Hide Idle (>14 d.) Chans


← 2022-01-27 | 2022-01-29 →
signpost[asciilifeform]: anyone recall what'd be the most recent source of "vtools" patches?
signpost[asciilifeform]: http://paste.deedbot.org/?id=oQlW << I've got just about all the vpatches cut, except for the few patches there that need to be converted and vtools.
shinohai[asciilifeform|billymg] sadly doesn't remember the link to diana_coman's v starter kit ....
signpost[asciilifeform|billymg]: hm I thought it was bvt
signpost[asciilifeform]: in fact, mostly phf, looks like
signpost[billymg|asciilifeform] has an aggressive gc between ears.
signpost[asciilifeform]: anyway phf's site works for me
signpost[asciilifeform]: whaack: any idea how this diverges from the one I linked?
signpost[asciilifeform|billymg]: ah, this is the perl v
signpost[asciilifeform|billymg]: I'm gonna stick with the shell one, and folks are welcome to then build whichever alternative they like.
signpost[billymg] intends to lop off the perl dependency for building gcc if possible.
PeterL[asciilifeform]: reduction of dependencies is a noble goal
signpost[asciilifeform]: yeah, going for minimizing that, and maximizing operator choice.
signpost[asciilifeform|billymg]: I don't need to tell you how to run your life; we just need a basis for sharing the traditions of our own houses.
signpost[asciilifeform]: "we make bread thus" and etc, civilization atop that.
bitbot: Logged on 2022-01-28 15:46:00 signpost[asciilifeform]: http://paste.deedbot.org/?id=oQlW << I've got just about all the vpatches cut, except for the few patches there that need to be converted and vtools.
signpost[asciilifeform]: hesitant to pull python into base set either
asciilifeform: makes sense
signpost[asciilifeform|billymg]: though python pbuild's welcome later, for blatta
PeterL[billymg|asciilifeform]: with no python or perl, are you just doing everything in Bash?
asciilifeform historically sat on gentoos and assumed the presence of 1 python or anuther. but it's a hog indeed
asciilifeform after misadventures in attempting serious proggies in bash rather pessimistic re 'replace perlisms w/ bash'
signpost[asciilifeform|billymg]: PeterL: yeah, pretty simple script is turning the crank here. if more complicated tools are needed, would like them to be ada.
signpost[asciilifeform|billymg]: asciilifeform: yep, shell scripts aren't any place to do serious programming. I've been laboriously grinding the script smaller, weighs about 400 loc atm
signpost[asciilifeform]: but gotta not haul in the whole world as a dependency to get started.
signpost[asciilifeform|billymg] will probably cut that ~400 into separate items too. pretty clear perforations in there.
asciilifeform: this is one reason for asciilifeform's interest in small schemes (e.g. 'tinyscheme') -- as potential replacements for perlism/bashism for bringup
signpost[asciilifeform|billymg]: http://paste.deedbot.org/?id=psvG << wip item in question. squashfs/overlayfs-related stuff likely to be carved off into a separate script, as they're only relevant to embedded hardware dev or the distributed job-runner system I've discussed.
signpost[asciilifeform]: asciilifeform: yep, would gladly port the whole thing to a saner language in time.
asciilifeform has a # of boxen w/ squashfsisms, but indeed they're 'toasters'
asciilifeform: btw fwiw 'litmus' worx, in principle if you have a gnat, dun need ye olde gpg for vtronics (for read-only vtronics, that is)
asciilifeform: as of the most recent vpatch, is 100% edible afaik
signpost[asciilifeform]: defo intend to build ffa on this.
signpost[asciilifeform|billymg]: and to rip out heathen items and replace with allied items as can be done.
asciilifeform in fact has coad for several unpublished ch's worth of ffa written, debugged, etc. and collecting dust, sadly lacking time for aeons nao to touch it
asciilifeform: spent what little spare cycles had in recent times on pestronics.
signpost[asciilifeform|billymg] has been grunting in his shitwaders in pentacle for what, 2yrs now, and fountaintron about same.
signpost[asciilifeform]: good things in time or not at all.
signpost[asciilifeform] appreciates the hell out of the pest spec, eager to turn more attention to it, just loathe to put down items already loaded into my skull just yet.
asciilifeform hopes to live to finish not only ffa but e.g this
asciilifeform: signpost: ty. imho is important to carry on the thing that's loaded into head, otherwise gotta reload, and lotsa wasted sweat.
awt[billymg|asciilifeform]: asciilifeform: I see no reference to checking the version of packets in the common prologue. Is it intentionally left out?
asciilifeform: awt: thus far
asciilifeform: awt: in 0xfa, supposing we go with some or all of the proposed format-breaking changes, will need to check ver
awt[billymg]: In blatta it's step 5.
asciilifeform: current draft, quite reworked from 0xfb but not yet including any of the things discussed this wk
awt[asciilifeform]: asciilifeform: also I definitely like step 5. Removes some duplication in my working version.
asciilifeform: which 5 ?
asciilifeform: ah that's 0xfb
awt[asciilifeform]: billymg: thanks for the key mapping snippet, btw - haven't had a chance to try it yet.
awt[billymg]: asciilifeform: regarding the order buffer, what if two or more messages have the same timestamp?
asciilifeform: awt: ordering is by chain
asciilifeform: possibly awt is asking re the 'The Order Buffer may contain packets bearing duplicate messages, and is indexed by Time of message receipt.' text
bitbot: Logged on 2022-01-28 17:30:47 asciilifeform: possibly awt is asking re the 'The Order Buffer may contain packets bearing duplicate messages, and is indexed by Time of message receipt.' text
asciilifeform: that subsection does need massage
asciilifeform welcomes suggestions
asciilifeform: the order buffer is req'd for dealing w/ out-of-order msg arrivals and losses (i.e. missing msgs which create chain gaps)
asciilifeform ambivalent re: whether a whole buffer is even needed, it uncomfortably resembles the old pre-trb trb's 'orphanage'
awt[asciilifeform]: asciilifeform: perhaps could index by timestamp -> message_hash.
asciilifeform: (in original, implied that can have 1:1 index by timestamp, which is obv. not the case, as awt pointed out)
awt[asciilifeform|billymg]: yeah - got burned by that already
awt[billymg|asciilifeform]: meh can just store the messages in an arbitrarily ordered array by timestamp.
awt[billymg|asciilifeform]: lol right alk key doesn't even report a key code using xev
awt[billymg|asciilifeform]: asciilifeform: reading the order buffer section and step 10 of the common prologue it appears that a packet may loop in and out of the order buffer indefinitely if the packet requested via GETDATA never arrives.
asciilifeform: awt: this is entirely good point
asciilifeform: awt: what prolly oughta happen to packets which drop outta that buffer w/out getting reconnected chainwise is /dev/null
asciilifeform: (rather than reentry into order buffer, which, lol, yes infinite carousel)
u0_a185[asciilifeform|billymg]: The 24-Hour VWAP for BTC is $ 37858.38 USD
u0_a185[asciilifeform]: The 24-Hour VWAP for BTC is $ 37858.38 USD
u0_a185[asciilifeform|billymg]: The 24-Hour VWAP for BTC is $ 37858.38 USD
u0_a185[asciilifeform]: The 24-Hour VWAP for BTC is $ 37858.38 USD
billymg: awt: np
billymg: oh hey, more pest bots
whaack[billymg]: hopefully i'll get trbexplorer in here soon as well
← 2022-01-27 | 2022-01-29 →