log☇︎
82000+ entries in 0.652s
asciilifeform: i do not point finger at trinque , this is here for the record strictly.
asciilifeform: e.g., i still have not forgiven, and do not intend to forgive, the acpi 'soft' power supply switch. ☟︎
asciilifeform: what i ~do~ have a strongly developed allergy to is attempts to redefine minimal expectations using 'new realities'
asciilifeform: understand, i have nothing against proggies that are slow because they are overly ambitious, common lisp on 386, 'quake' on 486-dx2 (i played, 5 frame/sec, yes), etc.
trinque: I will rape the thing with a huge benchmark and speak of it further at that time.
asciilifeform: neato. i will try this clim, at some point.
trinque: to restate, and then I'm bored of this, getting no perceptable lag whatsoever between firing events to the event queue and seeing the screen redraw.
asciilifeform: trinque: i was answering 'that's just what you read'
asciilifeform: aaaaha, lolk, it was what i read, what was actually there was proof of fermat's conjecture...
asciilifeform: the implicit claim was that i am somehow 'unreasonably' expecting realtime interaction with comp.
phf: asciilifeform: i don't actually know about that, i gotta try it myself now :) i think we might've both tried it around the same time
asciilifeform: and i will use the old reality.
asciilifeform: (phf suggested earlier that it is not yet fast enough. so i put off trying.)
asciilifeform: trinque: i have nfi, will have to try the new clim, won't i.
asciilifeform: because i USED realtime comp.
mircea_popescu: phf in my head, if the algo is correct but the problem too hard, then the problem is just too hard ; whereas if the algo is wrong for the problem, or badly implemented, then there's that. i suppose you could say "linear sort is dog slow", though.
asciilifeform: but i also do not know if he has any experience with fast comp.
mircea_popescu: i dun think that's the common meaning ; but if that's what you use, fair enuff.
phf: so depending on what sort of algorithms are used (i.e. knowing how mcclim works) a "get a faster machine" might be a valid response to "mcclim is slow", or it can be invalid, if, say, there's combinatorial exposition in the component interaction, or if there's a polynomial behavior in large text processing (those are two usual suspects)
asciilifeform: the counter to 'i don't feel any delays' is to sit the 'not feeler' in front of wordperfect on msdos and watch his face as he says 'this dun feel any faster'
mircea_popescu: as far as anyone knows, dog slow kinda means exactly that "irritatingly slow to me", to which "i don't feel any delays" is a valid counter.
mircea_popescu: i suspect this list may not be entirely correct. looks autogenerated.
asciilifeform: i have deeply nfi, but do know that winblows is pestilentially universal in ru, in all walks of life
mircea_popescu: i dunno about that.
asciilifeform: i'm still waiting to see them lose something expensive, e.g., the intel/amd microcode booby.
asciilifeform: i - also lolllered.
mircea_popescu: "CIA is cyber B-Team, yes? Where is cyber A-Team?" << i lollered. where, they're gone, locked in alf's basement.
mircea_popescu: trinque wasn't waht i meant. just, your spring sprung too far s'all
trinque: mircea_popescu: I can't develop on that imagined computer any more than Eulora wrote its own OS.
a111: Logged on 2016-10-31 17:56 trinque: I do not make decisions in life about what to do today based on somebody else's imagined tomorrow.
BingoBoingo: Step 1: I am powerless over lulz and my lyfe has become unmanageable!
phf: i found an old screenshot from last time i was hacking on clim :D http://glyf.org/screenshots/clim.png
asciilifeform: and what happens when i wanna plant the thing on bare vga later ?
ben_vulpes: i'm less up in arms about the vpatch than i am about continuing to lend credence to imperial operations like github et al. "dear mcclim donor".
trinque: I'm irritated he didn't genesis v-patch the thing ☟︎
phf: well, mcclim layers things in a way that i thought was part of the spec, but turns out it was beach's decision (to among other things support multiple different backends). when i was trying to optimize the x11 backend last time, i was hitting multiple "independent" layers, that all demanded attention. there wasn't really an 80/20 solution, which made me conclude that there's not a single subtrate that you can optimize (the way, say,
trinque: I guess it's not clear that I'm proposing *eating* mcclim, hence what I said to g_l
asciilifeform: only that i barfed mightily the last time i tried to clim.
asciilifeform: i have nfi if it were inherent - or not.
trinque: your bare assertion that there's some kind of inherent lag in the redisplay loop does not meet my observations, i.e. right now on my desk
asciilifeform: when i ran chirality numbers for al schwartz, it took ~6 mo of cluster. and that's small change.
asciilifeform: trinque can use whatever he likes on his box, but if i run a proggy and have to count '1, 2, 3,...' after pressing a key and before seeing screen draw, it goes in the shit bin.
trinque: but I will do some benchmarking, should be interesting
phf: anyway, i want to try and deploy masamune, but i'm also trying to spin up clim2 to see if it's maybe less of a dog (i imagine it would be since it was probably battle optimized for all that ported-from-legacy contractor software)
asciilifeform: see, i am willing to buy 10k box, or whatnot, to get fast clim. but i am not willing to close eyes and call a turd sausage.
phf: trinque: i don't think anyone's telling you what to do, but to honestly consider what you can observe, and compare it to what we can observe. asciilifeform is saying that clim is slow for him, i'm saying that there's something in clim design that makes it tricky to make it fast. it's a converstion.
