<<Prev Next>> Scroll to Bottom
Stuff goes here
L1[00:00:03] <Izzy> https://cdn.shadowkat.net/media/43/b6/8e/43b68e8010543fa37152a210d0afe4ae4b3c5dbc1d8e2f261e09a844b19ea1c3.jpg
L2[00:00:20] <Mic​hiyo> https://discordembeds.pc-logix.com/live/1534350304659898519/1785888019768-ce0f8ca8c7b3.mp4
L3[00:01:29] <SoraFi​restorm> speaking of screens, I think the holoscreens are a little derped
L4[00:01:34] <Izzy> oh hell yeah
L5[00:01:46] <Corded> > <SoraFi​restorm> speaking of screens, I think the holoscreens are a little d…
L6[00:01:47] <Mic​hiyo> Not surprised. lol. What's up?
L7[00:02:01] <SoraFi​restorm> well a couple things
L8[00:02:16] <SoraFi​restorm> the manual says that putting a keyboard in the inventory slot should make that go
L9[00:02:19] <SoraFi​restorm> it does not appear to be the case
L10[00:02:31] <Izzy> did I post my KVM switch? :3 https://uranus.shadowkat.net/tmp/060c.jpg https://uranus.shadowkat.net/tmp/8dcf.jpg
L11[00:02:34] <SoraFi​restorm> secondarily, the holoscreen disappears when the projector block itself goes offscreen
L12[00:03:01] <Mic​hiyo> sigh yeah that's a stupid MC thing that I had to fix in LC too lol
L13[00:03:12] <Mic​hiyo> ooh DVI KVM, nice
L14[00:03:21] <Izzy> DVI-I KVM, no less
L15[00:03:30] <Izzy> in theory if I had appropriate displays it can take digital and analog inputs
L16[00:03:33] <Mic​hiyo> very nice
L17[00:03:35] <Izzy> work in progress
L18[00:03:43] <Mic​hiyo> my shitty KVM is still VGA lol
L19[00:03:54] <Izzy> you can get HDMI ones off aliexpress for cheap
L20[00:04:01] <Izzy> but uhhhhh this was $100 and works a lot better
L21[00:04:29] <Mic​hiyo> yeah, then I get to pay another 90,000% in tariffs, or not? who knows it's $DAYOFTHEWEEK
L22[00:05:15] <Izzy> I had to get this one from the US actually x3
L23[00:05:27] <Izzy> breaking my rule of not buying anything from the US
L24[00:05:51] <Izzy> F1DN104E is the model
L25[00:17:03] <SoraFi​restorm> oh, and using a dye to change the color of the screen also appears to not be supported
L26[00:17:51] <Mic​hiyo> Ok, I pushed the KVM it... works so who knows how badly it'll break. Wife wants to play some games, so I'm AFD
L27[00:22:34] <Corded> > <Iz​zy> did I post my KVM switch? :3 https://uranus.shadowkat.net/t���
L28[00:22:34] <Va​ur> oh dear lord, those are the old mouse and keyboard connectors, that thing is ancient 😄
L29[00:23:29] <Mic​hiyo> Good Ole ps2
L30[00:32:46] <SoraFi​restorm> I'm not likely to use the holoscreens tbh so like... probably fix them but not because I'm just jonesing for them
L31[00:32:48] <Izzy> Vaur: 2012 or so I believe
L32[00:33:24] <Izzy> KVM switches kept PS/2 forever, especially if they also supported VGA somehow
L33[00:42:36] <Izzy> welp https://uranus.shadowkat.net/tmp/4ab7.png
L34[03:06:19] <Spider​EveryOS> %tonk
L35[03:06:19] <MichiBot> Geez! Spider​EveryOS! You beat Forec​aster's previous record of <0 (By 5 hours, 42 minutes and 19 seconds)! I hope you're happy!
L36[03:06:20] <MichiBot> SpiderEveryOS's new record is 5 hours, 42 minutes and 19 seconds! SpiderEveryOS also gained 0.00571 tonk points for stealing the tonk. Position #1.
L37[04:49:16] ⇦ Quits: vidak (~vidak@2407:e400:7800:2c01:c4db:d9c4:aac4:94f6) (Ping timeout: 195 seconds)
L38[06:34:00] *** Joins: Hawk777 (~Hawk777@2605:1700:1080:de00:5c4a:f766:d705:c4e8) Hawk777
L39[06:38:54] ⇦ Quits: Hawk777 (~Hawk777@2605:1700:1080:de00:5c4a:f766:d705:c4e8) (Quit: Leaving.)
L40[08:51:26] ⇦ Quits: Amanda (~m-yt727s@cloud.camnet.site) (Read error: Connection reset by peer)
L41[08:52:51] *** Joins: Amanda (~m-yt727s@cloud.camnet.site) AmandaC
L42[09:00:40] <Izzy> I have compiled OC for 1.12.2
L43[09:00:44] <Izzy> let's see if I can backport this shit
L44[09:06:21] ⇦ Quits: Amanda (~m-yt727s@cloud.camnet.site) (Read error: Connection reset by peer)
L45[09:07:17] *** Joins: Amanda (~m-yt727s@cloud.camnet.site) AmandaC
L46[09:19:06] <Izzy> oh boy we've hit the last file and now it looks like I might need something resembling an understanding of what I'm doing
L47[09:22:19] <SquidDev> Screaming. Everything about how the 1.21.1 port handles block/item/BE registration is insane.
L48[09:22:55] <Izzy> I'm currently trying to figure out how the new loadData/saveData maps to old load/save
L49[09:23:19] <Izzy> I think the underlying metadata in the world changed between versions :D
L50[09:24:06] <Izzy> me rn https://uranus.shadowkat.net/tmp/7021.jpg
L51[09:24:24] <Redston​eParkour> i believe its related to minecraft moving to 'components' instead of NBT in modern versions
L52[09:24:35] <Redston​eParkour> (components basically being a limited set of NBT you can set)
L53[09:26:53] <Izzy> I seeeeeeeeee
L54[09:27:04] <Izzy> hence ExtendedDataComponentHolder
L55[09:46:15] <Izzy> scala. how does it work.
L56[10:09:58] <Spider​EveryOS> %tonkout
L57[10:09:58] <MichiBot> Fiddlesticks! Spider​EveryOS! You beat your own previous record of 5 hours, 42 minutes and 19 seconds (By 1 hour, 21 minutes and 19 seconds)! I hope you're happy!
L58[10:09:59] <MichiBot> Spider​EveryOS has tonked out! Tonk has been reset! They gained 0.007 tonk points! plus 0.012 bonus points for consecutive hours! Current score: 2.6125506, Position #1
L59[10:13:24] <Izzy> okay I've got the errors down to just my save/load logic
L60[10:13:25] <Izzy> very good
L61[10:15:31] <luna​r_sam> hm
L62[10:15:40] <luna​r_sam> do debug hooks run on `__close`?
L63[10:46:46] <Izzy> okay I think I have save logic working
L64[10:47:00] <Izzy> now just to figure out how to like. turn this list into something usable
L65[10:52:37] * Amanda meows and looks around
L66[10:53:03] * Amanda orders MichiBot some %soup % and @Elfi a thimble full of her preferred meowing beverage
L67[10:53:03] <MichiBot> Waiter!! Waiter!! Theres an RTG in my soup!
L68[10:53:23] <Amanda> That's to keep it warm!
L69[10:56:28] <luna​r_sam> yippie pure lua implementation of frexp (i should probably test it first)
L70[11:06:33] <Amanda> Fr? Experimental? Frexp!
L71[11:06:41] <Izzy> me omw to write scala imperatively
L72[11:07:13] * Amanda curls up atop Izzy lap, meows about timetables and trains and the slow march to fascism
L73[11:07:25] <Izzy> just remember Amanda
L74[11:07:34] <Izzy> the one thing no authoritarian regime can do is make the trains run on time
L75[11:10:50] <Izzy> if you want the trains on time, you gotta have the unions running them
L76[11:15:43] <Amanda> trains on time? Sounds like some socialist dream!
