ByteCode.NewsSubmit
Use quotes for an exact word.

Subscribe

RSSPodcast
Log in

Chat Logs

  1. phalethany idea why these two respond fine https://js.stdlib.run/site.webmanifest https://bytecode.news/site.webmanifest and these two return 404 https://flog.run/site.webmanifest https://primate.run/site.webmanifest ?
  2. liolocmorning 👋
  3. phalethhi lioloc
  4. phalethlooks like you need a better irc client, one that can just auth you using SASL
  5. phalethalso some that gets you cloaked before joining channels
  6. bluehi
  7. phalethhi blue
  8. liolocyeah definitely, i'm using bIRC on macOS
  9. liolocif you have any suggestions I'm all ears
  10. phalethdo you like TUIs?
  11. phalethlooks like weechat can be installed on a mac https://kb.shells.com/tutorials/macOS/Weechat/
  12. bluelioloc flog 0.1.8 is out, and it supports `flog init`
  13. bluealso, repopack gains rolling release channels, I'm gonna add support for that later into flog and rp
  14. jreicherThe IRC clients in Emacs are excellent TUIs.
  15. liolocthanks will checkout the weechat
  16. liolocalso i saw that blue, actually used it as well, very neat
  17. liolocahhh i see, so some people use something like weechat and then they use tmux to persist it?
  18. liolocso that they can have it running on a vps or something and then use it on their phone?
  19. liolocor desktop
  20. blueyes lioloc, that's what I do
  21. blueI have irssi inside tmux running on a vps
  22. liolocgotcha
  23. liolocway better than my current setup, i'll do the same
  24. lioloctired of getting disconnected
  25. blueweechat is similar to irssi
  26. liolockk
  27. phalethso the file is there https://primate-master.repopack.app/site.webmanifest of course not on primate.run yet
  28. blueya soon, when 0.43 releases
  29. blueif you need to backport things in the future, just create a task with another --base, phaleth
  30. blueso if you want to change smth in the live website, you'd create a task against --base 0.42
  31. bluesince the live website deploys from that branch
  32. phalethno worries, it's gonna be out when it's gonna be out
  33. blue"Key finding: #196 must release @primate/angular as at least 0.14.0; Flog rejects 0.13.0 for the new Primate 0.43 peer range."
  34. blue<3 flog
  35. bluerelease sets are amazing, I don't get why pnpm and other package managers never built them in
  36. blueit's almost like nobody ever dedicates thought in designing actually useful software, especially for monorepos
  37. bluehttps://repopack.com/project/primate/tasks/254
  38. blue<3 flog
  39. bluehttps://repopack.com/project/flog/tasks/106
  40. blueunstable flog is coming, now with repopack's rolling channels
  41. phalethheh
  42. phalethit's just that dev tools are the worst tools of them all
  43. blueya, that's why we have flog
  44. blueto put a stop to that
  45. lioloc_hello from my vps
  46. phalethhello root
  47. phalethtype `/nick lioloc`
  48. liolocah better now
  49. lioloci created a service to have it autostart on load as well
  50. liolocthanks guys
  51. phalethnice
  52. phalethon VPS boot you mean?
  53. phaletha systemd service I guess
  54. liolocyes
  55. phalethsounds convenient
  56. phalethlioloc: what host OS is running on the VPS?
  57. phalethblue: is the post ready for a read since there are no XXXs anymore? https://primate-master.repopack.app/blog/primate-043
  58. nevetPrimate — Primate 0.43: diamond migrations, deferred navigation and direct SQL
  59. liolocphaleth: ubuntu
  60. phalethtry archlinux, it's way better
  61. liolocreally? how come?
  62. phalethubuntu package maintainers are way too busy messing around making software packages ubuntu specific and also pushing rust into every hinge they can
  63. phaletharchlinux is a rolling distro
  64. lioloci'm intrigued
  65. lioloci'm using hetzner, idk if i can change the os though
  66. liolocoh nvm i can
  67. phalethI used hetzner previous year and remember I could choose archlinux no problem
  68. liolocwhat do you use now?
