#trilema | 2018-01-27

ohai BingoBoingo

10:38

BingoBoingo: rack ready ?

10:39

BingoBoingo

Getting close

10:39

what's missing ?

10:40

BingoBoingo

Wire is expected to hit the datacenter account Monday. This weekend configuring and double checking the configuring of the switch and the musl Qntra

~ 23 minutes ~

11:04

"How Two, Bored, 1970's Housewives Helped Create the PC Industry www.fastcompany.com"

11:05

"Important Women in CS Who Aren't Grace Hopper www.hillelwayne.com"

11:17

https://thenevadaindependent.com/article/blockchain-company-buys-67000-acres-at-northern-nevada-industrial-park <<< New flaming-tire-in-shitpit factory?

11:17

(Won't archive )

~ 1 hours 12 minutes ~

12:29

!~later tell mircea_popescu http://btcbase.org/log/2018-01-26#1776901 <-- neato! ftr, it seems that pingback responses vary a lot across web server configs, wp versions etc. some blogs. for example wordpress.com-hosted blogs give undocumented (or sometimes empty) responses, probably as an anti-spam measure. (after filtering out non-pingbackabble sites, I sent all pingbacks manually, just to look at responses)

12:29

Logged on 2018-01-26 16:00 mircea_popescu: spyked all in.

12:29

jhvh1

spyked: The operation succeeded.

~ 46 minutes ~

13:15

mornin'

13:23

Heya mod6 .... afternoon! lol

13:26

i guess it is huh haha

13:27

*

mod6 worked hard yesterday, extra zzz's needed.

~ 45 minutes ~

14:12

Lords and Ladies of The Most Serene Republic: Please let me know if you have any updates for your Advertised Republican Nodes to conclude the month of January.

14:15

I have none besides advertised nodes I showed you yesterday.

14:16

oh from bitnodes?

14:17

yeah.

14:17

Ah, ok.

~ 39 minutes ~

14:57

spyked, fwiw, pingback arrived fine

15:01

hi diana_coman

15:01

hi mod6

15:01

how's eulora stuff these days?

15:02

re nodes: mine is still there but still not at the top; I even pressed and ran asciilifeform's patch but it's unclear to me if it helped tbh; I admit I did not have much time to really dig deeper there exactly

15:03

eulora is chugging along, atm focus is mainly on getting eucrypt done

15:03

*nod*

15:04

does your node use an SSD? mine doesn't and is @ 506042, and the net is @ 506402

15:05

heh, that moment where 'mimisbrunnr' is a block ahead of other blockchain sites...

15:06

so yeah, mine lags behind too because of sloooooow disk.

15:06

(and fwiw, im running the aggressive push blocks too, which should help. but too IO bound to matter with slow disk)

15:10

ah, that must be it then; no ssd :(

15:11

diana_coman, yeah, but your webserver is ultra-conservative on user agents. :) the one from http://btcbase.org/log/2018-01-19#1773164 didn't work, so I faked one from an existing browser, which eventually did the trick. (nb for whoever wants to craft xml-rpc requests manually)

15:11

Logged on 2018-01-19 18:34 mircea_popescu: !~later tell spyked hey, does your tarpit not send pingbacks or is there some error ? what's the response if you say curl -v -A "Mozilla/5.0" -r 0-4096 --connect-timeout 30 --max-time 10 "http://www.dianacoman.com/xmlrpc.php" --header "Content-Type: text/xml" --data "<?xmlversion="1.0"?><methodCall><methodName>pingback.ping</methodName><params><param><value><string>http://trilema.com/2017/re-reading-is-the-most-powerful-tool/

15:14

hm, what was the error you got on that one that didn't work?

15:18

diana_coman: here's the output: http://p.bvulpes.com/pastes/Nlokl/?raw=true and the script I used http://p.bvulpes.com/pastes/z0nep/?raw=true it would seem ngnix expects user agents to meet a particular spec (that I haven't figured out exactly) before delivering pages

15:19

(and it's not obvious from the error message either; but it's possible that there's a bit more info in the server logs)

15:21

spyked, thanks; looks familiar even so one of these days I'll have to get around to look at it and shoot it in the head

15:22

fwiw it looks like some mod_security mess

~ 23 minutes ~

15:46

*

spyked took a look through how mod_security does user agent-based filtering and stumbled upon http://archive.is/ELI4u#selection-1478.0-1518.2 lulz; no idea how a programmatically-generated string can be botched up just like that.

