Show Idle (>14 d.) Chans

← 2020-09-06 | 2020-09-08 →
mats: it really is, havent been able to buy ammo since late march
asciilifeform: !w poll
watchglass: Polling 12 nodes...
watchglass: : Alive: (0.021s) V=99999 (/ Jumpers=0x1 (TRB-Compat.) Blocks=647152
watchglass: : ( Alive: (0.083s) V=70001 (/ Jumpers=0x1 (TRB-Compat.) Blocks=647152
watchglass: : ( Alive: (0.120s) V=99999 (/ Jumpers=0x1 (TRB-Compat.) Blocks=647152
watchglass: : Alive: (0.119s) V=99999 (/ Jumpers=0x1 (TRB-Compat.) Blocks=647152 (Operator: asciilifeform)
watchglass: : Alive: (0.144s) V=70001 (/ Jumpers=0x1 (TRB-Compat.) Blocks=647152
watchglass: : Alive: (0.157s) V=99999 (/ Jumpers=0x1 (TRB-Compat.) Blocks=647152
watchglass: : Alive: (0.241s) V=70001 (/ Jumpers=0x1 (TRB-Compat.) Blocks=647152
watchglass: : Alive: (0.329s) V=99999 (/ Jumpers=0x1 (TRB-Compat.) Blocks=647152
watchglass: : ( Alive: (0.309s) V=99999 (/ Jumpers=0x1 (TRB-Compat.) Blocks=647152
watchglass: : Busy? (No answer in 20 sec.) (Operator: asciilifeform)
watchglass: : Busy? (No answer in 20 sec.) (Operator: jurov)
watchglass: : Busy? (No answer in 20 sec.)
asciilifeform: !w probe
watchglass: : ( Alive: (0.763s) V=99999 (/ Jumpers=0x1 (TRB-Compat.) Blocks=647152
asciilifeform: !w probe
watchglass: : Busy? (No answer in 20 sec.)
billymg: testing znc on the new rockchip... took a minute to figure out all the znc setup/config stuff again (seems to be working though)
asciilifeform: wb billymg . dun hesitate to ask q if you encounter puzzler re the rk/softs
asciilifeform uses znc ; but disrecommends the www config feature
asciilifeform: imho rk is perfect for znc, and even low-to-moderate traffic www
asciilifeform: for that matter, asciilifeform's logger sat on a rk from nov. '19 to feb. '20; and w/ very little palpable lag, except in searches
asciilifeform: ( today, lives on 'dulap'-style 32cpu box w/ raid10 )
trinque: asciilifeform: btw your dulap image is glibc
trinque: I have not had much luck bootstrapping from a glibc system
asciilifeform: trinque: of course it is. it's a trad gentoo
asciilifeform: i thought this was obv.
asciilifeform only ever had musltronic 'miniature' linuxen, e.g. the 1 on 'pogo', the mips linux for 'M', and coupla similars
trinque: I didn't say this was bad; I said it didn't work for a bootstrap host.
asciilifeform: hmm i thought that trinque had a 'rotor'-style bootstrapper that goes from trad linux to musl
trinque: I've been living on musl systems for a half-decade by now
trinque: I started from one of those
asciilifeform: aa neato
asciilifeform: trinque: am i right then to suppose you have a musl-compat. emacs ?
asciilifeform: iirc that was the 1 tough nut in muslism
trinque: << yep, iirc I used the method described here
trinque: the problem is that my old muslwad isn't formalized at all, just carried forward in tarballs
trinque: I'm going to just upload the genesis here in a sec, and folks can see what I mean.
asciilifeform: this is still nifty tho, asciilifeform genuinely had nfi how to bring up emacs on musltronic
trinque: making sure it builds; last night a novel error surfaced regarding libtool
asciilifeform: afaik that was the 1 proggy that wouldn't stand up by default
asciilifeform will read
asciilifeform: trinque: seems like if trinquetronic linux requires a working musl-based item to bootstrap, might make sense to distribute a dulap-style canned image to do so from
asciilifeform: this is actually 1 of the uses i had planned for 'M' . tho later realized that 'M'-powered build of anyffin nontrivial would prolly take aeons.
asciilifeform: ( on 3GHz opteron, M-linux runs at roughly speed of 486DX2... )
asciilifeform: ( the 1 upside was to be that 'M' can be run on literally any x64 linux (afaik) , and even ported to whatever future irons w/ minimal effort. )
trinque: yep, that's what I'm thinking. folks can either use the canned image or roll up their sleeves and bootstrap themselves.
asciilifeform would be satisfied w/ a bootable image that can be asked to repeatably build itself to given ext. disk
asciilifeform: $image dun need to carry support for exotic irons, would be quite enuff for it to know how to init x64 cpu, ram, and talk over rs232
asciilifeform: ( in contrast to e.g. asciilifeform's dulap & rk classical gentoo canned items, where tried to include support for most common devices actually present in asciilifeform's iron fleet )
trinque: asciilifeform and shinohai, find instructions to build the musltron here
trinque: I've included a "bootstrap" fs to make things easy, but the same instructions can be used on any musl linux with ave1's gnat handy.
trinque: I expect to revise this genesis if problems are found, but otherwise here's gnat and all the horrible shit necessary to build and modify it.
shinohai: ty trinque will review as soon as back @ desk
trinque: both ./install and ./install-set take a -j$NUMJOBS param, but I'd be careful about using that on the first build. I encountered a bug in texinfo just now using -j12
trinque: after the first build of stage-1 and stage-2, should be able to use -j to speed things up
trinque: this has been a truly miserable quest. most of the gnushits require themselves to be present in order to build, "of course"
trinque: but happy to have finally packaged the thing up.
trinque: the "build" dir contains what'd be gentoo's ebuilds, but simpler, and I intend to keep them simple
trinque: the ./install script is missing a bit yet, but mostly convenience items that can come as subsequent patches
trinque: to the extent work can be moved from stage-1 to stage-2, that'd lessen the dependency on w/e host system (assuming you don't use my bootstrap) and make it easier to bootstrap elsewhere.
trinque: that's a slog though, and I've had enough of it. I'll be focusing on moving forward from here, not building bridges to $RANDODEBIAN, but I wouldn't reject that if it results in a simpler item.
trinque: anyhow, if you reach the end, you'll have gnat 2016 and a functioning gprbuild
trinque: from here, oughta let as many people as are interested chew on it, and then it's time to graft on the vtools tree.
trinque: et cetera, ad infinitum.
asciilifeform: a+++ trinque ! will test tomorrow.
asciilifeform: pretty compact ( given as inside are gcc et al ) testbed, too -- < 300MB !
asciilifeform: gpg -a --verify genesis.vpatch.asc genesis.vpatch.gz
asciilifeform: gpg: Signature made Mon 07 Sep 2020 12:13:49 PM EDT using RSA key ID FAB953C4
asciilifeform: gpg: BAD signature from "Michael Trinque <>"
asciilifeform: ^ trinque might want to see whether you posted mismatched payload/sig pair
asciilifeform: 'bootstrap' verifies tho.
asciilifeform: or hm, is the sig for unpacked...
asciilifeform: gpg: Good signature...
asciilifeform: nao worx.
asciilifeform: ( logical : it's a vtronic sig. )
asciilifeform will return to item when awake enuff to actually think, lol
trinque: enjoy asciilifeform
← 2020-09-06 | 2020-09-08 →