:f45e:90a7:31f1:d8ef) (Quit: My MacBook has gone to sleep. ZZZzzz…) |
| 2021-08-24 12:36:28 | <merijn> | If you wanna pretend your a software *engineer*, start planning how that shit will work *long* term, instead of this one sprint |
| 2021-08-24 12:36:34 | <Taneb> | Oh god is that a mood. "This'll go wrong in ten years", I say, "In ten years neither of us or this system will still be here" is the response |
| 2021-08-24 12:36:46 | <merijn> | That includes, updating libraries, handling libraries disappearing, whatever |
| 2021-08-24 12:37:20 | <maerwald> | tdammers: not sure it's worth much... the greatest coders I met were wizards, some of them did know nothing of CS. But generally, everyone who is afraid of complexity is a good fit |
| 2021-08-24 12:37:32 | <maralorn> | When building with ghc 9.0 (in nixpkgs) in blaze-textual 0.2.1 I get "Ambiguous occurrence ‘quotRemInteger’" between base 4.15 and integer-gmp 1.1. This is a dependency of ghcide so it’s supposed to be working but I can‘t figure out in which way I have wrong versions. Can anyone help me? |
| 2021-08-24 12:37:37 | <tdammers> | Reminds me of that story about programmers working on a cruise missile control system. They found a memory leak, but it was a very consistent and predictable one, so they just calculated how much memory the code would leak over the duration of a mission, doubled that, and just added that much more RAM to the system |
| 2021-08-24 12:37:50 | <merijn> | tdammers: tbh, I can respect that |
| 2021-08-24 12:38:08 | <maerwald> | well, they don't just update the compiler, I hope :p |
| 2021-08-24 12:38:33 | <tdammers> | merijn: thing is, IMO programming is nothing like engineering. The constraints are just so radically different that comparing the two disciplines is not awfully useful. |
| 2021-08-24 12:38:35 | <merijn> | maerwald: Missile guidance people don't change software versions in their pipeline |
| 2021-08-24 12:38:36 | <maralorn> | maralorn: The fact, that integer-gmp 1.1 is not even on hackage does not help. |
| 2021-08-24 12:38:52 | <merijn> | tdammers: I agree that software isn't very similar to engineering |
| 2021-08-24 12:39:07 | <tomsmeding> | maralorn: how does it depend on it then :p |
| 2021-08-24 12:39:15 | <merijn> | tdammers: I'm just saying the attitude of most programmers is shite and they could learn a lot from engineers in many things |
| 2021-08-24 12:39:19 | <tdammers> | maerwald: the point being, you don't need fancy garbage collection in a computer system that, by design, literally explodes after one program run |
| 2021-08-24 12:39:29 | <Logio_> | software isn't like engineering only because not enough people have been killed by bad software (yet) |
| 2021-08-24 12:39:32 | <tdammers> | merijn: agree on this much. |
| 2021-08-24 12:39:52 | <maralorn> | tomsmeding: I suspect integer-gmp 1.1 is the one bundled with ghc 9.0? |
| 2021-08-24 12:39:59 | <tdammers> | Logio_: no, it's fundamentally different because information systems have different dynamics than physical machines and structures |
| 2021-08-24 12:40:09 | → | acidjnk_new3 joins (~acidjnk@p200300d0c72b9534e4a445cd6b407eb5.dip0.t-ipconnect.de) |
| 2021-08-24 12:40:30 | <Taneb> | maralorn: I think the right solution here will be drop the dependency on integer-gmp on that version of GHC. |
| 2021-08-24 12:40:36 | × | sjb0 quits (~stephen@2001:8004:5170:3b13:8c1d:1f1c:70aa:a42e) (Quit: Leaving.) |
| 2021-08-24 12:40:42 | × | maroloccio quits (~marolocci@37.100.40.252) (Quit: WeeChat 3.0) |
| 2021-08-24 12:40:51 | <dminuoso> | Taneb: Id say "not security" is like gambling, but the jackbot has a negative sign attached to it. |
| 2021-08-24 12:41:07 | <Logio_> | tdammers: not in any fundamental sense, IMO. Software is about controlling something in the real world at the end of the day |
| 2021-08-24 12:41:11 | <Taneb> | dminuoso: reverse gambling, and yet, the house still always wins |
