L1[04:02:55] <Forec​aster> %tonk

L2[04:02:56] <MichiBot> Yeah! F​orecaster! You beat V​aur's previous record of 6h 19m 54s (By 5h 33m 18s)! I hope you're happy!

L3[04:02:56] <MichiBot> Forecaster's new record is 11h 53m 12s! Forecaster also gained 0.03892 (0.00556 x 7) tonk points for stealing the tonk. Position #2. Need 0.5315712 more points to pass S​piderEveryOS!

L4[06:22:53] <Corded> > <Ocawes​ome101> I'm contemplating implementing a HTTP client / server for u…

L5[06:22:54] <Kristo​bear38> Gopher would be more usable in OC. @lunar_sam had a client

L6[06:26:44] <Izzy> ^

L7[06:26:52] <Izzy> fget is gopher lite, if such a thing is possible, also

L8[06:30:15] <Corded> > <Kristo​bear38> Gopher would be more usable in OC. @lunar_sam had a client

L9[06:30:15] <luna​r_sam> hi yes

L10[06:30:25] <luna​r_sam> i think i was working on a minitel implementation of gopher

L11[06:30:28] <luna​r_sam> but i think it had issues

L18[11:28:20] <Amanda> %splash Forecaster with mutable gold potion

L19[11:28:20] <MichiBot> You fling a mutable gold potion (New!) that splashes onto Forecaster. Forecaster turns into a dragon until hell has a mild winter.

L22[11:28:47] <MichiBot> Waiter!! Waiter!! There's a Magic furai furai furai! (25%) in my soup!

L23[12:05:27] <Amanda> %choose games or just halucinations

L24[12:05:27] <MichiBot> Oh no, not "just halucinations" again! I'll have "games" instead.

L25[15:29:16] *** Joins: RedstoneParkour (~RedstoneP@109.36.138.222) RedstoneParkour

L27[15:47:23] <Amanda> %remindme 1m check the thing

L28[15:47:23] <MichiBot> I'll tell you "check the thing" in 1m.

L29[15:48:23] <MichiBot> Amanda REMINDER: check the thing

L30[17:04:58] <Brisingr​ Aerowing> %splash @Forecaster

L31[17:04:59] <MichiBot> You fling a smooth apple potion (New!) that splashes onto @Forecaster. @Forecaster gains a negligible amount of luck.

L33[17:55:34] <Forec​aster> %tonkout

L34[17:55:34] <MichiBot> Yow! F​orecaster! You beat your own previous record of 11h 53m 12s (By 1h 59m 25s)! I hope you're happy!

L35[17:55:34] <MichiBot> F​orecaster has tonked out! Tonk has been reset! They gained 0.013 tonk points! plus 0.024 bonus points for consecutive hours! Current score: 2.9351494, Position #2 Need 0.4945712 more points to pass S​piderEveryOS!

L36[18:25:37] <Ocawes​ome101> https://discordembeds.pc-logix.com/live/1551660686449447055/1790015136481-f658360bc1bb.png

L37[18:32:35] <Forec​aster> What about artificial flavouring?

L38[18:38:58] <Ocawes​ome101> none of those either, that's a different motd

L39[18:46:07] <Ocawes​ome101> https://discordembeds.pc-logix.com/live/1551665848295620720/1790016366991-465e8104fe82.png

L40[19:26:30] <Forec​aster> But there's gotta be preservatives right?

L41[19:49:43] <Amanda> %remindme 30m check ci

L42[19:49:43] <MichiBot> I'll tell you "check ci" in 30m.

L43[20:19:43] <MichiBot> Amanda REMINDER: check ci

L44[22:00:23] <Brisingr​ Aerowing> %splash @Forecaster with mutable potion

L45[22:00:23] <MichiBot> You fling a mutable jumbonium potion (New!) that splashes onto @Forecaster. @Forecaster turns into a red iron boy until they p̸̮̪͗͝r̶̛̺͕ä̸̡y̴̢̮̚ ̵͎̉ẗ̴̹̜̊o̷̭͕͝ ̵̰̟̾t̶̩̿̎h̸̗̝̾͠ĕ̵̝ ̶̺̓͛ṿ̸̓o̵̗͔̎ǐ̵̱d̷͈̤̅.

L46[22:45:18] <Amanda> %choose rainbox time?

L47[22:45:18] <MichiBot> Yes.

L48[22:51:26] <Izzy> could one write a scheduler designed to exploit component call limits by batching and re-ordering component I/O?

L49[22:51:52] <Izzy> like, you get a lot of disk I/O per tick but only one network message or redstone change

L50[22:52:55] <Izzy> you could queue them separately by like, class, and then do max-1 disk I/O calls and then your network I/O

L51[23:11:00] <Ocawes​ome101> i've thought about doing that and it seems mostly like a lot of extra work for relatively little gain

L52[23:11:37] <Izzy> it seems interesting but also a lot of work yeah

L53[23:14:47] <Ocawes​ome101> the other thing is a lot of code assumes component calls are synchronous and that would sort of require they be asynchronous unless you have a lot of processes going on one machine

L54[23:17:21] <Izzy> My thinking would be they'd queue until you hit the max you can get away with and then the call would just yield

L55[23:18:10] <Izzy> and you keep doing that until either you reach the end of the process list or you fill up all your queues

L56[23:18:18] <Ocawes​ome101> oh i gotcha

L57[23:18:26] <Ocawes​ome101> so this wouldn't have a lot of impact for small systems

L58[23:18:34] <Izzy> Then you flush the queues and continue

L59[23:18:42] <Ocawes​ome101> that makes more sense

L60[23:18:51] <Izzy> yeah it'd make the most difference under high I/O load

L61[23:19:44] <Izzy> it would be Interesting but primarily helpful on either a graphical system or a multiuser system

L62[23:23:45] <Izzy> graphics to keep networking working under load, and multiuser for just squeezing as much I/O out of the system as possible

L63[23:24:07] <Ocawes​ome101> those are both good applications for this tbh

L64[23:24:15] <Ocawes​ome101> would be interesting to try out