Post 1300640 by eal@social.sakamoto.gq
 (DIR) More posts by eal@social.sakamoto.gq
 (DIR) Post #1300600 by kurisu@iscute.moe
       2018-11-18T16:08:04.840008Z
       
       1 likes, 1 repeats
       
       the problem with cat-v is that suckless and p9 are actually good software and that they're right half the time in their opinions on softwareall the other "philisophical" stuff is pretty much BS though
       
 (DIR) Post #1300640 by eal@social.sakamoto.gq
       2018-11-18T16:09:40.043524Z
       
       1 likes, 1 repeats
       
       @kurisu >they're right half the time in their opinions on softwareIf you randomly say "foo = good, bar = bad ...", you will reach the same accuracy
       
 (DIR) Post #1300677 by kaniini@pleroma.site
       2018-11-18T16:09:39.482240Z
       
       2 likes, 0 repeats
       
       @kurisu honestly suckless software is not very good much of the time, and it's 9p not p9.
       
 (DIR) Post #1300693 by kurisu@iscute.moe
       2018-11-18T16:11:20.568190Z
       
       2 likes, 0 repeats
       
       @eal they at least havea healthy bent towards simplicity which helps a lot
       
 (DIR) Post #1300698 by kurisu@iscute.moe
       2018-11-18T16:11:30.663793Z
       
       1 likes, 0 repeats
       
       @kaniini 9p is the protocol, p9 is the OS
       
 (DIR) Post #1300700 by lanodan@queer.hacktivis.me
       2018-11-18T16:11:07.374052Z
       
       1 likes, 0 repeats
       
       @kurisu suckless software is better but I wouldn’t really qualifies it as “good”.And their opinion tend to be quite trash, XML and JSON can be good formats, it just needs to be used right (ActivityPub should have been in XML instead of JSON-LD btw).
       
 (DIR) Post #1300759 by kurisu@iscute.moe
       2018-11-18T16:13:44.345693Z
       
       0 likes, 0 repeats
       
       @lanodan you can reasonably write a JSON parser from scratch in a day, you can't write a fully-featured standards-compliant XML parser in a month. I'd much prefer AP on JSON instead of XML, it's the JSON-LD bit that sucks.
       
 (DIR) Post #1300795 by a_breakin_glass@cybre.space
       2018-11-18T16:14:51Z
       
       0 likes, 0 repeats
       
       @lanodan @kurisu YAML pls
       
 (DIR) Post #1300796 by kurisu@iscute.moe
       2018-11-18T16:15:13.748599Z
       
       0 likes, 0 repeats
       
       @a_breakin_glass @lanodan almost as bad to parse, honestly
       
 (DIR) Post #1301052 by trufas@niu.moe
       2018-11-18T16:20:31Z
       
       1 likes, 0 repeats
       
       @kurisu broken clock is right twice a day and all that
       
 (DIR) Post #1301054 by lanodan@queer.hacktivis.me
       2018-11-18T16:17:12.071938Z
       
       0 likes, 0 repeats
       
       @a_breakin_glass @kurisu A format is for a purpose, have your ever heard of “design”?
       
 (DIR) Post #1301055 by kurisu@iscute.moe
       2018-11-18T16:23:23.522679Z
       
       0 likes, 0 repeats
       
       @lanodan @a_breakin_glass I agree, but XML always has a better, simpler alternative
       
 (DIR) Post #1301244 by lanodan@queer.hacktivis.me
       2018-11-18T16:24:49.419743Z
       
       0 likes, 0 repeats
       
       @kurisu @a_breakin_glass - XML is good for: Extensibility (it was made for that, dammit), Document Formatting- JSON is good for: Pretty-printing and compactness, multi-level- YAML is good for: INI but with lists- CSV(Character Separated Value) is good for: strongly defined one-level key → value(s) (ie. /etc/passwd), sucks for lists or when the separation character is in the value.- INI is good for: same as CSV but with one value per key
       
 (DIR) Post #1301245 by kurisu@iscute.moe
       2018-11-18T16:32:04.613897Z
       
       0 likes, 0 repeats
       
       @lanodan @a_breakin_glass XML is a markup language, using XML inside AP message bodies makes sense. Using it as a data exchange format is pretty annoying because of the difference in data models between the structs, lists and pointers that all languages provide and XML's model which was designed for added metadata to text.YAML is far far more than INI with lists, and it's very difficult to parse correctly due to being a feature-bag. Probably as hard as XML.CSV and INI are simpler but underspecced. They work well despite that in practice.
       