L77[11:17:52] <Corded> > <luna​r_sam> yippie pure lua implementation of frexp (i should probably …
L78[11:17:52] <luna​r_sam> i had to handle a few edge cases (infinity and nan)
L79[11:17:58] <luna​r_sam> rather, specifically those two
L80[11:18:14] <luna​r_sam> because they set the exponent bits to all 1s
L81[11:18:16] <Amanda> ... great, I paused a video to use the litter tray, and now nebula won't resume it!
L82[11:19:53] <Amanda> There we go, had to seek back a couple seconds
L83[11:30:13] <Izzy> fuck me it compiled??
L84[11:31:35] <Amanda> Well, if you insist.
L85[11:34:55] <Izzy> Okay. Game doesn't crash. Item doesn't exist.
L86[11:42:11] <Izzy> YOOOOO IT WORKS
L87[11:42:44] <Izzy> forgot to set the item title, and it doesn't have a recipe, but,
L88[11:43:10] <Izzy> https://uranus.shadowkat.net/tmp/1d0b.html
L89[11:44:30] <luna​r_sam> yeah i'm just going ahead and implementing `math.random` and `math.randomseed` in lua for the hell of it
L90[11:45:46] <Izzy> lunar_sam: I backported Michiyo's 4-head T2 graphics card
L91[11:49:02] <Izzy> https://uranus.shadowkat.net/tmp/1da2.jpg https://uranus.shadowkat.net/tmp/f16e.png
L92[11:49:07] <Izzy> this is going in SKS v13
L93[11:50:04] <luna​r_sam> sweet
L94[11:50:21] <luna​r_sam> also turns out i'm an idiot and lua 5.4 stores the random state in an upvalue
L95[11:50:28] <Izzy> oh, T3 server only takes two T3 cards
L96[11:50:29] <Izzy> oop
L97[11:50:41] <luna​r_sam> so it's not a single random state shared between all lua instances anymore
L98[11:52:55] <luna​r_sam> anyways, also threw in implementations of all the hyperbolic trig functions
L99[11:58:04] <Izzy> https://uranus.shadowkat.net/tmp/ad88.webm
L100[11:59:20] <luna​r_sam> neat
L101[12:01:07] <Izzy> two graphics cards in there
L102[12:29:59] <Izzy> hm. okay. so ~ at the start of a destination address in minitel means it's a broadcast / multicast
L103[12:30:12] <Izzy> should I allow subscribing to multicast groups, and possibly repeating specific groups?
L104[12:31:30] <Izzy> though. eh. they all turn up as net_broadcasts
L105[12:31:42] <Izzy> so maybe selective repeating
L106[12:35:37] <Forec​aster> %tonk
L107[12:35:37] <MichiBot> Kapow! Forec​aster! You beat Spider​EveryOS's previous record of <0 (By 2 hours, 25 minutes and 39 seconds)! I hope you're happy!
L108[12:35:38] <MichiBot> Forecaster's new record is 2 hours, 25 minutes and 39 seconds! Forecaster also gained 0.00243 tonk points for stealing the tonk. Position #2. Need 0.2589312 more points to pass Spider​EveryOS!
L109[12:49:30] <as​ie> would be interesting if OC:Rebooted got a 1.12.2 branch
L110[12:50:08] <Izzy> probably want to pull from MP/OC first
L111[13:02:45] <Corded> > <as​ie> would be interesting if OC:Rebooted got a 1.12.2 branch
L112[13:02:45] <spo​nge> so a backport of a port of a port
L113[13:04:38] <spo​nge> you could also fork the 1.12.2 version and implement the oc:r stuff there
L114[13:05:20] <spo​nge> * one
L115[13:06:22] * Izzy waves
L116[13:16:52] <spo​nge> yes hello irc people
L117[13:20:28] <SquidDev> Uhhh, the 1.21.1 port just entirely deleted the access point and switch.
L118[13:20:51] <SquidDev> "// TODO Remove in 1.7" Yes, yes it is.
L119[13:23:45] <spo​nge> i don't know why we need so many item classes when some of them don't even do anything
L120[13:24:39] <SquidDev> There's some really nasty places where the item name is derived from the class name. I want to clean that up at some point.
L121[13:25:51] <spo​nge> where exactly
L122[13:27:46] <SquidDev> https://github.com/CaitlynMainer/OpenComputers/blob/0857b399d2b17a900b34ab5d9b967dc463af5216/src/main/scala/li/cil/oc/common/item/traits/SimpleItem.scala#L45
L123[13:28:02] <SquidDev> All the places overriding getDescriptionId are broken — that should be driven by components now.
L124[13:28:39] <spo​nge> oh
L125[13:28:57] <spo​nge> that really is quite nasty
L126[13:38:58] <Izzy> Shame there's nothing for bitblt between GPUs
L127[13:39:15] <Izzy> Could do some Sun Microsystems flavoured trickery easily then
L128[13:40:22] *** Joins: Vexatos (~Vexatos@2001:4647:af75:0:a19f:4608:f47:914) Vexatos
L129[13:40:22] zsh sets mode: +v on Vexatos
L130[13:51:55] <Redston​eParkour> now why isn't the shell appearing... https://discordembeds.pc-logix.com/live/1534559578765922355/1785937914635-381faa34f630.png
L131[14:05:44] <Vex​atos> @asie uh am I supposed to put the OC update on curseforge too, manually?
L132[14:05:54] <as​ie> yes, because the scripts blew up
L133[14:05:58] <Vexatos> aight
L134[14:06:02] <as​ie> this is the final release so after that you can relax forever
L135[14:06:02] <Vexatos> do I just take them from modrinth?
L136[14:06:04] <as​ie> yes
L137[14:06:06] <as​ie> or github
L138[14:06:08] <as​ie> either works
L139[14:06:10] <Vexatos> aight
L140[14:06:16] <Forec​aster> "AI stories rated better than human ones study shows" --The guardian
L141[14:06:16] <Forec​aster> Doubt that's true...
L142[14:06:34] <as​ie> that's not what the article says
L143[14:07:00] <as​ie> there was a study. they gave 1682 people six stories to read - three AI generated, three human written
L144[14:07:18] <as​ie> they told them some were AI generated and some were human, but not always correctly
L145[14:07:22] <as​ie> and asked them to rate those stories
L146[14:07:49] <Forec​aster> Which article, the guardian one?
L147[14:07:49] <as​ie> https://www.cambridge.org/core/journals/judgment-and-decision-making/article/bot-or-not-can-people-tell-the-difference-between-stories-written-by-a-human-or-by-an-ai-system/45E6DC0BB90AA648654D5AE243F6C667
L148[14:07:55] <as​ie> yes. here's the study itself
L149[14:07:58] <Redston​eParkour> hm but they weren't not told if it was human or ai
L150[14:08:13] <Forec​aster> That's what the headline I just saw said
L151[14:08:14] <as​ie> no, they had a second study where they didn't tell them anything
L152[14:08:23] <Corded> > <Forec​aster> That's what the headline I just saw said
L153[14:08:23] <as​ie> yes, and the article is just guardian's editorializing of this study
L154[14:08:32] <as​ie> and the headline is guardian's editorializing of this article
L155[14:08:39] <as​ie> > In Study 1, participants (1,682 adults recruited from Prolific) read either a human-written or AI-generated story. Half were told that their story was written by a human, and half were told that their story was written by ChatGPT. Participants rated the AI-generated story as higher quality and more absorbing, but they rated the story
L156[14:08:40] <Corded> better on both measures when they were told it was written by a human. In Studies 2 and 3, participants (905 adults recruited from Prolific) read both types of story and guessed each story’s origins. Participants were no better than chance at differentiating between the human-written and AI-generated stories. Self-reported expertise with AI, but
L157[14:08:40] <Corded> not with fictional literature, was positively correlated with correct story identification. AI programs thus can generate creative work that is viewed as being at least as good as human-generated work but are not perceived as being able to do so.
