Newest at the top
| 2025-12-26 05:26:07 +0100 | <haskellbridge> | <Zemyla> You could potentially have a Map Monad instance, since the Ord constraint is on k, not a. The main problem is it doesn't let you implement pure. |
| 2025-12-26 05:21:37 +0100 | tabemann | (~tabemann@12.215.215.61) (Ping timeout: 264 seconds) |
| 2025-12-26 05:21:13 +0100 | merijn | (~merijn@host-cl.cgnat-g.v4.dfn.nl) merijn |
| 2025-12-26 05:19:24 +0100 | humasect | (~humasect@dyn-192-249-132-90.nexicom.net) humasect |
| 2025-12-26 05:14:10 +0100 | acidjnk | (~acidjnk@p200300d6e7171922bda2454d8bce8e38.dip0.t-ipconnect.de) (Ping timeout: 255 seconds) |
| 2025-12-26 05:10:24 +0100 | merijn | (~merijn@host-cl.cgnat-g.v4.dfn.nl) (Ping timeout: 252 seconds) |
| 2025-12-26 05:05:49 +0100 | merijn | (~merijn@host-cl.cgnat-g.v4.dfn.nl) merijn |
| 2025-12-26 04:54:14 +0100 | bitdex | (~bitdex@gateway/tor-sasl/bitdex) (Remote host closed the connection) |
| 2025-12-26 04:53:43 +0100 | merijn | (~merijn@host-cl.cgnat-g.v4.dfn.nl) (Ping timeout: 244 seconds) |
| 2025-12-26 04:49:04 +0100 | merijn | (~merijn@host-cl.cgnat-g.v4.dfn.nl) merijn |
| 2025-12-26 04:45:59 +0100 | kaskal | (~kaskal@2a02:8388:1a8b:7d00:5666:a937:cb77:6c3b) kaskal |
| 2025-12-26 04:44:33 +0100 | emmanuelux | (~emmanuelu@user/emmanuelux) (Remote host closed the connection) |
| 2025-12-26 04:44:32 +0100 | kaskal | (~kaskal@84-115-235-223.cable.dynamic.surfer.at) (Ping timeout: 240 seconds) |
| 2025-12-26 04:44:28 +0100 | emmanuelux_ | (~emmanuelu@user/emmanuelux) emmanuelux |
| 2025-12-26 04:38:25 +0100 | merijn | (~merijn@host-cl.cgnat-g.v4.dfn.nl) (Ping timeout: 264 seconds) |
| 2025-12-26 04:33:33 +0100 | merijn | (~merijn@host-cl.cgnat-g.v4.dfn.nl) merijn |
| 2025-12-26 04:30:10 +0100 | <glguy> | Probably the easiest starting point is to use to list and memoize with a list argument |
| 2025-12-26 04:29:18 +0100 | <Leary> | Though `splitRoot` is essentially a look into the internals, so you can probably hack a `HasTrie` instance with it. |
| 2025-12-26 04:27:38 +0100 | <geekosaur> | same reason why you can't have `Set` or `Map` `Monad` instances |
| 2025-12-26 04:27:01 +0100 | <geekosaur> | that said, a more practical reason for not having `Set` or `Map` instances is that instances that require additional constraints have severe limitations when possible at all, and both those instances would require an added `Ord` instance |
| 2025-12-26 04:26:52 +0100 | <Leary> | iqubic: The approach that 'MemoTrie' and related pure approaches take is to index into a lazy data type with one position for each possible argument (holding the result of applying the function). If you can design such a type for `Set` or `Map` then you can use it, but you'll likely be stymied by the fact that they're abstract types. |
| 2025-12-26 04:25:56 +0100 | <geekosaur> | large lists are kinda their own punishment 🙂 |
| 2025-12-26 04:24:23 +0100 | <iqubic> | Why does memoizing over a list make sense? |