| 2021-08-24 12:41:24 | <Logio_> | the implementation details are not what defines engineering |
| 2021-08-24 12:41:27 | <tdammers> | Logio_: the difference is in the economics of the production process |
| 2021-08-24 12:41:33 | → | maroloccio joins (~marolocci@37.100.40.252) |
| 2021-08-24 12:42:11 | <maralorn> | Taneb: Sorry, I don‘t understand. Do you mean building ghc without gmp? |
| 2021-08-24 12:42:14 | <tomsmeding> | maralorn: is blaze-textual a dependency of your project? Are you building using cabal or stack? |
| 2021-08-24 12:42:35 | <tdammers> | Logio_: in industrial design, "production" is elaborate, expensive, and bound to severe physical constraints. In programming, the "production" part is a matter of running the build script. Changing an industrial production process is expensive; changing a deployment script is cheap. |
| 2021-08-24 12:42:43 | <Logio_> | tdammers: not the economics of the production, but of failure. As I said, once bad software starts killing enough people than things will change |
| 2021-08-24 12:42:54 | <dminuoso> | merijn: Regarding your bridge argument, it's my opinion that the main problem is that there's no liability for software. If you simply neglect security for hospital software, and this allows an attacker to turn off the entire ICU equipment, nobody goes to jail because of the dozens people that die. Nobody pays even a fine. |
| 2021-08-24 12:43:01 | <Taneb> | maralorn: no, I mean blaze-textual has a dependency on integer-gmp, but on newer GHCs I don't think this should be necessary |
| 2021-08-24 12:43:05 | <merijn> | dminuoso: Agreed |
| 2021-08-24 12:43:20 | <tdammers> | Logio_: no, the economics of production as well. It's just that people call the process of writing code "production", when that's really the equivalent of "design" in an industrial engineering process |
| 2021-08-24 12:43:44 | <maralorn> | tomsmeding: blaze-textual is a dependency of ghcide. I am building with nix which has the famous "not-doing-it" dependency resolution algorithm. |
| 2021-08-24 12:43:51 | <Taneb> | maralorn: the direct dependency shouldn't be necessary, I mean. There may still be an indirect dependency |
| 2021-08-24 12:44:02 | <maralorn> | Taneb: Ah, thanks. |
| 2021-08-24 12:44:26 | <Logio_> | tdammers: engineering cheap things is similar whether it's hardware or software, see chinese manufacturing |
| 2021-08-24 12:44:30 | ← | jakalx parts (~jakalx@base.jakalx.net) (Error from remote client) |
| 2021-08-24 12:44:37 | <Taneb> | maralorn: I reserve the right to be horribly wrong about this ;) |
| 2021-08-24 12:45:06 | <tomsmeding> | maralorn: the cabal.project in the haskell-language-server repo for ghc 9.0.1 has a specific git dependency on blaze-textual on a particular commit |
| 2021-08-24 12:45:09 | × | doyougnu quits (~user@c-73-25-202-122.hsd1.or.comcast.net) (Read error: Connection reset by peer) |
| 2021-08-24 12:45:21 | <tomsmeding> | maralorn: https://github.com/haskell/haskell-language-server/blob/master/cabal-ghc901.project#L33 |
| 2021-08-24 12:45:28 | <tomsmeding> | perhaps this is your problem |
| 2021-08-24 12:45:33 | × | shailangsa_ quits (~shailangs@host86-185-102-120.range86-185.btcentralplus.com) () |
| 2021-08-24 12:46:00 | <tdammers> | Logio_: even if you produce in China, changing your design in the middle of a production run is a lot more cumbersome and time consuming |
| 2021-08-24 12:46:12 | <tomsmeding> | maralorn: that seems to be exactly your problem |
| 2021-08-24 12:46:18 | × | maroloccio quits (~marolocci@37.100.40.252) (Client Quit) |
| 2021-08-24 12:46:22 | <Logio_> | tdammers: that's still just a quantitative difference, not a qualitative one |
| 2021-08-24 12:46:57 | <maerwald> | it's as if "agile" isn't an actual engineering practice :p |
| 2021-08-24 12:47:12 | <maralorn> | tomsmeding: Thanks! I’ll try passing integer-simple. |
| 2021-08-24 12:47:19 | <maerwald> | you don't build a sky-scraper agile |