L158[14:10:23] <as​ie> https://osf.io/d8f6m/ and here is the data. the study also lists the exact human written stories used, and prompts used to generate their AI counterparts
L159[14:10:49] <Redston​eParkour> closer https://discordembeds.pc-logix.com/live/1534564337774100510/1785939049129-de0826930591.png
L160[14:11:24] <Redston​eParkour> and it almost crashed my browser
L161[14:11:25] <Redston​eParkour> whoops
L162[14:18:06] <luna​r_sam> in agony
L163[14:18:35] <luna​r_sam> running the pure lua pattern matching against the lua 5.4 tests
L164[14:18:45] <luna​r_sam> gotta decipher the ancient runes now to fix it
L165[14:18:53] <luna​r_sam> (it's failing against a 5.3 test)
L166[14:19:02] <Redston​eParkour> uh oh
L167[14:19:18] <luna​r_sam> and the pattern matching functions specifically said they were tested against 5.2's tests
L168[14:19:26] <Redston​eParkour> and do they pass those?
L169[14:19:29] <Corded> > <Redston​eParkour> and it almost crashed my browser
L170[14:19:29] <luna​r_sam> and yeah that'll happen
L171[14:19:34] <Corded> > <Redston​eParkour> and do they pass those?
L172[14:19:34] <luna​r_sam> i assume so idk
L173[14:19:47] <Corded> > <luna​r_sam> and yeah that'll happen
L174[14:19:47] <Redston​eParkour> yeah its spamming the tty for some reason
L175[14:20:14] <Redston​eParkour> its either my cursed unicode library or my cursed gpu-on-tty library
L176[14:20:14] <luna​r_sam> i also needed to disable using the native implementation since we can't yield in gsub callbacks
L177[14:20:26] <luna​r_sam> though i could also just change that for only gsub
L178[14:20:27] <luna​r_sam> lol
L179[14:20:39] <luna​r_sam> in fact, i think i will
L180[14:21:20] <Redston​eParkour> also fun fact: the vt100 escape code to move the cursor is `\27[<y>;<x>H` (iirc), but openos' (and psychos') vt100 use `\27[<x>;<y>H`
L181[14:21:30] <luna​r_sam> also the negative memory bug is back
L182[14:21:35] <Corded> > <Redston​eParkour> also fun fact: the vt100 escape code to move the cursor is …
L183[14:21:35] <luna​r_sam> YEAH
L184[14:21:37] <Redston​eParkour> yay i love free ram
L185[14:21:39] <luna​r_sam> I NOTICED THAT
L186[14:21:49] <luna​r_sam> 😭
L187[14:22:17] <Redston​eParkour> anyways i'm betting its the unicode library
L188[14:22:35] <luna​r_sam> i'm shocked you got it that close to booting openos as is
L189[14:22:48] <Redston​eParkour> cursedness upon cursedness upon cursedness
L190[14:22:57] <luna​r_sam> i had zero plans for that
L191[14:23:02] <Redston​eParkour> i've effectively made a hypervisor
L192[14:23:18] <luna​r_sam> how do you like the disks being sqlite DBs lol
L193[14:23:36] <Redston​eParkour> weird but quite nice actually with how the hypervisor handles filesystems
L194[14:24:06] <Redston​eParkour> (its very plan9-like fyi)
L195[14:24:21] <luna​r_sam> also gonna have some more normal block storage devices but that's for a later date
L196[14:26:22] <luna​r_sam> but yeah, specifically failing on
L197[14:26:26] <luna​r_sam> https://paste.pc-logix.com/zuhocaguse
L198[14:26:27] <luna​r_sam> so lol
L199[14:26:46] <Redston​eParkour> oh that's some weird stuff
L200[14:27:47] <luna​r_sam> yeah
L201[14:27:56] <luna​r_sam> gonna see what it outputs first
L202[14:31:44] <Corded> > <Squi​dDev> https://github.com/CaitlynMainer/OpenComputers/blob/0857b39���
L203[14:31:44] <Mic​hiyo> 🤮
L204[14:31:54] <luna​r_sam> you know, i'm about to just say "wontfix" and move on with my day
L205[14:32:04] <luna​r_sam> my brain ain't cut out for this today
L206[14:33:54] <luna​r_sam> oh i forgot to set `sandbox._G`
L207[14:34:06] <luna​r_sam> (this is not the first thing i forgot to pass to the sandbox)
L208[14:34:20] <Amanda> %choose repair the ship or save the money for trading
L209[14:34:20] <MichiBot> Ama​nda: I have a pamphlet that says never to engage in "repair the ship", so you should definitely do it!
L210[14:35:11] <Corded> > <luna​r_sam> (this is not the first thing i forgot to pass to the sandbo…
L211[14:35:12] <luna​r_sam> yeah, like how i forgot pcall
L212[14:37:04] <Mic​hiyo> https://paste.pc-logix.com/ojilebejag
L213[14:38:31] <Redston​eParkour> bruh
L214[14:38:48] <luna​r_sam> `checkerror("invalid capture index %%2", string.gsub, "alo", ".", "%2")`
L215[14:38:51] <luna​r_sam> epic, this is failing
L216[14:38:58] <Redston​eParkour> also `with upgradeRenderer`?
L217[14:39:07] <Redston​eParkour> err, UpgradeRenderer
L218[14:39:11] <spo​nge> i was thinking that maybe SimpleItem should be made a class
L219[14:39:14] <Redston​eParkour> but i doubt every single item is an upgrade
L220[14:39:31] <Amanda> uh...
L221[14:39:41] <Amanda> Well that's special...
L222[14:40:00] <Redston​eParkour> also uhh these reimplementations do nothing
L223[14:40:01] <Redston​eParkour> what the fuck
L224[14:40:04] <Amanda> I used station services to repair my ship.. it seems to have printed an entire second ship underneath it
L225[14:40:27] <Redston​eParkour> buy one get one free
L226[14:40:51] <luna​r_sam> actually, even funnier
L227[14:40:57] <luna​r_sam> it's not even failing in the intended way
L228[14:40:57] <spo​nge> there's a bunch of item classes that do nothing
L229[14:41:07] <Amanda> no wonder the minor damage cost 84k credits
L230[14:41:26] <luna​r_sam> it's actually failing on pattern matching against the error
L231[14:41:38] <luna​r_sam> fucking comedy
L232[14:41:50] <Corded> > <spo​nge> there's a bunch of item classes that do nothing
L233[14:41:50] <spo​nge> those can be removed and we can just instantiate a SimpleItem
L234[14:42:13] <spo​nge> i might work on that when i'm
L235[14:42:15] <spo​nge> home
L236[14:43:40] <luna​r_sam> oh this is a simple error to fix
L237[14:47:11] <luna​r_sam> oh that's really funny
L238[14:49:01] <luna​r_sam> it's failing on `checkerror("invalid use of '%%'", string.gsub, "alo", ".", "%x")` because it successfully returns `xxx`
L239[14:50:26] <Redston​eParkour> lol
L240[14:51:51] <luna​r_sam> but that is a shrimple enough bug to fix
L241[14:57:08] <luna​r_sam> oh huh
L242[14:58:18] <luna​r_sam> it's testing returning the original string from gsub if there's no matches
L243[14:58:21] <luna​r_sam> that's a wontfix
L244[15:05:15] ⇨ Joins: vidak (~vidak@2407:e400:7800:2c01:d0be:76f8:cc84:bd4a)
L245[15:07:33] <luna​r_sam> also gonna throw in both lua 5.2's math tests and lua 5.4's math tests
L246[15:07:40] <Redston​eParkour> a bit closer https://discordembeds.pc-logix.com/live/1534578644641648710/1785942460154-4c6501a29ea0.png
L247[15:07:49] <luna​r_sam> since i implemented functions that were removed from 5.3+
L248[15:08:16] <luna​r_sam> and just commenting out the 5.2 tests that fail due to 5.3's changes to numbers
L249[15:16:16] <Mic​hiyo> Working on cleaning up the mess that create left.
L250[15:16:37] <Mic​hiyo> I've got corrupt jars in my gradle cache, and flywheel is failing to load.
