Show Idle (>14 d.) Chans
← 2023-01-01 | 2023-01-04 → ↓
crtdaydreams[asciilifeform]: asciilifeform, phf: first time compiling cadr usim.tar.gz here on x86_64 void linux kver=5.18.19_1, I'm wondering what issues you've had with compilation (besides the misname -Wall flag) it appears a multilib system is a necessity? 32bit SDL seems to be a dep, are there any ways around this or patches?
crtdaydreams[asciilifeform]: yep musl, not gonna fly :<
phf: http://logs.nosuchlabs.com/log/pest/2023-01-02#1020120 << i haven't compiled brad's code in forever, the bulk of CADR work is happening here https://tumbleweed.nu/lm-3/
dulapbot: Logged on 2023-01-02 02:54:57 crtdaydreams[asciilifeform]: asciilifeform, phf: first time compiling cadr usim.tar.gz here on x86_64 void linux kver=5.18.19_1, I'm wondering what issues you've had with compilation (besides the misname -Wall flag) it appears a multilib system is a necessity? 32bit SDL seems to be a dep, are there any ways around this or patches?
phf: and usim lives here, https://tumbleweed.nu/r/usim. it's fossil repo
phf: i think brad's version had both sdl and x11, but we removed sdl, because nobody who works on this (3 people, myself included) care. ithere's a helper script in https://tumbleweed.nu/r/l/doc/trunk/README.md to bring up the whole thing
phf: but there should be no reason for the tumbleweed code not to compile on usim, except for some gnuism here and there
asciilifeform[cgra]: hey apparently even thinking folx often come to believe they can 'model' pyramid. so prolly will never run out.
dulapbot: (asciilifeform) 2022-06-14 phf: which is that you're trying to prove an argument, rather than make an honest discovery into the subject. you have no skin in game beyond the substance of the argument. if you for example, and by accident or intentionally, had some money running on a pyramid, might be a lot more interested in attempting to present viable models. even if "wake up from coma discovered that all money in mmm" if you were then to pull out, would have +ev, yet
asciilifeform: billymg: apropos of nuffin, seems like one can't easily link to an /action in yer logotron
asciilifeform[cgra]: ( the '*' aint clickable )
billymg[asciilifeform]: asciilifeform: the handle is clickable
asciilifeform: oh nm indeed
asciilifeform[cgra]: the column shift confusing tho imho
billymg[cgra]: imo should be the timestamps that are clickable
billymg[cgra]: now that those are displayed in their own column
billymg[asciilifeform]: and is traditionally how it's done elsewhere on the web
asciilifeform[cgra]: oldthrd asciilifeform was to link, re prev thrd
shinohai[cgra]: Oh heya cgra!
asciilifeform: hmm oddball echo ^ ? ( possib. on acct of desynced loggers )
bitbot[asciilifeform|cgra]: (asciilifeform) 2022-01-19 asciilifeform: realizes that he has sumthing like a personal 'laws of robotics' for 'when to spend coin' (or rather, when not to) :
asciilifeform: welcome to pestnet, cgra !
asciilifeform: dunno how useful the display of timestamps, esp. given billymg's spiffy 'and then x hrs passed' pheature
cgra speaking with duct-taped tkinter-blatta, irc-specifics and order buffer pulled out and/or taped shut. hi all!
billymg[cgra|asciilifeform|cgra]: though would have to be both handle and timestamp, since timestamps get hidden below a certain viewport width
billymg[cgra|asciilifeform|cgra]: asciilifeform: which is why hidden on smaller viewports
billymg[cgra|asciilifeform|cgra]: but on larger viewports, also no need to have mile-wide lines of text, makes for shit readability
phf: 􏿽http://logs.nosuchlabs.com/log/pest/2023-01-02#1020128 << i'm a strong believer that an important property of `thinking folk` is to be able to think any kind of thought a thinking folk might want to, and separate it from the doing aspect. e.g. once can theorize on how it feels to b
dulapbot: Logged on 2023-01-02 12:21:53 asciilifeform[cgra]: hey apparently even thinking folx often come to believe they can 'model' pyramid. so prolly will never run out.