  69. phalethI was hosting a couple of pointless wp sites and told the owner of those sites to take care of that so I don't need that VPS anymore
  70. liolocah gotcha
  71. phaleththese days I'm messing with two blue's VPSes
  72. phalethboth run archlinux without a problem
  73. phalethanyway ubuntu just needs to stop being used, their maintainers and devs are just wasting time
  74. liolocdidnt know that
  75. phalethrewriting core utils that never had a single CVE and are essentially feture complete for decades only to make them worse is a huge waste of effort
  76. liolocoof
  77. bluephaleth: yes, https://primate-master.repopack.app/blog/primate-043 is essentially ready for a read
  78. nevetPrimate — Primate 0.43: diamond migrations, deferred navigation and direct SQL
  79. phalethblue: alright
  80. blueit won't change substantially. flog found operative situations, and they were fixed
  81. bluelioloc: once superarch is done, we will migrate from archlinux to superarch. but superarch is some ways off right now
  82. bluearchlinux is currently the best we have until superarch arrives. we will then likely switch to bsd-based superarch installs, if podman is properly supported on that
  83. liolocvery cool
  84. blueonce repopack matures a bit more, I could host orz (the superarch package manager) and its packages on repopack. the first iteration of orz will likely just download packages from upstream (so node stuff from nodejs.org), but later iterations will build locally
  85. bluelinting with eslint takes so long... `flog lint` has to come sooner than later
  86. phaletheslint is there to stay
  87. blueit's very slow, we don't need it. we need fast, but not biome, biome is terrible and doesn't cover most of the useful eslint rules
  88. blueget this, phaleth
  89. bluehttps://repopack.com/project/flog/tasks/73
  90. bluethis is the future, but without Dog instrumentation. linting/formatting/test coverage will be handled by a Flog library called Stick
  91. blueDog will benefit from Stick as well later when it needs to detype .ts files
  92. bluethe reasons for Stick is that I don't wanna wait for Dog so long, Dog will still take some time
  93. blueand Flog needs to get some real useful capabilities
  94. blueI have two Dog experiments that failed, it's currently stuck on 9x V8 so it still needs a lot of work
  95. phalethnot sure what is the meaning of stick here
  96. blueDog has now moved to a pure C++ pilot, we'll see where that gets us
  97. phalethalso clanker simply has no chance as usual
  98. bluestick is essentially an ast tree interpreter, not a full engine
  99. blueyou need that for formatting/linting/etc. mostly static analysis, not a full-blown engine
  100. phalethso it's supposed to be used by dog then?
  101. bluemostly by flog for `flog format`, `flog lint` and `flog test --coverage`, but also by dog for some things. like running .ts files
  102. blueI'm gonna turn flog into a monorepo with flog and stick in it. dog will join them later, I will delete the dog project, since they'd be sharing libs
  103. bluedon't worry about it, it's all lessons learned
  104. phalethwell, flog should be able to initiate, but execution will still happen running on the engine
  105. bluedog will still be running JS
  106. phalethso I just don't think stick is an appropriate name
  107. bluenone of these are appropriate names, it's just fanciful names, they don't mean much
  108. bluebut don't worry, stick won't have a webpage and you won't need to design a logo for it :P
  109. blueit's just going to be a library used by flog and dog
  110. phalethmaking logo is not difficult, it's just the repopack project is a mess having ton of duplicates here and there
  111. phalethpretty sure sh script lines of code already exceeded 20k in that project
  112. phalethprolly some bash in there too
  113. blueit's mostly the provision part, no?
