Thursday, 2023-04-06

sadoon[m]You know when they say if you have a problem in "programming" you need to take a walk or touch grass to solve it03:56
sadoon[m]I needed a 3 day break and I finally found out what's wrong with my mini-buildd setup lmao03:56
sadoon[m]Short version is you can't feed it .dsc files without the source packages being in the same directory, which I should have known already03:57
programmerjakewell, oops03:57
sadoon[m]I'm trying to rely less on my internet facing server by having all the source packages local to the VM running mini-buildd03:57
sadoon[m]So, time to see how much space that would take up03:57
sadoon[m]In other words to make the whole thing self contained and perhaps not even needing internet access03:58
programmerjakethat's nice...definitely not needing internet helps with reproducibility03:59
sadoon[m]I already cloned the repos so all I need is to tarball it and send to the VM04:00
*** ghostmansd[m] <ghostmansd[m]!> has quit IRC09:19
lkclprogrammerjake, once you actually submit, it says "you are not endorsed to submit in the category cs.HA, please obtain endorsements from other arxiv users"09:19
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has joined #libre-soc09:20
lkclyes sorry i didn't want to hit the cancel button as it tends to send out dozens of messages (on future calendar invites)09:20
lkclha, irony. everyone who read the first line would not have been there to tell everyone who missed it... whoops09:25
programmerjakeyes, that's expected per arxiv's policy09:25
programmerjakefor the meeting, you could rename just that one event (but not following events) to include the word "canceled" in the title, or something09:28
programmerjakewe might be able to ask the people who are working on ec25519, icr their names rn09:44
lkclprogrammerjake, oh! does [[!include]] actually work?? (in pandoc and ikiwiki)?10:22
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has quit IRC11:50
lkclnext phase of the LD/ST-with-DD-FFirst is there (
lkclgaah so many steps13:28
lkclokaay ready to start on ISACaller LD/ST-DD-Ffirst.  need tea :)13:43
*** jn <jn!~quassel@user/jn/x-3390946> has quit IRC14:39
*** jn <jn!~quassel@2a02:908:1066:b7c0:20d:b9ff:fe49:15fc> has joined #libre-soc14:40
*** jn <jn!~quassel@user/jn/x-3390946> has joined #libre-soc14:40
Ryuno-KiAndrJaenI don't understand much of it, but this might be an opportunity to raise awareness for this project:
programmerjake> oh! does [[!include]] actually work?? (in pandoc and ikiwiki)?19:14
programmerjakeyes, because I pass it through the inlining script before handing it off to pandoc19:14
programmerjakeit's natively supported by ikiwiki19:15
lkclprogrammerjake, ahh excellent.19:41
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has joined #libre-soc20:01
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has quit IRC20:10
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has joined #libre-soc20:11
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has quit IRC20:15
*** ghostmansd[m] <ghostmansd[m]!> has joined #libre-soc20:16
lkclah ha!23:42
lkcljust got word that this is new23:42
lkclRyuno-KiAndrJaen, the bit that kicks everything else into touch is that they did a completely uniform 32-bit ISA which then allowed them to spam 100% throughput on multi-issue execution23:43
lkclIntel x86 with its variable-length encoding is so embarrassingly complex: it could take *3 clock cycles* just to identify where the length starts from!23:43
lkclthey actually have to begin decoders at *every byte*, then cancel the ones that quotes didn't work out quotes23:44
* lkcl face-palm23:44

Generated by 2.17.1 by Marius Gedminas - find it at!