phf: 􏿽e fucked in the ass, without it necessarily leading to one going to a homo club the next day over. in fact thought policing of that manner is a quintessential non-thinking folk behavior. `what are you some kind of fag?`
phf: 􏿽likewise, one can attempt to theorize a best game theoretical strategy for whatever situation. professional poker players like Sklansky took a degenerate, `lost the estate, the horse and the wife` activity, and modeled highly conservative winning strategies out of it. this is also
phf: 􏿽what places like goldman do when it comes to any kind of money markets. but that doesn't mean that i think that ~i~ could play an mmm, and win, and that's why i don't do it, because it's not my area of expertise. never the less to suggest that there's no analyzable, optimal strateg
phf: 􏿽y given acceptable risk tradeoffs is the opposite side of the mmm-rube coin, the all wise doomer skeptic
asciilifeform: waitasec wat happened there? phf -- all good points, but the msg yer replying to is actually from june ?!
dulapbot: Logged on 2023-01-02 12:21:53 asciilifeform[cgra]: hey apparently even thinking folx often come to believe they can 'model' pyramid. so prolly will never run out.
dulapbot: Logged on 2022-06-16 13:57:25 asciilifeform: hey apparently even thinking folx often come to believe they can 'model' pyramid. so prolly will never run out.
asciilifeform: but replayed without replay marker ?!
asciilifeform: not showed up in asciilifeform's station today, either (tho visible on dulapnet logger)
phf: lol, i was trying to figure out in vain why you decided to lift an old thread
phf: there's all kinds of mess going on if you compare bitdash and nosuchlabs.
asciilifeform: pretty strange -- even if were quirk in cgra's station, the ancient msg oughta've been rejected as stale/dupe by erryone neh
asciilifeform: phf: nosuchlabs logger is pretty useless atm, given as doesn't mark replays nor strip []s
asciilifeform looks fwd to replacing it with an actual logger
asciilifeform: this lul wasn't a regular replay tho
asciilifeform: was clearly sumthingelse
asciilifeform: q is -- wat
awt[asciilifeform]: cgra: please publish a patch!
asciilifeform looked at own debug log, defo didn't emit a re-edition of that line today
asciilifeform: dollars to doughnuts it's a bug in getdata handler in blatta; in debug log, asciilifeform sees over9000 getdata responses to cgra
phf: well, as far as `why extra packets` could be that nosuchlabs doesn't have history of packets going that far? so when cgra getdata'd for packets, some other station provided him with those packets and nosuchlabs decided to fill its own holes?
asciilifeform: phf: also possible. but doesn't reveal why ancient msg got processed as fresh somewhere
phf: yeah, combined with some kind of bug in getdata (like maybe station when receiving getdata thinks it's a fresh packet, and broadcasts it
asciilifeform: (timestamps, recall, are part of the message)
phf: i remember awt saying, but i might be misquoting, that timestamps are not checked at all in blatta. i don't check timestamps either though
awt[asciilifeform]: timestamps are checked
phf: apologies
awt[asciilifeform]: In the past we've had issues due to clocks being out of sync
phf: yeah, but in this case by 6 months
awt[cgra]: yes, not proposing that's the cause here
cgra: awt: seems like it's still a mess, don't see stuff in correct order yet on my window, for one. plus whatever else funny going on underneath, judging by the current pestnet events
signpost[cgra|asciilifeform]: ah stochastic conversation starter feature being tested
awt[cgra]: dulapbot was likely catching up on cgra's selfchain, that's for sure.
cgra: awt: but i'll do, once in some kinda shape
awt[asciilifeform]: cgra awesome
awt[asciilifeform]: getdata responses are the only types of messages that can sneak through the timestamp check.