| 2025-12-26 04:23:19 +0100 | <iqubic> | Yeah, that does make sense. |
| 2025-12-26 04:22:31 +0100 | merijn | (~merijn@host-cl.cgnat-g.v4.dfn.nl) (Ping timeout: 264 seconds) |
| 2025-12-26 04:21:21 +0100 | <geekosaur> | the other issue is that a large `Set` or `Map` would function in that case as an arbitrary number of "parameters", which would make memoization more expensive possibly to the point that it would have diminishing returns |
| 2025-12-26 04:20:05 +0100 | pebble | (~pebble@37.63.33.138) pebble |
| 2025-12-26 04:15:30 +0100 | merijn | (~merijn@host-cl.cgnat-g.v4.dfn.nl) merijn |
| 2025-12-26 04:15:23 +0100 | <iqubic> | Well, I'm using this 'Map' to store the state of the system I'm modeling. |
| 2025-12-26 04:14:21 +0100 | <geekosaur> | I don't usually have much use for memoization, and the fact that most memoizing functions use `Map` internally makes me wonder if you're just asking for the wrong thing to begin with |
| 2025-12-26 04:14:16 +0100 | <iqubic> | Hmm... that's annoying. MemoTrie looks good, but it doesn't have instances for Set or Map... |
| 2025-12-26 04:13:31 +0100 | <geekosaur> | not off the top of my head |
| 2025-12-26 04:13:18 +0100 | <iqubic> | geekosaur: Do you know if there are any off the shelf ways to memoize functions taking Sets or Maps? |
| 2025-12-26 04:11:49 +0100 | <lambdabot> | https://hackage.haskell.org/package/MemoTrie |
| 2025-12-26 04:11:49 +0100 | <iqubic> | @Hackage MemoTrie |
| 2025-12-26 04:11:25 +0100 | <iqubic> | I was trying to use MemoTrie, but that doesn't seem to work... |
| 2025-12-26 04:04:15 +0100 | merijn | (~merijn@host-cl.cgnat-g.v4.dfn.nl) (Ping timeout: 240 seconds) |
| 2025-12-26 04:00:03 +0100 | merijn | (~merijn@host-cl.cgnat-g.v4.dfn.nl) merijn |
| 2025-12-26 03:58:15 +0100 | humasect | (~humasect@dyn-192-249-132-90.nexicom.net) (Remote host closed the connection) |
| 2025-12-26 03:42:11 +0100 | <Leary> | iqubic: Presumably you can find this on hackage somewhere, but: https://gist.github.com/LSLeary/33a854e1ec1230e7fbcc09e44e1ee71d |
| 2025-12-26 03:33:43 +0100 | merijn | (~merijn@host-cl.cgnat-g.v4.dfn.nl) (Ping timeout: 240 seconds) |
| 2025-12-26 03:29:16 +0100 | merijn | (~merijn@host-cl.cgnat-g.v4.dfn.nl) merijn |
| 2025-12-26 03:23:43 +0100 | annamalai | (~annamalai@157.32.199.28) (Ping timeout: 240 seconds) |
| 2025-12-26 03:22:38 +0100 | bitdex | (~bitdex@gateway/tor-sasl/bitdex) bitdex |
| 2025-12-26 03:15:43 +0100 | merijn | (~merijn@host-cl.cgnat-g.v4.dfn.nl) (Ping timeout: 240 seconds) |
| 2025-12-26 03:13:47 +0100 | bitdex | (~bitdex@gateway/tor-sasl/bitdex) (Ping timeout: 252 seconds) |
| 2025-12-26 03:11:09 +0100 | polykernel_ | polykernel |
| 2025-12-26 03:11:09 +0100 | polykernel | (~polykerne@user/polykernel) (Ping timeout: 260 seconds) |
| 2025-12-26 03:11:07 +0100 | merijn | (~merijn@host-cl.cgnat-g.v4.dfn.nl) merijn |
| 2025-12-26 03:08:43 +0100 | polykernel_ | (~polykerne@user/polykernel) polykernel |