[00:52:13] The coding box is too small and there's no way to make it bigger. This is my fullscreen browser. : https://tools-static.wmflabs.org/bridgebot/257ae39e/file_68078.jpg [00:54:01] Disable limited width mode in settings? (re @Toby: The coding box is too small and there's no way to make it bigger. This is my fullscreen browser.) [00:55:09] I should learn how to do that! But also I'm more after a design solution that helps most users. [00:55:26] Zoom in? [00:55:43] Reset zoom? [00:56:39] I can read everything in the box, but I want a bigger box so that more of the implementation fits in the box. (re @Feeglgeef: Zoom in?) [00:57:24] User CSS? (re @Toby: I can read everything in the box, but I want a bigger box so that more of the implementation fits in the box.) [00:58:43] Class is: ext-wikilambda-app-code__code-editor [01:05:24] It’s not exactly “mobile first” 😏 (re @Toby: I should learn how to do that! But also I'm more after a design solution that helps most users.) [01:05:58] I put something in my user CSS if you'd like to grab and change it (re @Toby: The coding box is too small and there's no way to make it bigger. This is my fullscreen browser.) [01:12:28] I wrote entire sentences with this coding box, and there was a way to change that NOOOO (re @Toby: The coding box is too small and there's no way to make it bigger. This is my fullscreen browser.) [01:44:13] Yes, I understand that is why this has happened. But isn't it possible to make something that works okay for both? (re @Al: It’s not exactly “mobile first” 😏) [01:44:44] Yeah (re @Toby: Yes, I understand that is why this has happened. But isn't it possible to make something that works okay for both?) [04:24:21] Was doing some translations on MediaWiki and this came up hahaha lol : https://tools-static.wmflabs.org/bridgebot/dcfdacc0/file_68083.jpg [04:30:04] Just finished the abstract Wikipedia team translation queue! :) [04:31:36] The information from that page is quite a long ways out of date (re @MolecularPilot: Was doing some translations on MediaWiki.org and this came up hahaha lol) [04:33:00] IIRC [04:37:25] Not really?? I mean all the info about Wikifunctions should probably be present, not future tense i.e. "is" not "will be" but it's all true and still valid [04:37:51] Ahh the one on Meta is the out of date one (re @MolecularPilot: Was doing some translations on MediaWiki.org and this came up hahaha lol) [04:38:03] The one on MediaWiki looks fine [04:39:05] For some reason whoever did the translation markup did it wrong cause there's just random sentences in lede it won't let me translate as well as entire sections that aren't marked up [06:00:29] Wow there's over 200 translations that need doing on meta! Starting now... [06:05:21] Actually i misread, it's 7000 😬 [06:16:55] I think I just got Z21796 working. At least it now passes all the Tests I put in. If you can find a way to break it, I'd appreciate that. [06:22:26] If you have a non-0 numerator, but the denomintor is 0, it throws ZeroDivisionError and returns void (re @Toby: I think I just got Z21796 working. At least it now passes all the Tests I put in. If you can find a way to break it, I'd apprec...) [06:23:08] Ideal behavior (re @MolecularPilot: If you have a non-0 numerator, but the denomintor is 0, it throws ZeroDivisionError and returns void) [06:24:10] True! [06:28:41] okay so this is weird?? i did netural sign 1/10 with 200 sigfigs and it returned "0.000..." instead of "0.100..." [06:29:19] Also ideal behavior (re @MolecularPilot: okay so this is weird?? i did netural sign 1/10 with 200 sigfigs and it returned "0.000..." instead of "0.100...") [06:30:37] but isn't 1 divided by 10 0.1? [06:31:05] 1 times 0 divided by 10 is 0 (re @MolecularPilot: but isn't 1 divided by 10 0.1?) [06:31:38] i meant like this : https://tools-static.wmflabs.org/bridgebot/6bf2bceb/file_68085.jpg [06:32:43] If you changed the sign the negative one would not say "1 divided by 10 is 0.1, so -1/10 is also 0.1" (re @MolecularPilot: i meant like this) [06:32:59] oh, thank you for explaining! 🙂 [06:33:08] The sign is a multiple, essentially [06:33:49] 99of9 it's a very good function, couldn't find anything to break it! 🙂 [06:34:49] Perhaps it would have been more intuitive to make it a Boolean? (re @MolecularPilot: i meant like this) [06:35:20] As positive 0/10 and neutral {any}/10 are the same [06:36:50] Actually, maybe the type converter should error? It is debatably an invalid input, as the only number that can have a neutral sign is 0 (re @Feeglgeef: Also ideal behavior) [06:39:52] Yes that is bad. Until we get a validator, we could prevent this in type conversion. (re @Feeglgeef: Actually, maybe the type converter should error? It is debatably an invalid input, as the only number that can have a neutral si...) [06:41:12] This will eventually be prevented by the validator. But it's good to know the error is intelligible until then. (re @MolecularPilot: If you have a non-0 numerator, but the denomintor is 0, it throws ZeroDivisionError and