asciilifeform: awt: likely the boojum is in where the thing decides whether to rebroadcast a msg (i.e. erroneously rebroadcast a getdata response)
phf: signpost, a 2am irish pub brawl of a conversation starter!
signpost[cgra]: !!whiskey signpost
signpost[cgra]: the battle cry before the nut kick, yes.
signpost[cgra]: welcome cgra!
signpost[cgra]: peering looks active aside the chain sync chaos
awt[cgra]: asciilifeform: possibly - obviously tested syncs without that happening, however.
lobbes[cgra|asciilifeform]: welcome cgra! likewise, peering appears active from my station logs
cgra: lobbes: neat!
crtdaydreams[cgra]: http://logs.bitdash.io/pest/2023-01-02#1020059 << downloaded and ran http://0x0.st/oR9n.png but cursor interaction is completely borked, teleports and changes size any attempt to interact w/ RMB works but the teleporting bugs it out and closes the window
bitbot[cgra]: Logged on 2023-01-02 11:46:44 phf[deedbot|awt]: but there should be no reason for the tumbleweed code not to compile on usim, except for some gnuism here and there
crtdaydreams[cgra]: nvm, still usable, it only happens when moving cursor at speeds, but as soon as movement slows down below an arbitrary threshold no artifacting
crtdaydreams[asciilifeform|cgra]: lol ok, prolly direct result of running on a non-multilib system but it's well and truly broken
phf: http://logs.bitdash.io/pest/2023-01-02#1020118 << << that's perfectly valid result in base-8
bitbot[cgra|asciilifeform]: Logged on 2023-01-02 15:01:22 crtdaydreams[jonsykkel|awt|signpost]: lol ok, prolly direct result of running on a non-multilib system but it's well and truly broken
crtdaydreams[cgra]: phf: wait is everything in cadr octal?
phf: it's maclisp on cadr, it's got all kinds of idiosyncracies, which has to do with the fact that it's a system language used initially to implement the system itself. ibase and base are set to 8 by default, but can be changed as neededd, on e.g. per file basis
crtdaydreams[cgra] will excuse self to go read chineual
phf: like 25. with a dot is 25 base-10, 25 is whatever ibase, but you also have 2_11 which base/ibase 8 means (ash 2 11) => 2000
crtdaydreams[cgra|asciilifeform]: unlike in the manual though the card machines initial radix is 8
phf: you'll have to link to that, there's like 5 different editions of chinual, but i'm pretty sure none of them ibase or base set to 10
crtdaydreams[asciilifeform]: right, which is rather confusing because they're set to 10 but everythings in the octal system so it's 8 but in octal so "10"
crtdaydreams[cgra|asciilifeform]: I'm not quite sure what ASH does how and why
phf[asciilifeform]: well no, the reader and the printer are separate. how read interprets numbers is set by `ibase`, how print prints numbers is set by `base`
phf: ash is in chinual, it's a bit shift operation, https://tumbleweed.nu/r/lm-3/uv/chinual.html#index-ash not particularly useful to know upfront. just more of an illustration of the kind of concerns the system has: dealing with binary dadata
crtdaydreams[asciilifeform|cgra]: oh right, a(rithmetic sh(ift)
crtdaydreams[cgra]: i recall fondly the quake 3 fast inverse square root function
cgra: awt: any idea why blatta's logging.basicConfig() don't seem to have any effect on my desk? python 2.7.12. it doesn't seem to work in a manual python shell test either
crtdaydreams[cgra]: must bbl o/
awt[asciilifeform]: cgra: can you describe the behavior you're expecting?
cgra: awt: i'd expect the log level to change, and logging format to change, both according to the logging.basicConfig() input (of the one and only call). but it's stuck into defaults, level=ERROR for example
cgra: awt: just found out that even a call like logging.debug() prior to the logging.basicConfig() call will make any further basicConfig() calls no-op
cgra: that threw me off when trying a stand-alone test. but now it's sorted
← 2023-01-01 | 2023-01-04 → ↑
Random(pest) | Download hourly DB snapshot | Get Source Code