15:49

LUL

15:51

hanbot

~ 1 hours 19 minutes ~

17:10

!~ticker --market all

17:10

jhvh1

ben_vulpes: Bitstamp BTCUSD last: 11551.14, vol: 10348.41527763 | Bitfinex BTCUSD last: 11570.0, vol: 34357.10597093 | Kraken BTCUSD last: 11529.1, vol: 5932.07693913 | Volume-weighted last average: 11561.3543874

17:10

http://btcbase.org/log/2018-01-27#1777388 << at the risk of repeating -- mechanical hdd is ~useless for trb work

17:10

Logged on 2018-01-27 20:10 diana_coman: ah, that must be it then; no ssd :(

17:11

it sorta, possibly, worx, when in striped-raid; and worked ok prior to fullblox era

17:11

but today -- nope.

17:11

a mechanical-hdd node is possibly even worse than none. causes chronic frustration for its owner and anybody who tries to sync from it

17:12

really a kind of tarpit.

17:13

potentially they could work at something like normal speed, with something like http://btcbase.org/log/2018-01-25#1776380 . but that's not a Troo Adult Noad

17:13

Logged on 2018-01-25 19:29 asciilifeform: btw i'll mention again a simple method to speed up sync (to a point) 1,000x or moar -- signed hashes ( a la the programmable checkpoint thread from 2yrs ago )

17:16

http://btcbase.org/log/2018-01-27#1777381 << i was thinking, the '% of wallclock time spent waiting for blockverifier' , '% of wallclock time spent waiting for newblock' stats really oughta be part of standard 'getinfo' output

17:16

Logged on 2018-01-27 20:02 diana_coman: re nodes: mine is still there but still not at the top; I even pressed and ran asciilifeform's patch but it's unclear to me if it helped tbh; I admit I did not have much time to really dig deeper there exactly

17:16

mod6 et al ^

17:17

it will readily reveal the cause of a 'behind' node

17:17

though i will point out, for completeness, that i saw NO substantial 'behinditude' on zoolag since feeding it the pill against shitoshi's sync retardation.

17:18

i had the above stats patch on the conveyor, but abandoned when saw that there is no mystery anymoar for it to resolve..

17:19

works just fine with striped raid

17:19

not a bad idea asciilifeform

17:19

ben_vulpes: plox to quantify 'just fine'

17:20

dulap-II had a serious striped-raid ; still took 5x as long to verify same block as fist-sized, raidless ssd zoolag.

17:20

the latter having ~1/32th the cpu horse of the former.

17:20

obviously is going to be slow, but mimi has zero trouble keeping up with the tip-o-the-chain since applying your patch

17:20

striped raid might solve write speed issue, but not seek time issue

17:21

phf: actually opposite

17:21

thinkaboutit

17:21

err, actually scratch the whole thing

17:22

well actually it'd be : ~read~ seek-time is shortened on striped-raid; write-seek -- slightly increased ; absolute contiguous speed of r/w -- unchanged

17:22

what i meant to say was read speed issue, but not seek time issue, but i'm thinking that seek time ~might~ improve because of parallelizable seeking

17:22

( exercise for the reader, to see why )

17:23

( in 'striping' , the platters turn deliberately out of phase )

~ 15 minutes ~

17:39

mod6: http://p.bvulpes.com/pastes/ImgDW/?raw=true

17:39

*

mod6 looks

17:40

*

ben_vulpes zoidbergs

17:41

ben_vulpes: got it

17:48

*

asciilifeform polishing off ch9. which will contain, among other things, The Seekrit Of all-iron guaranteed-constanttime mul , with only ~8% penalty over iron mul.

17:49

which afaik has never been published anywhere.

17:49

Lords and Ladies of The Most Serene Republic, I present to you a pre-patched *EXPERIMENTAL* version of my vtron (99993), along with a patch to show changes from V (99994): http://therealbitcoin.org/ml/btc-dev/2018-January/000287.html

17:50