  114. blueit's all gonna disappear once we move to Go tooling
  115. phalethso as long as the website is clean it should be a no brainer to get a new logo up there
  116. phalethno clue, most likely
  117. bluek
  118. phalethit's like clanker self inflicting problems all the time
  119. bluewell, we need flog, dog and stick. they all solve real problems
  120. blueflog is already improving the way to publish packages in rcompat/primate and prevents me publishing nonsense
  121. phalethno idea what is it on about here https://repopack.com/project/flog/tasks/107
  122. phalethsounds like it invented some bash script with pnpm deeprooted and now it insists on me running that generating some more scripts to make the thing get bloated for enternity
  123. phalethalso it mentiones warnings from new apt, totally irrelevant as long as it can install whatever is needed
  124. bluesubmit it again
  125. phalethok
  126. bluethis is clanker nonsense. AGENTS.md says every change needs benchmark measurements
  127. blueI will change that guidance to "every change to .go" files
  128. phalethalright
  129. bluehttps://repopack.com/project/primate/tasks/259
  130. bluethis is likely gonna make it to 0.43 as well, because it affects repopack
  131. bluehttps://repopack.com/project/flog/tasks/108
  132. blueI'm thinking of adding shortlinks to repopack
  133. bluemaybe repopack.com/t/flog:108
  134. bluephaleth: thoughts?
  135. blueit'd just be a redirect, but easier to share
  136. blueactually the : does nothing there, could be repopack.com/t/flog/108 as well
  137. phalethcould primate@259 be backported if 0.43 was out?
  138. blueit could be, but I'm doing it now so it'll soon be done
  139. blue0.43 has passed two audits, one by sol 6.0 and one by sol 6.1. it did find a few things, those were mostly pre-0.43 stuff
  140. phalethno reason to bother with short urls
  141. blueI do wish clanker would understand that when I ask it to review a release, I mean to review it against changes to the last release, not in general
  142. phalethI've said every byte counts but only when it comes to browser DOM abuse
  143. bluek
  144. phalethheh
  145. blueI mean, it's ok finding bugs that are unrelated, but they've been there for a while and no one complained
  146. blueI don't even know if anyone uses @primate/angular
  147. blueit's ok to fix a bug in it, but I wouldn't consider it a blocker for 0.43
  148. bluepity clanker doesn't understand that nuance
  149. phalethheh
  150. bluewhich is ok.. clanker doesn't know primate's userbase
  151. phalethclanker wants to clanker
  152. bluebut still, I asked it specifically to review the release, not to audit the project
  153. blueit has a tendency to overengineer/overthink stuff which is really annoying
  154. blueI feel like repopack's chat will be able to curb some of these tendencies by building-in good guidance
  155. bluebut clanker will clanker, as you say
  156. phalethat least it didn't start pushing rust into every hinge yet
  157. bluewell, rust is a no-go for repopack or any project I have a say in
  158. blueI will never accept code in a language you need a phd to understand
  159. bluethat's currently rusk and haskell for me
  160. bluerust*
  161. bluemy rule of thumb is that if I can't understand a piece of code, then I won't merge it
  162. blueI don't need to understand every detail, but I need to be able to map the code and understand what it does, and I can't do that with rust
  163. phalethyeah
  164. phalethsome languages are made for some people's amusements
  165. blueyup. rust is a purely ego-driven language
  166. blueit might satisfy some people's needs to show off.. but I can't accept it
  167. bluetoo pity zig turned into such a worthless dumpster fire
  168. blueI really liked it :(
  169. bluejust shows you sometimes good people with bad intentions can ruin good projects
  170. phalethzig as a lang is still fine, the zig standard lib got a bit weird
  171. phalethanyways, no need for a system lang for now, also there are alternatives anyway
  172. phalethI like the database optimizations in primate 0.43, that's actually the way to go
  173. bluediamond migrations?
  174. phalethwell that's nice too
  175. blueya
  176. bluethe repopack looks like a friggin' alien in the favicon, and I love it
  177. bluerepopack logo*
  178. phalethheh, after I scaled it horizontaly it's got a bit weird I think
  179. bluedo you want to rework it?
  180. blueit's got alien vibes right now
  181. blueif htat's intended, then it's ok
  182. phalethalien vibes are ok
  183. phalethit had them before anyway
  184. phalethit looked like a reptile in a mask in the favicon
  185. bluelol
  186. bluehttps://repopack.com/project/flog/tasks/109
  187. bluestick work has started
  188. bluephaleth: clanker review 749
  189. bluereviewed*
  190. blueit found some missing old logos
  191. bluewe should really centralise those assets
  192. phalethyeah, good review
  193. dreamrealnext BCN article on its way...
