Show Idle (>14 d.) Chans


← 2020-08-30 | 2020-09-01 →
10:45 asciilifeform http://logs.nosuchlabs.com/log/asciilifeform/2020-08-30#1020827 << the unfortunate part is that in practice, quite often needed. e.g. 'SO_REUSEADDR' .
10:45 snsabot Logged on 2020-08-30 23:00:47 verisimilitude: I meant the library would cover general TCP, and ignore any fancy options, which probably aren't needed.
10:45 asciilifeform !w poll
10:45 watchglass Polling 12 nodes...
10:45 watchglass 205.134.172.26:8333 : Alive: (0.060s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=646153
10:45 watchglass 205.134.172.6:8333 : (172-6.core.ai.net) Alive: (0.133s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=646153
10:45 watchglass 205.134.172.4:8333 : (172-4.core.ai.net) Alive: (0.136s) V=70001 (/therealbitcoin.org:0.7.0.1/) Jumpers=0x1 (TRB-Compat.) Blocks=646153
10:45 watchglass 205.134.172.27:8333 : Alive: (0.164s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=646153 (Operator: asciilifeform)
10:45 watchglass 208.94.240.42:8333 : Alive: (0.136s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=646153
10:45 watchglass 192.151.158.26:8333 : Alive: (0.152s) V=70001 (/therealbitcoin.org:0.7.0.1/) Jumpers=0x1 (TRB-Compat.) Blocks=646153
10:45 watchglass 108.31.170.3:8333 : (pool-108-31-170-3.washdc.fios.verizon.net) Alive: (0.291s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=646153 (Operator: asciilifeform)
10:45 watchglass 213.109.238.156:8333 : Alive: (0.321s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=646153
10:45 watchglass 143.202.160.10:8333 : Alive: (0.264s) V=70001 (/therealbitcoin.org:0.7.0.1/) Jumpers=0x1 (TRB-Compat.) Blocks=646153
10:45 watchglass 103.36.92.112:8333 : (terebe.ns01.net) Alive: (0.571s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=646153
10:45 watchglass 188.121.168.69:8333 : (rev-188-121-168-69.radiolan.sk) Alive: (0.716s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=646154
10:46 watchglass 176.9.59.199:8333 : Busy? (No answer in 20 sec.) (Operator: jurov)
10:46 asciilifeform !w probe 205.134.172.28
10:46 watchglass 205.134.172.28:8333 : Alive: (0.155s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Return Addr=0.0.0.0:8333 Blocks=485329
10:51 * adlai was almost killed by one of these recently, although was fortunate enough to defuse actual violence by judicious application of silence >> http://logs.nosuchlabs.com/log/asciilifeform/2020-08-07#1018147
10:51 snsabot Logged on 2020-08-07 15:37:30 asciilifeform: i thought plebefodder delivery were up +1000%..
10:52 asciilifeform adlai: how's that ? are you a motorcyclist or wat
10:52 adlai I was just about to write a "too long; don't ask"
10:52 adlai I illegally cut the motorbike's path with my meat-powered bicycle.
10:52 asciilifeform adlai: btw are we even thinking of same thing ? over here, those trucks go at ~walking pace and stop erry 2 blox, like garbage truck
10:52 asciilifeform aa
10:53 asciilifeform adlai: what've you been up to anyway ?
10:53 asciilifeform i recall a 'going on long vacation'
10:53 adlai dude's subsequent challenge was along the lines of "you rat-shit crazy slumdog, never do that ever again!"
10:54 asciilifeform adlai: put a motor on yer donorcycle, outrun them instead of them -- you..
10:55 adlai mostly reading, in the hope of tossing the most pulpworthy books into the paper bins, and leaving the less garbage ones at a reasonably neutral location near a secondhand bookstore.
10:56 adlai I have also begun a concerted effort on writing the first of my own, although it is likely to take about as long to reach publishability as the long-tail end of cetacean pregnancies.
10:56 * adlai has almost no wish to outrun motorized vehicles, this is a wonderful way to "get places", without any idea what places exist other than 'a' and 'b'.
10:59 * adlai is also avoiding the use of keyboards for writing, primarily due to the chiropractic consideration, and resolve strengthened by the generalized distrust of the various pathetic excuses for 'computer'
11:00 adlai posts on the 'blog' might appear slightly more often than S.WOL statements did, back when they did, although not much.