returns void) [06:42:14] It says only "runtime error", but it's pretty easy to deduce why once you look at your input (re @Toby: This will also eventually be prevented by the validator. But it's good to know the error is intelligible until then.) [06:42:46] The error is in the type converter, which is convenient as according to the naming recommendations those errors will be displayed to users (re @Toby: This will also eventually be prevented by the validator. But it's good to know the error is intelligible until then.) [06:43:24] While coding it felt like patching up a leaky ship rather than a smooth build. So I guess it can be improved, even if not broken. (re @MolecularPilot: 99of9 it's a very good function, couldn't find anything to break it! 🙂) [06:43:40] That's how coding always feels hahaha [06:44:25] It never goes perfectly, but honestly as an outside observer on your function you've done it really well [06:46:16] Wow that code is long. Will look at cleaning it up when I get the chance (re @Toby: While coding it felt like patching up a leaky ship rather than a smooth build. So I guess it can be improved, even if not broken...) [06:47:09] Yes that's my feeling too. Some of it is still debugs, but even without them it felt a bit of a mess. (re @Feeglgeef: Wow that code is long. Will look at cleaning it up when I get the chance) [06:47:35] MolecularPilot I wonder from your name, are you a chemist too? [06:50:28] I'm still in high school but I want to be a biochemist, definitely have an interest in chemistry and bioinformatics, that's why I picked the username lol :) (re @Toby: MolecularPilot I wonder from your name, are you a chemist too?) [06:52:46] Awesome. I'm a computational chemistry academic in my day job. Maybe we can build a nice collection of chemistry functions now that we have some numerical types. [06:53:29] Or maybe propose a chemical element Type. [06:54:11] Wow, definitely, that would be so great, I've already worked on a couple smiles and InChI functions but it would be great to have more coverage!! You have such a cool job omg :) [06:55:39] Like with a natural number to represent the number on the periodic table (are those electrons? or protons? I don't remember)? (re @Toby: Or maybe propose a chemical element Type.) [06:55:47] Yes I noticed that you picked those up. Thanks. (re @MolecularPilot: Wow, definitely, that would be so great, I've already worked on a couple smiles and InChI functions but it would be great to hav...) [06:56:10] Ahh it's protons (re @Feeglgeef: Like with a natural number to represent the number on the periodic table (are those electrons? or protons? I don't remember)?) [06:56:54] Yes, the primary key would be the atomic number (protons). But then maybe we link to Wikidata for more??? (re @Feeglgeef: Like with a natural number to represent the number on the periodic table (are those electrons? or protons? I don't remember)?) [06:57:18] Or we could have functions to do that? (re @Toby: Yes, the primary key would be the atomic number (protons). But then maybe we link to Wikidata for more properties???) [06:57:50] Google has a really cool widget which I would want to implement something similar in functionality for on WF [06:58:31] Would be very interesting to have your opinion on [06:58:31] https://www.wikidata.org/w/index.php?title=Wikidata%3AProperty_proposal%2Fchemical_formula#molecular_formula (re @Toby: Awesome. I'm a computational chemistry academic in my day job. Maybe we can build a nice collection of chemistry functions now t...) [06:58:52] Do the getting from WD? Or try to reproduce the properties locally? (re @Feeglgeef: Or we could have functions to do that?) [06:59:06] Up to functions (re @Toby: Do the getting from WD? Or try to reproduce the properties locally?) [06:59:27] I strongly disagree with keys on types that don't disambiguate [07:00:58] As the things we are over disambiguating for can be done by functions and would be really really annoying to work with (re @Feeglgeef: I strongly disagree with keys on types that don't disambiguate) [07:01:04] Oops, I was probably tagged in that. I'll try to find some time to think it through (re @adafede: Would be very interesting to have your opinion on [07:01:04] https://www.wikidata.org/w/index.php?title=Wikidata%3AProperty_proposal%2Fche...) [07:02:02] Imagine "next chemical element" needing to figure out all the properties of the next one. I consider it more productive to just have a number (re @Toby: Yes, the primary key would be the atomic number (protons). But then maybe we link to Wikidata for more properties???) [07:02:03] And I would be absolutely delighted to hear that Wikifunctions has solutions for this kind of cases :) [07:02:47] An Enum type may also work well as we are working with a limited, but IIRC slowly expanding group (re @Feeglgeef: Imagine "next chemical element" needing to figure out all the properties of the next one. I consider it more productive to just ...) [07:05:37] Maybe neutrons (to identify different isotopes) and