L251[15:17:55] <luna​r_sam> https://discordembeds.pc-logix.com/live/1534581226176905416/1785943075572-30e4bec4b1dd.png
L252[15:18:01] <luna​r_sam> awesome
L253[15:18:21] <luna​r_sam> only one 5.2 test failed because of changes to integers so i just
L254[15:18:24] <luna​r_sam> commented it out
L255[15:18:25] <luna​r_sam> lol
L256[15:18:44] <luna​r_sam> and the other required a minor modification because it was float behavior (`0` to `0.0`)
L257[15:27:27] <luna​r_sam> bah i have to actually fix the change in 5.3
L258[15:51:21] <luna​r_sam> right, might have to make a custom xpcall to prevent the yielding bugs
L259[15:52:20] <luna​r_sam> since it can occasionally try to yield in the xpcall error handler that literally only sets two variables lol
L260[15:55:19] <Mic​hiyo> sigh fucking maven
L261[16:01:18] <Mic​hiyo> (╯°□°)╯︵ ┻━┻
L262[16:12:21] <spo​nge> what now
L263[16:12:23] <spo​nge> for me it works fine
L264[16:12:42] <spo​nge> +is it
L265[16:13:51] <Mic​hiyo> I don't have the energy to explain it. I pushed the build.gradle back to before my experiments.
L266[16:29:16] <SquidDev> Everything about ExtendedRecipe is so cursed :/.
L267[16:30:26] <SquidDev> Ironically CC:T actually has a system to handle most of what it's trying to do, but I really don't want to have to port that to Scala :D:.
L268[16:30:26] <Mic​hiyo> %s/ExtendedRecipe/this port/
L269[16:30:26] <MichiBot> <SquidDev> Everything about this port is so cursed :/.
L270[16:33:04] <SquidDev> This bit I think can be blamed on Sangar, but some of it's just inevitable when the core of a mod hasn't been touched in 10+ years.
L271[16:40:13] <Redston​eParkour> ok the input is being eaten...
L272[16:40:14] <Redston​eParkour> weird
L273[16:44:47] <Corded> > <Redston​eParkour> ok the input is being eaten...
L274[16:44:47] <luna​r_sam> have you been using the "new computer" button
L275[16:44:54] <luna​r_sam> if so, i think it's broken and i dunno how i pulled that one off
L276[16:45:03] <Redston​eParkour> i've been restarting the server every time
L277[16:45:08] <luna​r_sam> ah
L278[16:45:09] <Redston​eParkour> but i believe refreshing + new computer also works
L279[16:45:12] <luna​r_sam> yeah
L280[16:45:21] <luna​r_sam> that's what i've been doing lol
L281[16:45:32] <Redston​eParkour> oh nvm
L282[16:45:37] <Redston​eParkour> i see the issue
L283[16:45:54] <Redston​eParkour> my tty read -> key_down thing only process one char but could read multiple
L284[16:47:06] <Redston​eParkour> yea seems to work now mostly
L285[16:47:41] <Corded> > <Redston​eParkour> my tty read -> key_down thing only process one char but cou…
L286[16:47:41] <luna​r_sam> oh, yeah
L287[16:47:55] <luna​r_sam> `next` tries to return the next entire keypress
L288[16:47:58] <Redston​eParkour> its probably totally my fault
L289[16:48:02] <Redston​eParkour> no not that
L290[16:48:04] <luna​r_sam> oh
L291[16:48:07] <Redston​eParkour> it goes through a read buffer first
L292[16:48:13] <luna​r_sam> yeah `read()` without any args--oh
L293[16:48:14] <luna​r_sam> lmao
L294[16:48:35] <luna​r_sam> currently adding a way to have critical regions of code (for machine.lua)
L295[16:48:47] <luna​r_sam> because turns out
L296[16:48:56] <luna​r_sam> yeah there's a lot of places the automagic yield will break
L297[16:49:04] <Corded> > <Squi​dDev> Ironically CC:T actually has a system to handle most of wha…
L298[16:49:05] <as​ie> honestly
L299[16:49:11] <as​ie> the less Scala code OC has the better
L300[16:49:19] <Mic​hiyo> You're NOT wrong.
L301[16:49:35] <as​ie> so if you want to yank out some part of OC and replace it with Java code, that's a win-win to me
L302[16:49:38] <Mic​hiyo> I'm considering slowly porting the scala out... lol
L303[16:49:40] <as​ie> it means more people can contribute to that part
L304[16:49:49] <luna​r_sam> there's a few entire functions that'll be marked as critical (mostly `table.sort`)
L305[16:50:14] <Corded> > <luna​r_sam> yeah there's a lot of places the automagic yield will break
L306[16:50:14] <Redston​eParkour> yeah i've already had to work around that a bit
L307[16:50:30] <luna​r_sam> yeah i'm adding another vmstate flag
L308[16:50:33] <Redston​eParkour> (the disk reader crashing because its yielding in the for loop's iterator)
L309[16:50:54] <Corded> > <Redston​eParkour> (the disk reader crashing because its yielding in the for l…
L310[16:50:54] <luna​r_sam> LMFAO WHAT
L311[16:50:56] <luna​r_sam> that's wild
L312[16:51:21] <Redston​eParkour> a for loop is C code
L313[16:51:27] <Redston​eParkour> you can also get that in openos btw
L314[16:52:04] <luna​r_sam> man
L315[16:52:31] <luna​r_sam> perhaps i should make the hook trigger on lines
L316[16:52:43] <luna​r_sam> instead of instructions
L317[16:52:59] <Redston​eParkour> maybe
L318[16:53:06] <Redston​eParkour> buuut that can be defeated quite quickly
L319[16:53:17] <luna​r_sam> i mean, yeah
L320[16:53:21] <luna​r_sam> there's still the hard deadline
L321[16:53:23] <Redston​eParkour> maybe do bytecode patching to rewrite for loops into the pure lua equivalent
L322[16:53:35] <luna​r_sam> that's horrible
L323[16:53:43] <luna​r_sam> i hope it doesn't come to that
L324[16:55:27] <as​ie> honestly
L325[16:55:38] <as​ie> the ideal solution is to rebuild JNLua into Java bytecode, using say WASM as an intermediate step
L326[16:55:53] <as​ie> then you can get rid of the whole Eris patchset and just use vanilla Lua with the WASM virtual machine acting as the sandbox
L327[16:56:06] <as​ie> this would make Lua updates much easier to merge in, too
L328[16:56:20] <as​ie> at least, that's what I think
L329[16:56:22] <Redston​eParkour> hmm
L330[16:56:27] <Redston​eParkour> that could work yeah
L331[16:56:49] <luna​r_sam> i'm doing all my mess in C#
L332[16:57:30] <luna​r_sam> right, now, HOPEFULLY error handlers work
L333[16:57:46] <Redston​eParkour> i've yet to update so we'll see
L334[16:58:05] <Kristo​bear38> And we could finally have preemptive multitasking by way of delegating this to the WASM VM
L335[16:58:10] <Redston​eParkour> anyways got ctrl+c working https://discordembeds.pc-logix.com/live/1534606452403081357/1785949090036-613173dc80a1.png
L336[16:58:16] <Redston​eParkour> weirdly the input appears top-left
L337[16:58:18] <Corded> > <Redston​eParkour> i've yet to update so we'll see
L338[16:58:18] <luna​r_sam> i have yet to push it
L339[16:58:19] <Redston​eParkour> dont ask
L340[16:58:34] <Corded> > <Redston​eParkour> anyways got ctrl+c working
L341[16:58:34] <luna​r_sam> wew lad
L342[16:58:51] <Redston​eParkour> also yes the background's gray
L343[16:58:53] <Redston​eParkour> also don't ask
L344[16:59:01] <luna​r_sam> yeah i know about that one
L345[16:59:11] <luna​r_sam> that just happens
L346[16:59:28] <luna​r_sam> now i get to make a buncha changes to machine.lua again
L347[17:01:26] <Redston​eParkour> now i'm going to have to implement `gpu.copy`...
L348[17:01:59] <luna​r_sam> f
L349[17:02:25] <Redston​eParkour> ah well let's hope the screen buffer code works
L350[17:04:31] <luna​r_sam> trying to think if i have to mark any other functions as critical
L351[17:04:34] <luna​r_sam> other than table.sort
L352[17:05:05] <luna​r_sam> and the error handler used in ypcall
L353[17:05:32] <Redston​eParkour> metamethods perhaps?
L354[17:05:48] <luna​r_sam> `fuck`
L355[17:06:12] <luna​r_sam> yeah those aren't yieldable, are they?