 (DIR) Post #1301263 by lanodan@queer.hacktivis.me
       2018-11-18T16:26:26.088692Z
       
       0 likes, 0 repeats
       
       @kurisu @a_breakin_glass Also if you’re just gonna continue in the direction of “X is better than Y; Y sucks”, well that’s just a fucking opinion and we would probably just all have one format if that were true.
       
 (DIR) Post #1301264 by pndr@ap.kios.cc
       2018-11-18T16:32:43.555783Z
       
       1 likes, 0 repeats
       
       @lanodan @kurisu @a_breakin_glass the thing is about picking the right tool for the jobBesides writing an XML parser is probably far easier than writing a YAML parser
       
 (DIR) Post #1301286 by roka@pl.smuglo.li
       2018-11-18T16:33:39.202572Z
       
       1 likes, 1 repeats
       
       @kurisu @lanodan i'd rather just use properly formatted CSV tbh2433f0576cb7cb1091938d474a7a2d9…
       
 (DIR) Post #1301326 by kurisu@iscute.moe
       2018-11-18T16:35:45.840346Z
       
       0 likes, 0 repeats
       
       @roka @lanodan CSV only gives you a flat relational table structure where all the columns are strings. It's not particularly powerful but when the data fits the model it is great.
       
 (DIR) Post #1301393 by lanodan@queer.hacktivis.me
       2018-11-18T16:37:01.611558Z
       
       0 likes, 0 repeats
       
       @kurisu @a_breakin_glass CSV has a standard about it I think it’s from ISO or something and I think there is a RFC version too.
       
 (DIR) Post #1301394 by kurisu@iscute.moe
       2018-11-18T16:38:51.132010Z
       
       0 likes, 0 repeats
       
       @lanodan @a_breakin_glass yeah, it does. Unfortunately it's not always adhered to. Better than INI though where there's no real spec.As I said, they both work fine in practice and I have very few complaints about them. I mostly dislike XML and YAML's complexity and CSV and INI are not complex.
       
 (DIR) Post #1301429 by roka@pl.smuglo.li
       2018-11-18T16:40:31.627168Z
       
       1 likes, 0 repeats
       
       @kurisu @lanodan dead simple to parse too :rokalife:
       
 (DIR) Post #1301458 by kurisu@iscute.moe
       2018-11-18T16:41:43.660707Z
       
       0 likes, 0 repeats
       
       @roka @lanodan not quite so simple as to use cut and grep as soon as sone of your values need escaping (yeah, escaping in CSV exists) but the escaping isn't too bad
       
 (DIR) Post #1301619 by lanodan@queer.hacktivis.me
       2018-11-18T16:44:42.429260Z
       
       0 likes, 0 repeats
       
       @kurisu @roka From what I remember the escaping is done with double-quotes, from a C point of view you can strtok blindly, then check for " at the start and if true another one with " at the end and strcat theses two.I could check this against official implementation/documentation but well, done, C parser for you.
       
 (DIR) Post #1301620 by kurisu@iscute.moe
       2018-11-18T16:48:35.027348Z
       
       0 likes, 0 repeats
       
       @lanodan @roka oh yeah, the escaping is really nice and simple I wasn't complaining about the escaping. However, sometimes people assume there is no escaping and create vulns because of that (user putting , in their value)
       
 (DIR) Post #1301715 by lanodan@queer.hacktivis.me
       2018-11-18T16:52:54.709842Z
       
       0 likes, 0 repeats
       
       @kurisu @roka There will always be vulns even with “stupid” formats.You could have one with original DNS just by putting a . at the end (canonical form, . are root NS) or omitting it.Same with email with a @ in the local-part (to which it is legal to put ~anything) or @ in the host part (thanks IDN).
       
 (DIR) Post #1301716 by kurisu@iscute.moe
       2018-11-18T16:53:56.316216Z
       
       0 likes, 0 repeats
       
       @lanodan @roka yeah, it's not really a complaint about CSV. We're pretty much agreed on CSV.
       
 (DIR) Post #1302880 by copyleftie@freeradical.zone
       2018-11-18T17:56:55Z
       
       0 likes, 0 repeats
       
       @kurisu cat-v thinks Milton Friedman is a "genius" and that definitely makes me uncomfortable with agreeing with as many of their opinions on software as I do
       
 (DIR) Post #1302881 by kurisu@iscute.moe
       2018-11-18T17:58:02.167739Z
       
       0 likes, 0 repeats
       
       @copyleftie you don't have to agree with them (i.e. all their opinions) to agree with some of their opinions
       
 (DIR) Post #1302893 by copyleftie@freeradical.zone
       2018-11-18T17:58:25Z
       
       1 likes, 0 repeats
       
       @kurisu yeah yeah I know but still