  194. bluedreamreal: `flog format` is coming
  195. bluelikely with a flog 0.2.z line
  196. dreamrealnoice!
  197. bot[ByteCode.News] That free beer has a brewer - https://bytecode.news/posts/2026/10/that-free-beer-has-a-brewer
  198. bluesweet, dreamreal
  199. dreamrealSomeone post that BCN post to HN please
  200. dreamrealbest if it's not blue this time because their filters watch for associations
  201. phalethdreamreal: https://news.ycombinator.com/item?id=50245121
  202. nevetHacker News
  203. dreamrealtodah!
  204. dreamrealsays "dead"
  205. dreamrealhrrrm
  206. phalethcannot even upvote?
  207. dreamrealit literally shows [dead] for me
  208. phalethoh, yeah, for me too
  209. phalethafter I logged out
  210. dreamrealNo idea why
  211. phalethyeah, well, hn does not like my account
  212. dreamrealhttps://news.ycombinator.com/item?id=50245457
  213. nevetThat free beer has a brewer | Hacker News
  214. bluedreamreal: I didn't peruse yet; node2shell might also merit mention
  215. bluethis, or rathe last week, has been a catastrophe for programming
  216. dreamrealOr something that's been LONG overdue, a butcher's bill that's been unpaid for far too long
  217. bluewell, php has always been a dumpster fire
  218. bluethis is nothing new
  219. dreamrealPHP isn't the point, though, and never has been
  220. dreamrealthis is a problem that PHP just ran into
  221. blueI beg to differ, mr dreamreal
  222. bluethis is very much a PHP point. good software, like openbsd, does not suffer from this
  223. * dreamreal thinks that's less true than blue might hope.. and the exploits are only EVIDENCE, not cause
  224. bluethere are superior and less superior models of programming. there *are* better programming languages, and more conscientious developers. all of this bears upon the product. the problem is NOT necessarily that PHP has one fulltime member. the problem is PHP is dumpster fire level garbage that should be retired ASAP
  225. dreamrealSure, but it's not ABOUT PHP
  226. dreamrealJava had the same kind of vulnerability
  227. bluethe place where your article and I differ, is that I generally do *not* believe that throwing money at problems fixes them. throwing money at stuff may, and often very poorly, alleviate symptoms. fixing problems is inextricably intertwined with uphill fighting many faults of the human nature
  228. blueand that is much, much harder than the article's promise
  229. blueif you give me a million dollars, that will *not* make primate better
  230. bluewhatever contributes to primate being good or bad is wholly connected with the culture around it
  231. bluethe people working on it, and with it
  232. dreamrealI don't think THROWING MONEY fixes problems
  233. dreamrealbut I think participating does
  234. bluethe reason I think Go is a great language, and I've only come to realise it recently, is because the c people wrote it, but also because it chose simplicity where others preferred complexity
  235. bluethose two things are completely unrelated to money
  236. bluethat is also my problem with rust: the "you-need-a-phd" complexity means people will produce conceptually bad software with it, and see those worthless rewrites of everything that crawls under the heavens, everywhere
  237. bluethat is a *bad* culture, no matter how much money the rust foundation has
  238. bluethe same criticism now goes to zig
  239. bluealthough it plays out differently there; the language is conceptually ok, the people are totally crazy
  240. bluedreamreal: phaleth: lioloc: https://primate.run/blog/primate-043
  241. nevetPrimate — Primate 0.43: diamond migrations, deferred navigation and direct SQL
  242. blueprimate 0.43 is released
  243. bluethe 0.44 cycle has now opened
  244. phaleththe website works fine, well done
  245. phalethmobile score is better https://pagespeed.web.dev/analysis/https-primate-run/zmilz1ezks?form_factor=mobile
  246. liolocNICE
  247. liolocwill check it out
  248. bluenice phaleth
  249. blueya lioloc
  250. lioloclots of changes, nice to see. Very nice. I'm off to bed, see you tomorrow!
  251. liolocCongrats on the release
  252. bluethx