L356[17:06:16] <Redston​eParkour> maybe
L357[17:06:18] <Redston​eParkour> maybe not
L358[17:06:26] <Redston​eParkour> maybe you could bytecode-patch them with the pure-lua equivalent
L359[17:06:33] <luna​r_sam> ACTUALLY
L360[17:06:38] <luna​r_sam> WAIT
L361[17:06:40] <Redston​eParkour> although
L362[17:06:44] <Redston​eParkour> there's no pure-lua equivalent
L363[17:06:47] <luna​r_sam> i think they are MOSTLY yieldable
L364[17:06:48] <Redston​eParkour> wait hmm
L365[17:07:01] <Redston​eParkour> the pure-lua equivalent would require intercepting `setmetatable` that's for sure
L366[17:07:07] <luna​r_sam> (they just can't yield across c boundaries)
L367[17:07:12] <Redston​eParkour> and providing a `rawgetmetatable` that ignores the `__metatable`
L368[17:07:23] <Redston​eParkour> then insert that as an upvalue
L369[17:07:27] <luna​r_sam> so don't do anything in C code lol
L370[17:07:41] <Redston​eParkour> hmm you could be right
L371[17:07:44] <luna​r_sam> any edge cases are
L372[17:07:50] <luna​r_sam> a future me's problme
L373[17:07:51] <Redston​eParkour> as the operator basically tail-calls the metamethod or something
L374[17:07:52] <luna​r_sam> *problem
L375[17:08:09] <Redston​eParkour> hmmmmm
L376[17:08:18] <Redston​eParkour> `tostring` and `pairs`?
L377[17:08:55] <Corded> > <Kristo​bear38> And we could finally have preemptive multitasking by way of…
L378[17:08:55] <Kristo​bear38> Prior research, notably all the listed VMs now support GC extension: https://gitlab.com/Hawk777/oc-wasm/-/boards?show=eyJpaWQiOiI4IiwiZnVsbF9wYXRoIjoiSGF3azc3Ny9vYy13YXNtIiwiaWQiOjE1OTAwNzA4Mn0%3D#/
L379[17:35:36] <as​ie> we don't need GC
L380[17:35:43] <as​ie> and Lua doesn't do preemptive multitasking AFAIK
L381[17:40:25] <Corded> > <Squi​dDev> Uhhh, the 1.21.1 port just entirely deleted the access poin…
L382[17:40:25] <SoraFi​restorm> Yeah, that got merged into one block - the relay
L383[17:42:11] <Redston​eParkour> well uhh close i guess https://discordembeds.pc-logix.com/live/1534617530071777352/1785951731285-2657693bb5ed.png
L384[17:43:21] <luna​r_sam> `error in critical region: C stack overflow`
L385[17:43:22] <luna​r_sam> awesome
L386[17:43:24] <Redston​eParkour> yay
L387[17:43:28] <luna​r_sam> how'd i do this one
L388[17:43:36] <Redston​eParkour> no clue
L389[17:43:51] <Redston​eParkour> C stack even
L390[17:44:07] <Redston​eParkour> tried increasing the stack size ulimit?
L391[17:44:21] <luna​r_sam> wait hold up
L392[17:44:24] <luna​r_sam> i may be stupid
L393[17:45:01] <Redston​eParkour> oh that's an issue
L394[17:45:04] <luna​r_sam> i was indeed stupid
L395[17:45:05] <Redston​eParkour> ctrl+w closes the tab
L396[17:45:11] <Redston​eParkour> so you can't exit edit
L397[17:45:13] <luna​r_sam> oops lol
L398[17:45:24] <luna​r_sam> yeah that's gonna uh
L399[17:45:31] <luna​r_sam> be a problem for later me
L400[17:45:34] <luna​r_sam> when i write an editotr
L401[17:45:35] <Redston​eParkour> fair enough
L402[17:45:55] <luna​r_sam> anyways, critical region code seems to™ work
L403[17:46:15] <luna​r_sam> i also fixed the hard deadline code by uh
L404[17:46:20] <luna​r_sam> ...making it larger
L405[17:46:31] <luna​r_sam> (also it yields after exiting the critical region)
L406[17:47:02] <luna​r_sam> https://paste.pc-logix.com/xifolodafi
L407[17:47:20] <luna​r_sam> initially i had the n_args assignment after pushing the upvalue onto the stack
L408[17:47:21] <luna​r_sam> so
L409[17:47:27] <luna​r_sam> that's what was causing my previous errors
L410[17:47:37] <luna​r_sam> of c stack overflow
L411[17:47:47] <Redston​eParkour> right
L412[17:47:59] <Redston​eParkour> because it tries to access an undefined stack element or something
L413[17:48:01] <Redston​eParkour> so UB
L414[17:48:09] <luna​r_sam> yeah
L415[17:48:13] <Corded> > <as​ie> the ideal solution is to rebuild JNLua into Java bytecode, …
L416[17:48:13] <SoraFi​restorm> At some future point that is definitely at least a few months off, I feel like making wasm the default arch internally and preserving the existing Lua stuff via some sort of backwards compatibility layer would make a lot of sense
L417[17:48:19] <Redston​eParkour> weird that it just overflowed the stack
L418[17:48:21] <SoraFi​restorm> It'd probably just be a total PITA to all make work
L419[17:48:31] <Corded> > <SoraFi​restorm> At some future point that is definitely at least a few mont…
L420[17:48:31] <as​ie> the problem is then preserving the Lua APIs
L421[17:48:31] <Redston​eParkour> and not the funny `segmentation fault (core dump)`
L422[17:48:36] <luna​r_sam> UB is more like magic behavior
L423[17:48:37] <as​ie> you don't want to pull an OC2
L424[17:49:03] <Corded> > <Redston​eParkour> and not the funny `segmentation fault (core dump)`
L425[17:49:04] <luna​r_sam> or if it somehow threw an uncaught error to the C# layer, `abort (core dump)`
L426[17:49:09] <SoraFi​restorm> Especially because we currently don't have a wide body of folks around to really hammer and be sure it stays compatible
L427[17:49:16] <Corded> > <as​ie> the problem is then preserving the Lua APIs
L428[17:49:16] <SoraFi​restorm> right yeah
L429[17:49:18] <luna​r_sam> oh yeah, i was debugging a bus fault error i was getting on the byond side
L430[17:49:36] <luna​r_sam> turns out, if you overwrite a library that byond has loaded via `load_ext`
L431[17:49:43] <luna​r_sam> and call that function handle
L432[17:49:49] <luna​r_sam> it'll just bus error
L433[17:49:54] <SoraFi​restorm> and the motivation is like... Lua is a language I do like, but I would prefer a different one
L434[17:49:54] <luna​r_sam> awesome byond moment
L435[17:50:20] <Redston​eParkour> well
L436[17:50:38] <Redston​eParkour> if it's close enough to OC i could probably make it OC
L437[17:51:01] <SquidDev> Does anyone know if the new flat screens in the 1.21 version are from an existing mod, or were vibed in? The recipes look like they have some copy-paste errors, but I'm not sure if it's intentional.
L438[17:51:02] <luna​r_sam> you've sure gotten close with my env that's not even that close to OC :v
L439[17:51:42] <Corded> > <Squi​dDev> Does anyone know if the new flat screens in the 1.21 versio…
L440[17:51:43] <Mic​hiyo> OpenScreens
L441[17:52:18] <luna​r_sam> i'm also going back and removing a bunch of debug prints
L442[17:53:38] <luna​r_sam> my computer locked up for a sec and i got a magic message in the terminal
L443[17:53:40] <luna​r_sam> `Watchdog triggered (268.6889199999996x limit)
L444[17:53:45] <luna​r_sam> * limit)`
L445[17:53:52] <SquidDev> Thanks!
L446[17:55:58] <luna​r_sam> i do love that xterm.js sometimes bugs out
L447[17:56:05] <luna​r_sam> and prints messages out of order
L448[17:56:10] <Redston​eParkour> is that xterm?
L449[17:56:34] <Redston​eParkour> i believe its the websocket or the stuff around it that's causing the reordering
L450[17:57:08] <luna​r_sam> pain, suffering even
L451[17:57:18] <luna​r_sam> agony, if you will
L452[17:57:56] <luna​r_sam> i plan to eventually™ make my own terminal emulator that i can more easily persist on the C# side
L453[17:58:38] <Redston​eParkour> uh oh
L454[17:58:44] <luna​r_sam> uh oh?