electrons (to identify ions/different oxidation states as well), but maybe that would be a seperate type to identify an individual atom (re @Toby: Yes, the primary key would be the atomic number (protons). But then maybe we link to Wikidata for more properties???) [07:06:55] Perhaps a type for each level of specificity? Like we have month, day of roman year, Gregorian date, each one containing the last (re @MolecularPilot: Maybe also have neutrons (to identify different isotopes) and electrons (to identify ions/different oxidation states as well), b...) [07:08:33] Also maybe for the atom time we should also store metastable state? Then we could have a function that takes an radioactive isotope and then performs the decay it would undergo on it [07:09:19] Haven't the slightest clue what any of that means but not opposed (re @MolecularPilot: Also maybe for the atom type we should also store metastable state? Then we could have a function that takes an radioactive isot...) [07:14:57] https://w.wiki/6N8q (re @Feeglgeef: Haven't the slightest clue what any of that means but not opposed) [07:15:56] A metastable state basically means a condition where the nucleus is "excited". It's not the most stable, but it's not immediately unstable. When a nucleus emits gamma radiation, a type of radiation, it loses its metastable/excited state and usually goes to the ground state. If you just consider particle numbers, there's no way to know if a nucleus is [07:15:56] excited/metastable or not (ak [07:15:57] a undergon gamma radioactive decay), so it might be helpful to store this info in the type as a bool. :) (re @Feeglgeef: Haven't the slightest clue what any of that means but not opposed) [07:17:23] Would most uses of the type use this? If yes, go ahead with putting it in the primary type. If not, I think we should instead have another type? (re @MolecularPilot: A metastable state basically means a condition where the nucleus is "excited". It's not the most stable, but it's not immediatel...) [07:18:21] Because every key we add is useful but also a liability for functions that don't need it. [07:27:40] Maybe something like this? It's a very valid point (thanks feegs!) that not all functions would need it and it would feel unneeded in a good amount of functions : https://tools-static.wmflabs.org/bridgebot/09899aa9/file_68087.jpg [07:29:23] Or the functions that need metastability could just accept the atom type and have it as a separate parameter? [07:41:30] https://youtube.com/shorts/LNTBc8ryzEQ?si=6NiSqa-NtgGjsi1l [07:41:31] This is me on Wikifunctions [07:43:02] Is this modeling radioactivity? Or something else (re @MolecularPilot: Maybe something like this? It's a very valid point (thanks feegs!) that not all functions would need it and it would feel unneed...) [07:54:39] We were just brainstorming what types would be useful for chemistry-related functions :) [11:47:22] Does anyone like continued fractions? I made one function, but there could be a whole lot. Z21835 [12:00:41] Kinda 😏 Could it be a separate Type, or just the basis for representing Real numbers? 🤔 (re @Toby: Does anyone like continued fractions? I made one function, but there could be a whole lot. And this first one only has a composi...) [12:08:17] Maybe it's too niche for its own type at this point? I'm not sure what it brings to the table that we can't do with rationals. (re @Al: Kinda 😏 Could it be a separate Type, or just the basis for representing Real numbers? 🤔) [12:33:38] Yeah, I was thinking of float64-free trigonometric function implementations, but it’s all beyond my mathematical education 🙄 Even just looking at pi, a function returning a value to a given number of decimal digits might be better using a continued fraction (https://oeis.org/A001203) [12:33:39] rather than recursively computing successive digits with Z14166 🤷‍♂️ (re @Toby: Maybe it's too niche for its own type at this point? I'm not sure what it brings to the table that we can't do with rationals.) [12:40:28] Yes your rational approximations to pi were what prompted me to tinker with it. (re @Al: Yeah, I was thinking of float64-free trigonometric function implementations, but it’s all beyond my mathematical education 🙄 Eve...) [15:06:31] I did the suggested change: https://www.wikifunctions.org/wiki/Z20885?uselang=en&diff=prev&oldid=159275 (re @Toby: I think there may be an error in Z20885, but only staff can change it (ping @vrandecic ). I believe line 76 should be

if...)
[16:11:49] 	  Where do you see it exactly? (re @MolecularPilot: Actually i misread, it's 7000 😬)
[23:06:40] 	  [[meta:Special:Translate]], the "Abstract Wikipedia" translation group for en>zh! 🙂 (re @amire80: Where do you see it exactly?)
[23:32:05] 	  WOAH (re @MolecularPilot: [[meta:Special:Translate]], the "Abstract Wikipedia" translation group for en>zh! 🙂)
[23:35:29] 	  So much to say about this
[23:35:38] 	  It's completely translated into Korean?!
[23:36:29] 	  And why does it exist there at all? Perhaps it should be moved to wikifunctions.org ?