Newest at the top
| 2026-01-08 23:15:24 +0100 | <TMA> | tomsmeding: I have encountered ring definition containing the axiom 0!=1 as well |
| 2026-01-08 23:15:20 +0100 | tomsmeding | doesn't follow; doesn't that assume a*0 = 0 from the get go? |
| 2026-01-08 23:14:36 +0100 | <monochrom> | for all b, b*(a*0) = (b*a)*0 = 0. Then appeal to uniqueness of 0: "if forall b b*foo=0, then foo=0" |
| 2026-01-08 23:14:24 +0100 | xff0x_ | (~xff0x@2405:6580:b080:900:cd9:802b:8b60:b254) |
| 2026-01-08 23:11:33 +0100 | <tomsmeding> | (and 0 = a * 0 because: 0 = 1 + -1 = a * a^-1 + -1 = a * (0 + a^-1) + -1 = a * 0 + a * a^-1 + -1 = a * 0 + 1 + -1 = a * 0 + 0 = a * 0; there's probably a simpler derivation lol) |
| 2026-01-08 23:11:19 +0100 | merijn | (~merijn@host-cl.cgnat-g.v4.dfn.nl) merijn |
| 2026-01-08 23:08:10 +0100 | <haskellbridge> | <loonycyborg> you can even divide by zero in this ring |
| 2026-01-08 23:07:12 +0100 | <tomsmeding> | if 1 = 0 then by the ring axioms, 0 = 0 * a = 1 * a = a, so all elements are zero, so it's the trivial ring, but it's allowed |
| 2026-01-08 23:06:49 +0100 | danz94513 | (~danza@user/danza) (Ping timeout: 246 seconds) |
| 2026-01-08 23:05:30 +0100 | <EvanR> | though we have good examples where they are, in crypto code |
| 2026-01-08 23:05:21 +0100 | <monochrom> | I agree. |
| 2026-01-08 23:04:37 +0100 | danza | (~danza@user/danza) danza |
| 2026-01-08 23:04:37 +0100 | <EvanR> | this whole time I was convinced that Num reflects a subcultural understanding of computer numbers and programmers are not usually thinking they're using a ring |
| 2026-01-08 23:04:32 +0100 | malte | (~malte@mal.tc) malte |
| 2026-01-08 23:04:26 +0100 | <geekosaur> | they're allowed to be the same… if there's only one value in the set |
| 2026-01-08 23:03:39 +0100 | <EvanR> | and a law saying they must be different? xD |
| 2026-01-08 23:03:31 +0100 | <EvanR> | zero and one |
| 2026-01-08 23:03:09 +0100 | malte | (~malte@mal.tc) (Ping timeout: 250 seconds) |
| 2026-01-08 23:00:58 +0100 | vanishingideal | (~vanishing@user/vanishingideal) vanishingideal |
| 2026-01-08 23:00:43 +0100 | Googulator82 | Googulator |
| 2026-01-08 22:59:19 +0100 | merijn | (~merijn@host-cl.cgnat-g.v4.dfn.nl) (Ping timeout: 240 seconds) |
| 2026-01-08 22:50:51 +0100 | <Leary> | monochrom: But sometimes that's the correct implementation. :) |
| 2026-01-08 22:50:25 +0100 | <monochrom> | You need a law to outlaw me trying "foldMap _ _ = mempty". |
| 2026-01-08 22:50:18 +0100 | <Leary> | Re `Num`, `abs` and `signum` should be moved to another class, and arguably `fromInteger` too (necessitating new `zero` and `one` methods). The rest could do to be split up or factored over `Monoid`, but it's otherwise fine and perfectly principled as a class for rings. |
| 2026-01-08 22:48:46 +0100 | merijn | (~merijn@host-cl.cgnat-g.v4.dfn.nl) merijn |
| 2026-01-08 22:43:58 +0100 | <Leary> | `Foldable` just doesn't need laws, since the type of `foldMap` and parametricity say it all. |
| 2026-01-08 22:43:18 +0100 | <mauke> | comfortably Num |
| 2026-01-08 22:42:33 +0100 | <EvanR> | at some point you heard a lot of Foldable has no laws, Foldable is essentially a place to put your toList function |
| 2026-01-08 22:42:02 +0100 | peterbecich | (~Thunderbi@71.84.33.135) peterbecich |
| 2026-01-08 22:41:43 +0100 | <tomsmeding> | I've ranted about Num before here |
| 2026-01-08 22:41:23 +0100 | <tomsmeding> | so that's a low bar |
| 2026-01-08 22:41:21 +0100 | <EvanR> | most things people try to write a Num instance for are contrived xD |
| 2026-01-08 22:41:19 +0100 | <tomsmeding> | Num was never principled or based in any way |
| 2026-01-08 22:41:08 +0100 | <EvanR> | (curiously) Foldable ended up being more principled and based than Num |
| 2026-01-08 22:40:32 +0100 | <tomsmeding> | but contrary to Foldable, where [a] is merely a common instance, such contrived Num instances are really contrived, I'd guess |
| 2026-01-08 22:40:09 +0100 | <tomsmeding> | there are probably contrived Num instances for which foldl' would be inappropriate |
| 2026-01-08 22:40:04 +0100 | <EvanR> | a long time ago |
| 2026-01-08 22:39:44 +0100 | <EvanR> | maybe, but there was a long thread on the mailing list defending the foldl version xD |
| 2026-01-08 22:39:09 +0100 | <tomsmeding> | that one was probably just an oversight |
| 2026-01-08 22:38:35 +0100 | <jreicher> | Oww. How did that happen in the first place? |
| 2026-01-08 22:38:33 +0100 | <EvanR> | good |
| 2026-01-08 22:38:25 +0100 | merijn | (~merijn@host-cl.cgnat-g.v4.dfn.nl) (Ping timeout: 264 seconds) |
| 2026-01-08 22:38:08 +0100 | <tomsmeding> | it is, but that was fixed; sum is now defined using foldl' |
| 2026-01-08 22:36:26 +0100 | trickard_ | trickard |
| 2026-01-08 22:36:25 +0100 | <EvanR> | this is probably dumb |
| 2026-01-08 22:36:08 +0100 | <lambdabot> | sum = foldl (+) 0 |
| 2026-01-08 22:36:08 +0100 | <EvanR> | @src sum |
| 2026-01-08 22:36:03 +0100 | <EvanR> | however the situation with sum |
| 2026-01-08 22:34:48 +0100 | spew | (~spew@user/spew) (Quit: nyaa~) |
| 2026-01-08 22:34:01 +0100 | Milan_Vanca | (~milan@user/Milan-Vanca:32634) (Quit: WeeChat 4.7.2) |