L455[17:58:49] <Redston​eParkour> its a terminal emulator
L456[17:59:04] <Redston​eParkour> will it do the \27[H the right or wrong way?
L457[17:59:14] <luna​r_sam> the right way
L458[17:59:28] <Mic​hiyo> sigh I don't understand this person's aversion to posting their damn MC log....
L459[17:59:39] <Spider​EveryOS> %tonk
L460[17:59:39] <MichiBot> Consarn it! Spider​EveryOS! You beat Forec​aster's previous record of 2 hours, 25 minutes and 39 seconds (By 2 hours, 58 minutes and 22 seconds)! I hope you're happy!
L461[17:59:40] <MichiBot> SpiderEveryOS's new record is 5 hours, 24 minutes and 1 second! SpiderEveryOS also gained 0.00891 (0.00297 x 3) tonk points for stealing the tonk. Position #1.
L462[17:59:45] <Mic​hiyo> I don't need snippets, and what yoiu think the issue is. I NEED THE LOG.
L463[17:59:46] <luna​r_sam> while the `y;x` order is confusing, gotta support it
L464[17:59:51] <Mic​hiyo> * you
L465[18:00:40] <SquidDev> My issue template has flaming animated text saying "Upload your logs", and people still don't two times out of three.
L466[18:00:58] <Mic​hiyo> I've asked 3 TIMES in this issue for the logs.
L467[18:01:02] <Corded> > <Mic​hiyo> sigh I don't understand this person's aversion to posting…
L468[18:01:03] <Redston​eParkour> probably out of fear some kind of personal info is in there
L469[18:01:54] <luna​r_sam> exploding people who don't post logs with my mind
L470[18:02:04] <Mic​hiyo> %stacktrace
L471[18:02:05] <MichiBot> http://i.imgur.com/XffI6QA.jpg {0}
L472[18:03:10] <luna​r_sam> "your mod stopped working"
L473[18:03:10] <luna​r_sam> post logs
L474[18:03:10] <luna​r_sam> 🦗 🦗 🦗 🦗 🦗 🦗 🦗 🦗
L475[18:04:44] <luna​r_sam> even worse is when your mod has conflicts with another mod but the other mod has no sause available
L476[18:05:09] <Redston​eParkour> optifine
L477[18:05:12] <luna​r_sam> sure i could decompile but what a fucking pain in the ass it is
L478[18:05:33] <Corded> > <Redston​eParkour> optifine
L479[18:05:33] <luna​r_sam> i'm convinced some unity game mods have no sause out of spite
L480[18:05:53] <luna​r_sam> when people have a github repo with just a readme and releases for the binaries, that really gets me angry
L481[18:07:27] <Redston​eParkour> ok so i think the screen corruption is because of the tty write reordering
L482[18:08:21] <Redston​eParkour> also it lags like hell
L483[18:09:42] <Corded> > <as​ie> we don't need GC
L484[18:09:42] <Kristo​bear38> I'm thinking ahead; it's as @SoraFirestorm says, having lua is nice and all, but having WASM is an order of magnitude nicer as you can use any language that compiles to wasm which is a whole lot nowadays. The catch is a few of them that I like require GC, tailcall and whatnot extensions. So if we're picking a wasm runtime, I'd
L485[18:09:42] <Corded> prefer it has these. Then we can do a wasm arch, and either run Lua on top of it or have a separate Lua arch that internally runs on wasm runtime as an implementation detail. And the point about preemptive multitasking is about doing that on a WASM VM level, not on Lua level.
L486[18:09:57] <Redston​eParkour> there's OC-WASM
L487[18:10:02] <Redston​eParkour> though that's your thing right?
L488[18:10:13] <Kristo​bear38> it's Hawk777's
L489[18:10:17] <Redston​eParkour> oh whoops
L490[18:10:35] <Corded> > <Kristo​bear38> I'm thinking ahead; it's as @SoraFirestorm says, having lua…
L491[18:10:35] <as​ie> I really wouldn't want OC1 to morph into OC2, is the concern
L492[18:10:48] <Redston​eParkour> so basically make it a really big implementation detail?
L493[18:11:14] <Corded> > <Redston​eParkour> probably out of fear some kind of personal info is in there
L494[18:11:14] <SoraFi​restorm> Your system username - which is typically someone's actual real name - does in fact leak in this situation
L495[18:11:33] <Mic​hiyo> https://github.com/CaitlynMainer/OpenComputers/issues/9#issuecomment-5195170634 IDK if they're trying to imply it's our issue or not lol
L496[18:11:34] <MichiBot> Title: the tier 3 case when i hold it / drop it its invisible | Posted by: MetalDev-why | Posted: Wed Aug 05 03:54:59 UTC 2026 | Status: open
L497[18:11:40] <Mic​hiyo> My windows username is "Michi" :P
L498[18:11:56] <Mic​hiyo> my Linux username is Michiyo lol
L499[18:12:13] <SoraFi​restorm> my usernames on my personal computers haven't been my real name for like a decade
L500[18:12:14] <Mic​hiyo> Be smart people lol
L501[18:12:22] <Redston​eParkour> that looks like shaders
L502[18:12:26] <Mic​hiyo> 💯
L503[18:12:30] <SoraFi​restorm> but I am a weirdo
L504[18:12:35] <Redston​eParkour> ask for a modlist at least
L505[18:13:09] <Corded> > <as​ie> I really wouldn't want OC1 to morph into OC2, is the concern
L506[18:13:09] <Kristo​bear38> The difference between my idea and OC2 would be that we'd still retain the same components and their methods
L507[18:13:23] <Corded> > <as​ie> I really wouldn't want OC1 to morph into OC2, is the concern
L508[18:13:23] <SoraFi​restorm> Yeah, I also don't really want this, and we would need to be careful on implementation
L509[18:13:32] <SoraFi​restorm> This is all pie in the sky stuff since there is like
L510[18:13:36] <Kristo​bear38> so you wouldn't run linux on it; the vibe would still be very much "implement your own little scripts"
L511[18:14:14] <SoraFi​restorm> Michiyo and Squid (competent), myself (not very competent), and maybe like 1 or 2 other people who are working on this
L512[18:14:20] <Redston​eParkour> well i wonder kristobear
L513[18:14:27] <Redston​eParkour> would it be possible to run ocvm on oc2
L514[18:14:28] <Elfi> I'm a sicko who would absolutely wanna run a lisp machine in MC
L515[18:14:32] <SoraFi​restorm> We don't have the manpower to consider this
L516[18:14:42] <Corded> > <El​fi> I'm a sicko who would absolutely wanna run a lisp machine i…
L517[18:14:42] <SoraFi​restorm> also me
L518[18:14:53] <Redston​eParkour> elfi: there's a lisp for lua on oppm
L519[18:15:01] <Mic​hiyo> competent? Me? your bar is low.
L520[18:15:16] <Brisingr​ Aerowing> My username most places is my fursona's name.
L521[18:15:25] <Kristo​bear38> yes and no; look at cosmosprime's port - don't underestimate the power of focused neurodivergents
L522[18:15:28] <SoraFi​restorm> Maybe, but you clearly know JVM land better than I do!
L523[18:15:53] <Corded> > <Redston​eParkour> elfi: there's a lisp for lua on oppm
L524[18:15:53] <SoraFi​restorm> Urn or Fennel? Or is this some other thing?