| 2021-08-24 12:47:26 | <tdammers> | Logio_: it becomes quasi-qualitative when you consider that "production" is practically always a commodity in programming, and practically never in industrial production |
| 2021-08-24 12:47:30 | <tomsmeding> | maralorn: or using that fork of blaze-textual |
| 2021-08-24 12:47:58 | → | maroloccio joins (~marolocci@37.100.40.252) |
| 2021-08-24 12:48:10 | <maralorn> | tomsmeding: Since I am trying to upstream this into nixpkgs I‘d prefer to not do that to much. |
| 2021-08-24 12:49:10 | <Taneb> | maralorn: nixpkgs does sometimes do that |
| 2021-08-24 12:49:56 | <maralorn> | I will go there if there is no other way. |
| 2021-08-24 12:50:12 | <Taneb> | e.g. https://github.com/NixOS/nixpkgs/blob/master/pkgs/development/haskell-modules/configuration-ghc-8.10.x.nix#L74 (there's a file for ghc 9.0 as well) |
| 2021-08-24 12:51:39 | <maralorn> | Yeah, I am familiar with those files. |
| 2021-08-24 12:51:51 | <maralorn> | Working on the 9.0 one right now. |
| 2021-08-24 12:52:11 | <tomsmeding> | all too familiar? :p |
| 2021-08-24 12:53:04 | <Taneb> | OK, I'll leave you to it |
| 2021-08-24 12:55:15 | × | ac quits (~aloiscoch@2001:470:69fc:105::65) (Quit: You have been idle for 30+ days) |
| 2021-08-24 12:55:16 | × | hsiktas[m] quits (~hsiktasma@2001:470:69fc:105::30d4) (Quit: You have been idle for 30+ days) |
| 2021-08-24 12:55:34 | × | yin[m] quits (~zwromatri@2001:470:69fc:105::1d4) (Quit: You have been idle for 30+ days) |
| 2021-08-24 12:57:26 | AlexNoo_ | is now known as AlexNoo |
| 2021-08-24 12:57:47 | → | jakalx joins (~jakalx@base.jakalx.net) |
| 2021-08-24 13:00:53 | × | vs^ quits (~vs@12.16.129.111) (Ping timeout: 250 seconds) |
| 2021-08-24 13:05:31 | → | markpythonicbtc joins (~textual@50.228.44.6) |
| 2021-08-24 13:06:16 | × | bitdex quits (~bitdex@gateway/tor-sasl/bitdex) (Quit: = "") |
| 2021-08-24 13:09:34 | → | trcc_ joins (~trcc@users-5343.st.net.au.dk) |
| 2021-08-24 13:10:21 | <maerwald> | ok, now I'm confused... my code type checks if I move a function to a different module... |
| 2021-08-24 13:12:07 | × | trcc quits (~trcc@users-5343.st.net.au.dk) (Ping timeout: 240 seconds) |
| 2021-08-24 13:13:12 | <maerwald> | do module boundaries affect some type level hackery? |
| 2021-08-24 13:14:35 | × | hendursa1 quits (~weechat@user/hendursaga) (Quit: hendursa1) |
| 2021-08-24 13:14:59 | → | hendursaga joins (~weechat@user/hendursaga) |
| 2021-08-24 13:16:24 | × | burnsidesLlama quits (~burnsides@dhcp168-019.wadham.ox.ac.uk) (Remote host closed the connection) |
| 2021-08-24 13:16:51 | → | burnsidesLlama joins (~burnsides@dhcp168-019.wadham.ox.ac.uk) |
| 2021-08-24 13:18:17 | → | lavaman joins (~lavaman@98.38.249.169) |
| 2021-08-24 13:18:49 | → | shailangsa joins (~shailangs@host86-185-102-120.range86-185.btcentralplus.com) |
| 2021-08-24 13:19:44 | × | burnsidesLlama quits (~burnsides@dhcp168-019.wadham.ox.ac.uk) (Remote host closed the connection) |
| 2021-08-24 13:19:51 | → | burnsidesLlama joins (~burnsides@dhcp168-019.wadham.ox.ac.uk) |
| 2021-08-24 13:20:12 | × | burnsidesLlama quits (~burnsides@dhcp168-019.wadham.ox.ac.uk) (Remote host closed the connection) |
| 2021-08-24 13:20:43 | → | burnsidesLlama joins (~burnsides@dhcp168-019.wadham.ox.ac.uk) |
| 2021-08-24 13:20:45 | ← | felixphew parts (~felix@cssa/committee/felix) () |
| 2021-08-24 13:21:08 | × | trcc_ quits (~trcc@users-5343.st.net.au.dk) (Ping timeout: 250 seconds) |
| 2021-08-24 13:21:29 | → | vs^ joins (~vs@12.16.129.111) |
| 2021-08-24 13:22:33 | × | lavaman quits (~lavaman@98.38.249.169) (Ping timeout: 250 seconds) |
| 2021-08-24 13:24:21 | × | mcglk quits (~mcglk@131.191.49.120) (*.net *.split) |
| 2021-08-24 13:24:21 | × | awth13 quits (~awth13@user/awth13) (*.net *.split) |
| 2021-08-24 13:24:21 | × | Techcable quits (~Techcable@168.235.93.147) (*.net *.split) |
| 2021-08-24 13:24:21 | × | Dykam quits (Dykam@dykam.nl) (*.net *.split) |
| 2021-08-24 13:24:21 | × | polux quits (~polux@51.15.169.172) (*.net *.split) |
All times are in UTC.