11:02 adlai in other n00se, I managed to regain consciousness one day without my old secondhand ru arhfmethka text, and as a result, am reducing the usage of both ethanol, and backpacks.
11:03 adlai "secondhand" is a colossal understatement; the leaves seemed about a decade away from turning brittle beyond the point of physical use.
11:03 * adlai was reading this for the language practice, and only incidentally for algorithms
11:05 adlai local deliveryfolks with whom I have discussed their work have invariably claimed that the quarantine has improved their earnings, although I have never actually discussed this question with ones who were on their way anywhere, only with the ones that were taking time off.
11:13 adlai http://logs.nosuchlabs.com/log/asciilifeform/2020-08-08#1018360 << yet another reason why I am taking a hiatus from writing software, in general; specifically, lisp; most specifically, the trading program. it has overflown head, if the metaphor is parseable, to the point where I am offering [a small amount of] coin for the the work of distanglement: https://github.com/adlai/scalpl/issues/13
11:13 snsabot Logged on 2020-08-08 16:42:57 asciilifeform: the 1 thing rms did that there can be 0 forgiveness for, is his ~deliberate~ choice to cultivate coad gnarl as 'theft defense' mechanism for e.g. gcc.
11:14 adlai the tl;dr of issues/13 is "write sane toplevel comments, between my insane toplevel forms, and you can receive somewhere between zero and 4/5ths of a coin"
11:15 * adlai intends to leave the programmatic paper chase on hold, until this issue is resolved, even if he has to write every single comment himself.
11:29 trinque adlai: do I understand correctly that you're offering a bounty for someone else to document your shitball *for you*?!
11:32 adlai trinque: almost; the only departure from correctness, that I can discern in your question, is the difference between comments and documentation.
11:33 * adlai wonders whether the hamsters who run laps against themselves die of heart attacks faster than the humans who run marathons on an indoor treadmill.
11:42 adlai since this might be an enlightening simile, I'll append it to my answer to trinque's question: stage performers occasionally pay for the attendance of audience members who have certain characteristics known in advance, to facilitate 'hot reading' (as opposed to 'cold'), and various other acts of showmanship; I am essentially offering a bounty for someone who is able to read Common Lisp, and write
11:42 adlai English, with the caveat that they although they might not be the same person, they should work together.
11:44 adlai if all this bounty does is accelerate the expenditure of my own coins, then sic transit pecunia fullstop
11:46 verisimilitude I like the idea, considering I've an advanced grasp of both Common Lisp and English.
11:46 verisimilitude I'm interested, adlai.
~ 38 minutes ~
12:25 asciilifeform http://logs.nosuchlabs.com/log/asciilifeform/2020-08-31#1020868 << talk to shinohai , iirc he actually specializes in similar things
12:25 snsabot Logged on 2020-08-31 11:14:59 adlai: the tl;dr of issues/13 is "write sane toplevel comments, between my insane toplevel forms, and you can receive somewhere between zero and 4/5ths of a coin"
12:27 asciilifeform http://logs.nosuchlabs.com/log/asciilifeform/2020-08-31#1020877 << him too, fella likes writing detailed docs
12:27 snsabot Logged on 2020-08-31 11:46:57 verisimilitude: I'm interested, adlai.
12:27 * asciilifeform also likes, but sadly very very busy..
12:39 trinque adlai, what I'm asking is why you think someone else would write a better spec for your work than you
12:39 trinque but I can't be bothered to sift through any more acid-cross-wired metaphor.
12:43 trinque the point is that documenting one's work causes you to rethink your work in the context of everything that was unsaid in it. somebody else can't do that.
12:43 trinque it's not decorative, like "sprinkle the documentation onto it, so it's superficially complete"
~ 22 minutes ~
13:06 asciilifeform http://logs.nosuchlabs.com/log/asciilifeform/2020-08-31#1020885 << hrm i assumed that this is precisely what he was asking for. 'read this megalith and simplify' etc
13:06 snsabot Logged on 2020-08-31 12:43:25 trinque: the point is that documenting one's work causes you to rethink your work in the context of everything that was unsaid in it. somebody else can't do that.