L525[18:15:59] <Redston​eParkour> fennel yeah
L526[18:16:08] <Mic​hiyo> The best I can say, is I was at least writting MC Mods before LLMs were a thing. lol
L527[18:16:16] <as​ie> WASM is an implementation detail here; the crux is "move from Eris, serializing the Lua state, and relying on machine.lua to control timeouts, to a layer the Lua C virtual machine runs on, which is itself serialized and timeouted"
L528[18:16:27] <as​ie> this would fix a lot of hacks machine.lua has to do to ensure computers generally time out correctly
L529[18:16:29] <Redston​eParkour> that could work
L530[18:16:36] <as​ie> and it would allow using vanilla Lua as opposed to Eris
L531[18:17:05] <Redston​eParkour> hmmm
L532[18:17:09] <luna​r_sam> man i shoulda did this for my project
L533[18:17:13] <SoraFi​restorm> right, but I'm saying at that point you might as well go another mile or 2 and just make WASM an exposed arch
L534[18:17:13] <Kristo​bear38> there are WASM implementations that allow preempting execution and serializing state I believe
L535[18:17:17] <Redston​eParkour> so that'd mean its the lua vm on the wasm vm on the java vm
L536[18:17:19] <as​ie> yes
L537[18:17:22] <Redston​eParkour> triple vm
L538[18:17:29] <as​ie> the problem is, if you just run a WASM interpreter, it might cause considerable performance overhead
L539[18:17:31] <SoraFi​restorm> very virtual
L540[18:17:36] <as​ie> but it would make for a good proof of concept at least
L541[18:17:47] <as​ie> a JIT for the WASM layer would be much better, but you can throw one inbetween later
L542[18:18:05] <as​ie> this is also why I wouldn't want users to throw in WASM code themselves - a JIT could carry with itself security risks
L543[18:18:38] <Kristo​bear38> I thought the WASM runtime was supposed to be secure? ;)
L544[18:18:48] <as​ie> so was Lua but DoS exploits for OC happened
L545[18:18:53] <as​ie> because of the machine.lua hacks
L546[18:19:09] <as​ie> also, once you add the WASM layer, if there's a Java interpreter fallback, you can drop LuaJ completely
L547[18:19:10] <luna​r_sam> yeah i'm kinda worried about that for my project
L548[18:19:11] <as​ie> that is too very good
L549[18:19:15] <as​ie> * that, too, is
L550[18:19:26] <SoraFi​restorm> Fennel is kinda not ideal to run in OC because it takes so much damn memory to run the system
L551[18:19:28] <as​ie> it will mean that even on machines which don't support natives, OC1 will work fine
L552[18:19:42] <Corded> > <SoraFi​restorm> Fennel is kinda not ideal to run in OC because it takes so …
L553[18:19:42] <as​ie> the RAM sticks need to be rebalanced IMO
L554[18:19:46] <Kristo​bear38> yes because Lua is insecure out of the box and you have to make an effort to make it secure, whereas wasm was designed to be secure sandbox from the beginning
L555[18:19:55] <SoraFi​restorm> Even after having gone through and done some compiler optimization work for them a couple years ago
L556[18:20:08] <Redston​eParkour> compiler?
L557[18:20:13] <Redston​eParkour> wait fennel compiles lisp to lua?
L558[18:20:17] <SoraFi​restorm> Yes
L559[18:20:18] <Kristo​bear38> anywa, lucky you, graalwasm is written in Java and it does JIT
L560[18:20:19] <Redston​eParkour> daaamn
L561[18:20:34] <SoraFi​restorm> It's basically Lua with a Lispy coat of paint
L562[18:20:36] <Redston​eParkour> but yeah that'd explain the ram eating
L563[18:20:55] <SoraFi​restorm> Because it deliberately maintains most of Lua's semantics
L564[18:20:59] <Corded> > <Kristo​bear38> anywa, lucky you, graalwasm is written in Java and it does …
L565[18:20:59] <as​ie> but can it do interruptions for CPU timeouts, and can it do memory restrictions?
L566[18:21:10] <as​ie> ... and does it run on a standard JVM?
L567[18:22:57] <Redston​eParkour> fixed redrawing mostly, but copying's still broken... https://discordembeds.pc-logix.com/live/1534627789054152826/1785954177194-86436e2e7ee0.png
L568[18:22:58] <Corded> > <luna​r_sam> yeah i'm kinda worried about that for my project
L569[18:22:58] <luna​r_sam> like, at least, on one hand, the lua instances run in a separate process from byond (for multiple reasons)
L570[18:23:06] <luna​r_sam> but on the other hand, i don't trust my code
L571[18:23:10] <Redston​eParkour> hmm
L572[18:23:19] <Kristo​bear38> I need to look into timeouts; as for memory restrictions iirc those are built into WASM spec itself but don't take my word for granted
L573[18:23:24] <Redston​eParkour> have the lua instances setuid to another user?
L574[18:23:31] <Elfi> Wait, you actually got an LVM interacting with BYOND? How'd you manage that?
L575[18:23:47] <Redston​eParkour> something something websockets i believe
L576[18:23:47] <luna​r_sam> i mean, in this case
L577[18:23:58] <Kristo​bear38> but uh you're right, jit relies on running on graalvm
L578[18:24:05] <luna​r_sam> the lua instances are contained in a C# process
L579[18:24:15] <luna​r_sam> which talks to byond over a unix socket
L580[18:24:23] <luna​r_sam> (a bit of C++ glue logic on the byond side)
L581[18:24:23] <Elfi> That must make developing for SS13 a little less miserable
L582[18:24:29] <luna​r_sam> it does not
L583[18:24:31] <Redston​eParkour> (or to a webbrowser-emulator over websockets)
L584[18:24:35] <Elfi> Of course it doesn't
L585[18:24:43] <Corded> > <Redston​eParkour> (or to a webbrowser-emulator over websockets)
L586[18:24:43] <luna​r_sam> yeah this has been a godsend
L587[18:24:47] <luna​r_sam> i'm so glad i did it this way
L588[18:24:58] <Elfi> God couldn't make SS13 development less miserable
L589[18:24:58] <luna​r_sam> just having it to the tty over a websocket
L590[18:25:15] <luna​r_sam> byond got webview2 so i've been embracing it
L591[18:25:57] <luna​r_sam> after i get a minimum viable product here, i'm gonna rip apart the character editor
L592[18:26:16] <luna​r_sam> but yeah, i previously built a lua layer into extools back in the byond 513 days
L593[18:27:56] <Corded> > <Brisingr​ Aerowing> reacted 👍 to <as​ie> this would fix a lot of hacks machine.lua has to do to ensure computers gener…
L594[18:28:14] <Corded> > <as​ie> the RAM sticks need to be rebalanced IMO
L595[18:28:15] <SoraFi​restorm> In fairness, the Fennel system is like 300k of code
L596[18:29:08] <SoraFi​restorm> +source
L597[18:29:27] <SoraFi​restorm> And could probably use more compiler optimization work
L598[18:30:55] <Brisingr​ Aerowing> I've been looking at the Teal language as well.
L599[18:41:07] *** Joins: Michiyo (~Michiyo@50.53.94.126) Michiyo
L600[18:41:07] zsh sets mode: +o on Michiyo
L601[18:42:00] <SquidDev> Not content with opening one useless issue without logs, they've opened two :/.
L602[18:43:25] <Mic​hiyo> sigh yeah
L603[18:46:49] <Corded> > <SoraFi​restorm> And could probably use more compiler optimization work
L604[18:46:49] <SoraFi​restorm> Just that the really easy low-hanging fruit fixes I already did the last time around
L605[18:46:50] <Redston​eParkour> they said something really weird at the end of the second issue
L606[18:47:00] <Redston​eParkour> > in the file manager i made it wont show the entire log thing because i made it automatically show where the problem is and forgot that i might need the entire file log thing one day lol
L607[18:47:18] <luna​r_sam> what
L608[18:47:26] <SoraFi​restorm> ???
L609[18:47:41] <luna​r_sam> also i might just start making patches for lua for my usecase tbh
L610[18:47:44] <Mic​hiyo> yeah no damn clue. lol <a:RubyShrug:1488397610195615794>
L611[18:48:02] <luna​r_sam> but now i have another dependency that has to compile on windows
L612[19:07:10] ⇦ Quits: Corded (~MichiBot@2604:2dc0:200:54::2) (Quit: Bridge shutting down)
L613[19:08:07] *** Joins: Corded (~MichiBot@2604:2dc0:200:54::2) Michibot
L614[19:08:07] zsh sets mode: +v on Corded
L615[19:09:15] <SoraFi​restorm> You can AOT compile Fennel to Lua, so it's still possible to run Fennel
L616[19:10:27] <Corded> > <Redston​eParkour> fennel yeah
L617[19:10:28] <SoraFi​restorm> I didn't realize that someone was shipping Fennel on OPPM... I probably wouldn't use that if it hasn't been updated in a while
L618[19:11:07] <SoraFi​restorm> One of my compiler optimizations fixed some dramatically awful memory churn in some circumstances
L619[19:12:10] <SoraFi​restorm> * compiler/codegen
L620[19:17:19] <Amanda> %choose get a second load?
