Logs: liberachat/#haskell
| 2021-06-17 07:20:42 | → | kuribas joins (~user@ptr-25vy0i9z0t743gg6aow.18120a2.ip6.access.telenet.be) |
| 2021-06-17 07:21:30 | × | warnz quits (~warnz@2600:1700:77c0:5610:7144:467c:eae6:37e7) (Ping timeout: 240 seconds) |
| 2021-06-17 07:22:44 | → | gehmehgeh joins (~user@user/gehmehgeh) |
| 2021-06-17 07:23:05 | → | Lycurgus joins (~juan@cpe-45-46-140-49.buffalo.res.rr.com) |
| 2021-06-17 07:25:24 | <kuribas> | I wrote about a 1000 lines of code, with the model and queries. The queries worked from the first time when testing on the REPL. |
| 2021-06-17 07:25:30 | × | BosonCollider quits (~olofs@90-227-86-119-no542.tbcn.telia.com) (Ping timeout: 268 seconds) |
| 2021-06-17 07:25:58 | <kuribas> | I am known as the "crazy haskell zealot", I'd like to see them show this in another language... |
| 2021-06-17 07:26:21 | <kuribas> | I don't get code right the first time in lisp/java/python/... |
| 2021-06-17 07:26:37 | <kuribas> | we don't even get code right in production. |
| 2021-06-17 07:26:39 | × | monochrom quits (trebla@216.138.220.146) (Quit: NO CARRIER) |
| 2021-06-17 07:27:00 | <kuribas> | Stuff just breaks after refactorings, dispite a currently very large testing setup. |
| 2021-06-17 07:27:00 | <Lycurgus> | not safety critical i presume |
| 2021-06-17 07:27:16 | <kuribas> | Lycurgus: it isn't. |
| 2021-06-17 07:27:26 | <kuribas> | Lycurgus: but if you ask me, it costs the company money. |
| 2021-06-17 07:27:52 | × | spirgel quits (spirgel@gateway/vpn/protonvpn/spirgel) () |
| 2021-06-17 07:28:02 | <kuribas> | Since bugs affect the frontend team (which is outsourced), takes more time to communicate those bugs, delays the product for the customer, who only pays when they get the feature. |
| 2021-06-17 07:28:05 | × | dunkeln quits (~dunkeln@94.129.65.28) (Ping timeout: 252 seconds) |
| 2021-06-17 07:28:11 | <Lycurgus> | well not breaking stuff in production ofc predates hs |
| 2021-06-17 07:28:50 | <kuribas> | Lycurgus: if you spend a lot of time testing before releasing... |
| 2021-06-17 07:28:54 | <kuribas> | We have a short cycle. |
| 2021-06-17 07:29:36 | <Lycurgus> | that and move fast and break stuff is or just was an ethos |
| 2021-06-17 07:30:28 | <kuribas> | I like that philosophy. I just don't like when it breaks because the tooling, language or libraries are poor. |
| 2021-06-17 07:30:39 | <Lycurgus> | believe it or don't software development in general is much more reliable and generally successful at the project level than it was a generation ago |
| 2021-06-17 07:30:57 | <kuribas> | because of move fast? |
| 2021-06-17 07:30:58 | <Lycurgus> | that with a concomittant narrowing of scope |
| 2021-06-17 07:31:16 | <Lycurgus> | no it's just maturity of the field |
| 2021-06-17 07:31:45 | <Lycurgus> | advance and intensification and less wild ass shit |
| 2021-06-17 07:31:55 | <kuribas> | sure, I think being late because you refactor code to be clean, or because you don't have well defined scope and overly generalize, are two very different things. |
| 2021-06-17 07:32:10 | <kuribas> | The first isn't really "being late". |
| 2021-06-17 07:32:35 | <Lycurgus> | well late used to mean never a lot of times |
| 2021-06-17 07:33:07 | × | gehmehgeh quits (~user@user/gehmehgeh) (Remote host closed the connection) |