trinque: the fuck, I'm supposed to develop for DOS now?
asciilifeform: trinque: i asked that you compare to a very factual 'yesterday', rather than imagined futures
trinque: if that proves impossible I'll move just like I did from commonqt
trinque: if clim gets too slow for any of my use cases I'll find out where and work on it
trinque: I do not make decisions in life about what to do today based on somebody else's imagined tomorrow. ☟︎
phf: asciilifeform: i can feel delays in x11 emacs after working on a hardware terminal, which makes me conclude that "feels" kind of changes once you get used to it. i can't use popular hipster software for example, because it's definitely slower
asciilifeform: phf: i dun see why 'inherent', if i gotta buy a hardware accelerator, i'll buy
trinque: I can tell you precisely how long redisplay-frame-pane or w/e takes depending on what's updated
asciilifeform: if i can feel delays, it is slow.
asciilifeform: 'dog slow' means that i can feel delays.
asciilifeform: trinque: in all fairness, i have nfi what you used before, maybe it was a telegraphist drawing on chalkboard. so yes, faster.
trinque: this is much better than the Qt I used for this same application before
asciilifeform: i'm comparing it to motherfucking 'word perfect' on msdos 5.
trinque: release the system; I'll use
trinque: I'm comparing it to other tk
asciilifeform: would be interesting demo, i'll watch.
trinque: I know the latency precisely
asciilifeform: ooh i almost forgot, on the cray i'll need a $10k payware lisp, yes
trinque: but like I said, want to lean on it to find out, not decided by any means
phf: i'm trying to spin up at least one instance of lispworks/allegro clim to see if it might be faster, but so far it's proving... difficult.
asciilifeform: (i'm still waiting for somebody to tell me which cray i gotta buy for it to work in MOTHERRFUCKING REAL TIME)
phf: trinque: i'd like to at least try it out. i'm unconvinced clim is a good idea, because there aren't any good implementations. mcclim is terribly over-engineered (in the best of java style, with delegates for proxies etc.), clim codebase that lispworks/allegro share is less so, but more hacky. which makes me wonder if clim spec itself is suspect ☟︎☟︎
asciilifeform: i always wondered if the very sad infrastructure in washington, a kind of frozen-in-time 1970s theme park, is at least partly on account of the expectation that anthill will be leveled 'any day' ☟︎
asciilifeform: general ripper: 'take this thing to the garage and hotwire it' grunt: 'i dunno how to make new detonator, the old one melted when i popped the cover with 3 wrong passwords'
asciilifeform: and i picture that it is largely against meatheads from... own side
asciilifeform: but i suspect that the acid will win in mircea_popescu's case. nuke builder does not want to make the booby overly sensitive, killing himself
mircea_popescu: you can't predict what i'll do, and you can't ensure against all the things i might do.
mircea_popescu: yes, but reality is more encompassing than engineering. so suppose i submerge the chest in fumaric acid and you lose.
mircea_popescu: also i have nfi how they imagine reality works, but in principle it's true that "the random process of nuclear decay is the gold standard of rngs" FOR THE REASON that you can't use arithmetics to distinguish one clump of radioactive substance from "another" clump.
mircea_popescu: i have no fucking idea how they imagine physical security is to work in the hands of an adversary.
mircea_popescu: i dunno what the bird was even, some kind of crow.
mircea_popescu: i watched yest the greatest thing ever. people took dog out to park, nice flat coated retriever. young, fulla life.
trinque: I watched it last week, treated it like a dog, and it was fine
mircea_popescu: i can tell by some pixels.
trinque: reminds my of my parents idiot dog that nervously stares at you as if to say "I am good boy???"
asciilifeform: (i imagine that a pyro makes the same face into a fire, a zoophiliac - while pounding into the sow; etc)
asciilifeform: ah i just assumed it was the sop grin of narcissist staring into camera lens
trinque: he's talking about the cuckgrin and weird goatee, I'm sure
phf: that buttery-smooth-emacs was a torture to read. it's like a text equivalent of those "heeeeeeeyyy guys it's me i'm back with yet another!! hello!!" youtube videos
mircea_popescu: come to think about it i'm getting fucking ridiculous over here, "oh, mp is against drugs and pot and doesn't watch porn and totally straightedge".
mircea_popescu: i suppose i should actually remedy this huh.
mircea_popescu: in which interval i doubt i watched a full hour of pronz.
mircea_popescu: well, i guess the livestream of my livingroom irl should count and that sorta thing. but i've not been following teh industry in just about a decade by now.
mircea_popescu: also should prolly be said i dun actually watch any porn.
asciilifeform: for some reason i thought that mircea_popescu cremated all of these when he moved out of ro
mircea_popescu: aha. i have tons of these things, not touched for >10+ years etc
asciilifeform: l0l, i recall the thread, it was on an ancient mac wasn't it
mircea_popescu: i think i may have a winamp somewhere also
asciilifeform: well yes, i also.
mircea_popescu: asciilifeform i have liek, old laptops etc.
asciilifeform: BingoBoingo: xchat (last i tried... has been a while) still built without dbus
asciilifeform: which is what i watched with on t40.
asciilifeform: when i learned that mircea_popescu watches his pr0n on an ibm t40, i could not picture it as with anything but mplayer
mircea_popescu: yeh but i dun have it so... ima have to look into it!