L621[19:17:19] <MichiBot> Ama​nda: "get a second load" doesn't really seem like a good idea right now.
L622[19:19:14] <Amanda> oh fml
L623[19:19:25] <Amanda> I didn't dock my miner right, so now it's been left behind in the jump
L624[19:23:13] <Mic​hiyo> oof
L625[19:29:06] <Forec​aster> It's gone forever now
L626[19:30:52] <Corded> > <SoraFi​restorm> Michiyo and Squid (competent), myself (not very competent),…
L627[19:30:52] <Forec​aster> I should join in and lower the average competence level further!
L628[19:32:38] <Corded> > <Mic​hiyo> reacted 💯 to <Forec​aster> I should join in and lower the average competence level further!
L629[19:34:46] <Amanda> bah, I give up. Stupid miner won't reconnect to be carried back, so I just yoinked it's fuel and o2
L630[19:40:52] <SquidDev> I forget how much *stuff* OC has. The recipe datagen is currently at 1.8k LOC, and I've only done the stuff I could automate migrating...
L631[19:42:16] <SoraFi​restorm> Yeah, OC is way more fleshed out, which is one of the appeals IMO
L632[19:44:07] <SquidDev> "Raw Circuit Board", "Circuit Board" (uncraftable, possibly deprecated?) and "Printed Circuit Board" is definitely fleshed out.
L633[19:44:30] <SoraFi​restorm> There is an alternate, harder recipe set
L634[19:45:03] <SoraFi​restorm> CB is from that
L635[19:45:56] <SquidDev> Ahh, yeah, the greg one has a separate step.
L636[19:46:31] <SoraFi​restorm> Grog used to be harder set only but later got used in the recipe for nanomachines
L637[19:46:57] <SoraFi​restorm> I think the recipe is different between normal and hard
L638[19:49:14] <SoraFi​restorm> It might possibly be worth not adding the hard mode items if not using hard mode
L639[19:50:50] <SquidDev> Noooo. Dynamic item registration is a Bad Idea.
L640[19:50:57] <SquidDev> One day we'll get datapack-driven items.
L641[19:52:53] <spo​nge> yeah i wish
L642[19:53:14] <SoraFi​restorm> I mean... I understand why, but look at what just happened because it wasn't like that: it made you sad and confused
L643[19:53:24] <SoraFi​restorm> Well... maybe not sad
L644[19:56:02] <SquidDev> Part of the issue here is that the greg recipes weren't included in the 1.21.1 version, so I just couldn't see any usages for it. Normally you can just grep for something in the code.
L645[19:56:23] <SoraFi​restorm> ... now that you have said that
L646[19:56:28] <SoraFi​restorm> Something just occurred to me
L647[19:56:52] <SoraFi​restorm> @Michiyo I think the port lost the ability to configure crafting recipes
L648[19:57:10] <SoraFi​restorm> The yaml/json/whatever-the-heck don't seem to generate anymore
L649[19:57:42] <SquidDev> That's what datapacks are for!
L650[19:57:52] <Mic​hiyo> sigh
L651[19:57:57] <SoraFi​restorm> Right sure
L652[19:58:11] <SoraFi​restorm> But I'm talking about a thing that used to exist and no longer does
L653[19:58:19] <SoraFi​restorm> So like
L654[19:58:38] <SoraFi​restorm> Is it worth not restoring that?
L655[19:58:56] <SoraFi​restorm> And just waiting to datapack-ize ut?
L656[19:59:25] <SoraFi​restorm> * it?
L657[19:59:34] <SquidDev> I'm not sure I understand what you're proposing? You can replace recipes now with datapacks — the thing which needs doing is migrating all the old *.recipe files to a datapack.
L658[19:59:55] <SoraFi​restorm> So there was an old system for doing recipes
L659[20:00:04] <SoraFi​restorm> I am musing that maybe instead of fixing it
L660[20:00:07] <SoraFi​restorm> We just don't
L661[20:00:19] <SoraFi​restorm> And do datapacks at some future point instead
L662[20:00:48] <SoraFi​restorm> So yeah, doing what you are suggesting and just transitioning now
L663[20:01:00] <SoraFi​restorm> Since the old system appears to be dysfunctional anyways
L664[20:07:02] <SoraFi​restorm> https://github.com/MightyPirates/OpenComputers/tree/master-MC1.7.10/src/main/resources/assets/opencomputers/recipes
L665[20:07:08] <SoraFi​restorm> From the original MP repo
L666[20:07:23] <SoraFi​restorm> Not sure if the files are still in OC:R... they probably are
L667[20:08:56] <SoraFi​restorm> Uhh... not obviously they aren't
L668[20:09:10] <SoraFi​restorm> If they are still there, they got moved
L669[20:09:36] <SoraFi​restorm> On my phone so difficult to go finding and grepping for things
L670[20:09:50] <SquidDev> No, they were removed.
L671[20:09:51] <Mic​hiyo> I regret picking this up a little more every damn time something like this comes up. lmao
L672[20:10:47] <Corded> > <Mic​hiyo> I regret picking this up a little more every damn time some…
L673[20:10:47] <SoraFi​restorm> That's why I'm trying to do what I can to be a useful contributor ^^;;
L674[20:11:05] <SoraFi​restorm> Especially early in the process where there's a lot of rolling up of the sleeves...
L675[20:38:28] ⇦ Quits: xal (~xal@mx1.xal.systems) (Quit: No Ping reply in 180 seconds.)
L676[20:38:35] ⇨ Joins: xal (~xal@mx1.xal.systems)
L677[20:39:59] <Michiyo> %join #oc-rebooted
L678[20:40:02] <MichiBot> Joined channel #oc-rebooted
L679[20:49:24] <Forec​aster> %sip
L680[20:49:24] <MichiBot> You drink a hexagonal aegisalt potion (New!). The potion contained a computer virus! But Forecaster's anti-virus routines destroy it.
L681[20:49:52] <Forec​aster> I can't believe someone would do that, put computer viruses into potions
L682[20:50:08] <Michiyo> Yeah Fore, I wonder how that happened, and who did it.
L683[20:52:36] <Forec​aster> We'll never know
L684[21:04:16] <Izzy> Michiyo: I wanna submit my quad GPU 1.12 backport some time, should I open a PR to fast forward the 1.12 branch to the MP/OC current state first?
L685[21:04:37] <Michiyo> 1.12 is marked as archived. I don't think you're gonna get it in.
L686[21:05:15] <Izzy> the 1.12 branch on OC:R?
L687[21:05:27] <Michiyo> Oh.
L688[21:05:47] <Michiyo> Umm... I have no idea if I'm going to do anything with sub 1.21.1 TBH... it's something I'm going to have to think on.
L689[21:05:57] <Izzy> Okiii
L690[21:07:33] <Michiyo> If you want to go ahead and submit the PR that's fine. But I'm just not sure which way I'm gonna go on that.
L691[21:08:54] <Izzy> Aight, will do some time this morning
L692[21:10:25] <Amanda> %choose more space or veg out to YouTube
L693[21:10:25] <MichiBot> Ama​nda: Somebody once told me to roll with "veg out to YouTube"
L694[21:10:33] <Izzy> Realised the OC:R branch I backported it to wasn't up to date with the 1.12 version in MP/OC we're running on the server which is not ideal
L695[21:10:37] <Forec​aster> Go both ways, at the same time
L696[22:03:54] ⇦ Quits: Vexatos (~Vexatos@2001:4647:af75:0:a19f:4608:f47:914) (Quit: Insert quantum chemistry joke here)
L697[22:27:57] ⇦ Quits: Corded (~MichiBot@2604:2dc0:200:54::2) (Quit: Bridge shutting down)
L698[22:28:05] *** Joins: Corded (~MichiBot@2604:2dc0:200:54::2) Michibot
L699[22:28:05] zsh sets mode: +v on Corded
L700[22:31:00] ⇦ Quits: Corded (~MichiBot@2604:2dc0:200:54::2) (Client Quit)
L701[22:31:06] *** Joins: Corded (~MichiBot@2604:2dc0:200:54::2) Michibot
L702[22:31:06] zsh sets mode: +v on Corded
<<Prev Next>> Scroll to Top