| 2021-06-17 07:33:09 | <kuribas> | I mean, I get this explanation, types and abstractions are overhead and delay the project, because you don't focus on business logic. |
| 2021-06-17 07:33:24 | → | cfricke joins (~cfricke@user/cfricke) |
| 2021-06-17 07:33:26 | <kuribas> | But that argument is flawed, because your abstraction and types should flow from the business logic. |
| 2021-06-17 07:33:28 | <Lycurgus> | in the late nineties I worked on a major refactoring of telephone systems customer service systems |
| 2021-06-17 07:33:40 | <Lycurgus> | it was a billion dollar project |
| 2021-06-17 07:33:47 | → | gehmehgeh joins (~user@user/gehmehgeh) |
| 2021-06-17 07:33:48 | <Lycurgus> | that failed |
| 2021-06-17 07:34:20 | <Lycurgus> | by the predecessor of verizon (nynex) |
| 2021-06-17 07:35:36 | → | chele joins (~chele@user/chele) |
| 2021-06-17 07:35:55 | <Lycurgus> | well the concerns of the hs community are not in general those of business |
| 2021-06-17 07:36:39 | <Lycurgus> | which has achieved the end goals of reliability and so forth by more traditional engineering practicss |
| 2021-06-17 07:37:14 | <kuribas> | reliability for us seem to mean mostly, iterate until there are no more bugs. |
| 2021-06-17 07:37:23 | <Lycurgus> | and continues to do so |
| 2021-06-17 07:38:14 | <kuribas> | sure, the initial effort to model the program is often harder with static types, and takes a bit more time. |
| 2021-06-17 07:39:01 | <kuribas> | By harder, I mean doing it type safe, where in java or lisp you would use NULL, or nil, partial hashmaps etc. |
| 2021-06-17 07:39:37 | → | dunkeln joins (~dunkeln@94.129.65.28) |
| 2021-06-17 07:39:50 | <Lycurgus> | that paper by sm oder on OOHaskell and dynamic exception types, if you know of it, in what sense are the types dynamic? |
| 2021-06-17 07:40:03 | → | pera joins (~pera@70.red-88-14-152.dynamicip.rima-tde.net) |
| 2021-06-17 07:40:26 | pera | is now known as Guest8423 |
| 2021-06-17 07:40:31 | <kuribas> | I don't know OOHaskell |
| 2021-06-17 07:40:46 | <Lycurgus> | ah, it's from the turn of the century |
| 2021-06-17 07:41:01 | <kuribas> | Java is technically statically typed, but it leaves a lot of room for runtime errors. |
| 2021-06-17 07:41:15 | <Lycurgus> | b4 c. '06 or so when hs started to become really practical |
| 2021-06-17 07:41:32 | <Lycurgus> | and didn need going forward to suck up to OO |
| 2021-06-17 07:41:50 | <kuribas> | IMO it's not haskell, but composability which makes your programs robust. Haskell is just a language that makes composability much easier. Java makes it rather hard. |
| 2021-06-17 07:41:55 | <kuribas> | Objects don't compose well. |
| 2021-06-17 07:42:46 | × | wroathe quits (~wroathe@c-68-54-25-135.hsd1.mn.comcast.net) (Ping timeout: 268 seconds) |
| 2021-06-17 07:42:49 | <Lycurgus> | hs, erlang, ocaml in rough order of popularity |
| 2021-06-17 07:43:07 | <Lycurgus> | it's still a small fraction of global software production |
| 2021-06-17 07:43:21 | <Lycurgus> | maybe 1 or 2 % tops |
| 2021-06-17 07:43:38 | <Lycurgus> | for all of em together |
| 2021-06-17 07:43:58 | <kuribas> | I am not saying reliable software engineering is not possible in mainstream language, I am just saying it is easier in haskell. |
| 2021-06-17 07:44:19 | <Lycurgus> | easier for you |
