|
|
| code logs -> 2024 -> Mon, 17 Jun 2024 | < code.20240616.log - code.20240618.log > |
| --- Log opened Mon Jun 17 00:00:03 2024 |
| 00:13 | | Emmy [Emmy@Nightstar-qo29c7.fixed.kpn.net] has quit [Ping timeout: 121 seconds] |
| 00:33 | | Vornicus [Vorn@Nightstar-8k6f0b.res.spectrum.com] has quit [Connection closed] |
| 03:10 | | Degi [Degi@Nightstar-hdo1ek.pool.telefonica.de] has quit [Ping timeout: 121 seconds] |
| 03:12 | | Degi [Degi@Nightstar-ekkac3.pool.telefonica.de] has joined #code |
| 10:31 | | bluefoxx [fuzzylombax@Nightstar-gmk.ubj.180.108.IP] has quit [Connection closed] |
| 10:32 | | bluefoxx [fuzzylombax@Nightstar-gmk.ubj.180.108.IP] has joined #code |
| 11:56 | | Kindamoody [Kindamoody@Nightstar-pqh9gl.tbcn.telia.com] has joined #code |
| 11:56 | | mode/#code [+o Kindamoody] by ChanServ |
| 12:30 | <@gnolam> | Paper title of the month goes to "ChatGPT is bullshit": https://link.springer.com/article/10.1007/s10676-024-09775-5 |
| 14:51 | | Vornicus [Vorn@Nightstar-8k6f0b.res.spectrum.com] has joined #code |
| 14:51 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 15:24 | | Kindamoody [Kindamoody@Nightstar-pqh9gl.tbcn.telia.com] has quit [Ping timeout: 121 seconds] |
| 15:35 | | Vornicus [Vorn@Nightstar-8k6f0b.res.spectrum.com] has quit [Connection closed] |
| 17:02 | | Emmy [Emmy@Nightstar-qo29c7.fixed.kpn.net] has joined #code |
| 19:15 | | Kindamoody [Kindamoody@Nightstar-pqh9gl.tbcn.telia.com] has joined #code |
| 19:15 | | mode/#code [+o Kindamoody] by ChanServ |
| 19:41 | < Alek> | thanks gnolam |
| 19:48 | <&McMartin> | Bullshit in the philosophical sense, no less |
| 19:49 | <&McMartin> | This amuses me because in my hobby project lately I fell afoul of an optimizing compiler that went berzerk and not only converted one of my FOR X=1 TO 100 loops into an infinite loop, it altered the logic of my "log current state" debugging so that flatly impossible stuff would happen |
| 19:50 | <&McMartin> | This was a consequence of rigid application of a set of rules of a mechanical theorem-prover, but it turns out you *really can't* talk about the behavior of a theorem prover without some way of saying that it knows and deduces things. |
| 19:50 | <&McMartin> | You can write around it -- I tried -- but it's very obvious that you're writing around a taboo set of words to not say that the theorem prover "deduces" the things that you are running the goddamn program to deduce. |
| 19:51 | <&McMartin> | Obligatory: https://staging.cohostcdn.org/attachment/e2d6d929-4692-4cc9-bc92-b67c12ab76dc/funsafe-math-optimizations-meme.png |
| 19:52 | <&McMartin> | But this wasn't that; it was an unintuitive consequence of the assumption "assume the program is valid", which optimizers are allowed to do |
| 21:35 | <&Reiver> | Speaking of algorathm optimisations: https://www.smbc-comics.com/comic/algorithm |
| 21:42 | <&McMartin> | Speaking also of "the program technically does not say what it obviously says," I was recently fixing up warnings in some old code, and one of the warning flags was this command: |
| 21:42 | <&McMartin> | return "<????>"; |
| 21:43 | <&McMartin> | Because C has a decrepit feature that I think they're getting around to removing from the language where ??< and ??> were used to replace { and } for those machines whose character sets lacked braces |
| 21:43 | <&McMartin> | So that got the warning "You, uh, you didn't _really_ want trigraphs here, did you? If you did, you'd better compile with this compiler option these days" |
| 21:43 | <&McMartin> | (Or, you can make the warning go away by escaping the last question mark <_<) |
| 22:11 | | Vornicus [Vorn@Nightstar-8k6f0b.res.spectrum.com] has joined #code |
| 22:11 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 22:13 | | Kindamoody [Kindamoody@Nightstar-pqh9gl.tbcn.telia.com] has quit [Ping timeout: 121 seconds] |
| 22:31 | | Emmy [Emmy@Nightstar-qo29c7.fixed.kpn.net] has quit [Connection closed] |
| --- Log closed Tue Jun 18 00:00:05 2024 |
|
| code logs -> 2024 -> Mon, 17 Jun 2024 | < code.20240616.log - code.20240618.log > |
|
|