13:08 * asciilifeform did quite a bit of similar work commercially
13:09 asciilifeform i can't fathom, tho, that anyone would pay for 'just put emacs-genned comments on subroutines'. tho adlai's req could be interpreted that way.
13:10 asciilifeform fwiw asciilifeform considered in past, and may consider again, to put btc bounties on finding mistakes in ffa.
13:12 asciilifeform however, the historic record re this kinda thing is disappointing imho.
~ 17 minutes ~
13:29 trinque yeah, trb comes to mind, not that the author asked, more forced it to happen via sloth.
13:29 trinque not a shining example of what to do.
13:30 asciilifeform trinque: trb from my pov is rather like 'crashed ufo'
13:30 trinque mhm
13:31 trinque I doubt anybody yet has a rigorous description of why that works (assuming for the moment that it does)
13:31 trinque for those following along, I'm not saying trb doesn't work, either.
13:32 asciilifeform trinque: plenty of folx have variously detailed descriptions, but the 1 troo test would be a 100%-working (over reasonably long term, and plentiful abuse) from-0 reimplementation. which absent.
13:36 trinque aha, have to prove it's not actually accidental behavior coming from e.g. openssl, bdb, etc
13:36 asciilifeform right
~ 15 minutes ~
13:52 asciilifeform trinque: at this pt i'm no longer convinced that the 'corners' of bdb, openssl, etc. actually matter in any sense other than 'reimplemented item must correctly verify historic blocks, and reject any hypothetical ones that violate known form' .
13:54 asciilifeform and in particular not convinced that it is necessary to replicate 100% of the rather asinine mechanics of the orig. mempool.
13:55 asciilifeform ( what with the node e.g. attempting to determine which tx is 'original' and which 'double' in mempool, a perennial source of misbehaviour )
14:06 trinque I'm most interested in whether the single-threadedness accidentally did anything useful. iirc you mentioned this the other day.
14:06 * trinque bbl
14:08 asciilifeform trinque: it 'useful' in the sense that the author found himself utterly unable to write functioning locks that'd permit any measure of parallelism (e.g. correctly handling verification of mempool tx while an incoming potential-block is being eaten; or for that matter responding to ~any other command while some form of verification is taking place)
14:08 asciilifeform he simply 'locked everything'
14:09 asciilifeform thing was written as if the year were 1958 and none of the thinking re db transaction logic had yet happened.
14:12 asciilifeform tbf correctly and safely parallelizing is ~extremely~ painful in cpp planet.
14:13 asciilifeform ( and worse, the Right Thing in cpp aint ~visually~ distinguishable from subtly-lethal brokenness )
14:14 asciilifeform a pill imho would be ada's 'tasks', 1 of the items that originally drew asciilifeform into the lang.
14:15 asciilifeform theoretically, the only lock absolutely req'd by bitcoinism algos, is around block-acceptance.
14:16 asciilifeform ( i.e. it is unacceptable to evaluate a proposed input mid-way through a block-eating )
14:17 asciilifeform a good chunk of the logic in trb re individual tx, where attempts to evaluate incoming mempool tx not only for correctness and the existence of inputs / validity of spending sigs , but 'not contradict each other' -- is dead weight imho.
14:18 asciilifeform there is no guarantee that 2+ tx bounding around globally in mempools dun 'contradict one another'. why should there be attempt of such resolution inside individual node's mempool ?
14:18 asciilifeform only ~blocks~ are absolutely required to avoid contradicting previous block on max-pow chain.
14:19 asciilifeform as i understand, tx winnowing was only put in trb because the conventional node was also expected to behave as miner.
14:21 asciilifeform ( miner -- indeed requires 'find me <1e6bytes of mempool tx that dun contradict one another' )
14:24 asciilifeform this is an O(N^2) job that really is imho the responsibility of a miner strictly. 0 reason for nodes to do it for phree.
14:24 asciilifeform ( not to mention, they -- as a global whole -- dun actually ~do~ the job ! merely attempt to. and arrive at ~random conclusions. )
~ 7 hours 6 minutes ~
21:31 feedbot http://mvdstandard.net/2020/08/china-warns-new-delhi-after-indian-military-incursion-in-disputed-aksai-chin-border-region/ << The Montevideo Standard -- China Warns New Delhi After Indian Military Incursion In Disputed Aksai Chin Border Region
← 2020-08-30 | 2020-09-01 →