| 2021-06-17 07:44:34 | → | merijn joins (~merijn@83-160-49-249.ip.xs4all.nl) |
| 2021-06-17 07:44:42 | <Lycurgus> | easier in principal for the pure fps |
| 2021-06-17 07:44:55 | <Lycurgus> | easier in fact and practice is another matter |
| 2021-06-17 07:45:04 | <Lycurgus> | *in principle |
| 2021-06-17 07:45:15 | <kuribas> | yeah, it may not be for everyone. |
| 2021-06-17 07:45:21 | <kuribas> | But you don't know unless you try. |
| 2021-06-17 07:45:52 | <kuribas> | but software engineering is hard in any language, ecosystem. |
| 2021-06-17 07:46:06 | <kuribas> | A lot of ecosystems just pretend it is easy. Haskell doesn't. |
| 2021-06-17 07:46:17 | <Lycurgus> | a thing which would be decisive and which I have intentions for |
| 2021-06-17 07:46:34 | <Lycurgus> | is if low level programming, end programming is automated |
| 2021-06-17 07:46:47 | → | fizbin joins (~fizbin@c-68-83-100-68.hsd1.nj.comcast.net) |
| 2021-06-17 07:46:54 | × | slice quits (~slice@user/slice) (Quit: zzz) |
| 2021-06-17 07:46:54 | → | ddellacosta joins (~ddellacos@86.106.121.100) |
| 2021-06-17 07:46:57 | <kuribas> | Monads and Applicatives sound like adding complexity on top of it, but in fact they allow you to reduce complexity. |
| 2021-06-17 07:47:02 | <Lycurgus> | in that case training programmers for hs is eliminated as an issue |
| 2021-06-17 07:48:18 | × | stefan-__ quits (~cri@42dots.de) (Ping timeout: 264 seconds) |
| 2021-06-17 07:48:32 | <kuribas> | training for hs is an investment. |
| 2021-06-17 07:48:40 | <kuribas> | it's a nonzero cost indeed. |
| 2021-06-17 07:48:50 | <kuribas> | but companies have money for training. |
| 2021-06-17 07:49:00 | <Lycurgus> | like in G/Labview, where people who've pinned their careers to it don't know or care how it produces code |
| 2021-06-17 07:49:02 | → | warnz joins (~warnz@2600:1700:77c0:5610:7144:467c:eae6:37e7) |
| 2021-06-17 07:49:56 | <Lycurgus> | they have money for training but not the kind of reeducation hs calls for |
| 2021-06-17 07:51:05 | <kuribas> | I think we can train for anything, as long as we show it helps for the job. |
| 2021-06-17 07:51:08 | × | fizbin quits (~fizbin@c-68-83-100-68.hsd1.nj.comcast.net) (Ping timeout: 244 seconds) |
| 2021-06-17 07:51:29 | <Lycurgus> | maybe in central eu oder, not here (US) |
| 2021-06-17 07:51:55 | × | tzh quits (~tzh@c-24-21-73-154.hsd1.or.comcast.net) (Quit: zzz) |
| 2021-06-17 07:52:01 | × | ddellacosta quits (~ddellacos@86.106.121.100) (Ping timeout: 268 seconds) |
| 2021-06-17 07:52:09 | → | stefan-_ joins (~cri@42dots.de) |
| 2021-06-17 07:52:41 | <kuribas> | US is crazy |
| 2021-06-17 07:52:44 | → | monochrom joins (trebla@216.138.220.146) |
| 2021-06-17 07:52:50 | <Lycurgus> | ikr? |
| 2021-06-17 07:53:12 | × | warnz quits (~warnz@2600:1700:77c0:5610:7144:467c:eae6:37e7) (Ping timeout: 244 seconds) |
| 2021-06-17 07:53:17 | <Lycurgus> | crazy ol bastards |
| 2021-06-17 07:53:23 | <kuribas> | I would just be independend in the US, or join a haskell company, rather than hoping for them to accept haskell. |
| 2021-06-17 07:53:36 | → | neceve joins (~quassel@2a02:c7f:607e:d600:a95a:ecd2:e57a:3130) |
| 2021-06-17 07:54:22 | <Lycurgus> | well FP is huge right now and hs is known as the first among equals in it |
All times are in UTC.