#ossasepia | 2020-01-20

http://younghands.club/2020/01/20/wh-review-of-week-14-jan-13th-jan-19th/ << Young Hands Club -- WH Review of Week 14 (Jan 13th - Jan 19th)

~ 1 hours 15 minutes ~

02:53

http://younghands.club/2020/01/20/ejb-review-week-4-jan-13-jan-19/ << Young Hands Club -- ejb review: week 4 (Jan 13 - Jan 19)

~ 15 minutes ~

03:08

http://younghands.club/2020/01/20/ejb-plan-week-5-jan-20-jan-26/ << Young Hands Club -- ejb plan: week 5 (Jan 20 - Jan 26)

~ 2 hours 52 minutes ~

06:01

http://younghands.club/2020/01/20/jfw-review-week-of-jan-13-2020/ << Young Hands Club -- JFW review, week of Jan 13 2020

06:08

http://younghands.club/2020/01/20/jfw-plan-week-of-jan-20-2020/ << Young Hands Club -- JFW plan, week of Jan 20 2020

~ 1 hours 15 minutes ~

07:23

http://younghands.club/2020/01/20/rmd-review-jan-11-19th-2020/ << Young Hands Club -- RMD review, Jan 11-19th, 2020

~ 23 minutes ~

07:47

http://younghands.club/2020/01/20/rmd-plan-jan-20-24th-2020/ << Young Hands Club -- RMD plan, Jan 20-24th, 2020

~ 28 minutes ~

08:15

dorion: you do like living dangerously, don't you.

~ 7 hours 9 minutes ~

15:25

diana_coman while I've definitely taken my share of risk - some smart, some dumb - I know I can improve the balance of the ratio in favor of more smart risk.

~ 18 minutes ~

15:43

dorion: some unbalance in that direction may be even better! but yes, setting yourself up for less dumb risk will do wonders and on more fronts than one really.

~ 22 minutes ~

16:06

for sure. set up to let compound interest of smart risk accrue.

~ 2 hours 33 minutes ~

18:39

dorion: note the difference there between your stated "more" (risk, be it smart) and my "less" (dumb risk); while "more of smart and less of dumb" may be the theoretical ideal, the practice is best served by focusing on the "less" first; it also makes for more calm and less stress really.

18:54

diana_coman makes sense++

18:57

dorion: so how's that Hawaii article going wrt to the current deadline for it?

19:04

diana_coman good. not sure if it'll be out before I sleep tonight, but good chance it will.

19:07

iirc deadline is tomorrow so that sounds fine, glad to hear it; and no need to rush/skip the final review either.

19:09

yes, deadline is tomorrow.

19:22

whaack: why were the bots kicked in the first place anyway?

19:23

diana_coman: this i have not yet investigated.

19:23

when you say they were kicked do you mean you looked and ~all were kicked or do you mean you saw some kicked and so "bots were kicked"?

19:24

whaack: well, before "fixing" it, you *have to* investigate what the fuck the problem is in the first place, you know?

19:26

diana_coman: I looked on one VM and I discovered I was banned from ~all of the 7-8 networks I had connected to. Then I looked into a few networks and saw that my bots had been kicked from channels and were constantly trying to rejoin them.

19:27

hrm, now there's even a sf story for you but finish the odyssey or it'll take longer than it took ulysses by the sounds of it.

19:27

whaack: the same on other VMs or was there only one connected?

19:28

this is quite important too because ahem, if you don't figure out and solve it, there's not going to be much systematic anything there.

19:29

diana_coman: I was running bots on two VMs and iirc yes the same thing happened on the other VM (checking now)

19:30

there was my previous clear statement re if kicked let the chan be but mk, test run and all that.

19:30

ossabot

Logged on 2019-12-19 16:41:24 diana_coman: let it join, what; collect data for 1 month and then see; if it gets kicked from a chan, let the chan be, not like I want to force anyone now out of the hole they enjoy, what.

19:31

whaack: any reason given on first kick? because the ban makes sense afterwards given the kick/rejoin dance, ofc.

19:33

I just checked, on the other VM I was banned from 3 networks w/ errors: "ERR_YOUREBANNEDCREEP: shire.digitalirc.org snup97 "You are not welcome on this network. G-Lined: join/part spam. Email https://tickets.digitalirc.org for more information."" and "NOTICE: Defiant.GeekShed.net snup97 "*** You are banned from GeekShed (You are banned from GeekShed for 8d! Reason: Malicious bots, clones and drones are prohibited by the Network TO

19:33

S." and "NOTICE: spinnaker.p2pchat.net roco44 "*** You are banned from P2PChat (either stay on or off (3 days))""

19:33

so hm, join/part spam aka...connection flicker??

19:34

no I think the join/part spam is more likely from the kick/join dance

19:34

no data re kicks themselves?

19:34

there should be data .. i have to dig for it

19:35

Maybe the bots need girl names and an Eliza plugged into them?

19:35

*

diana_coman hands whaack a shovel.

19:35

BingoBoingo: eh, I'm not (yet) trying to take over all chans too!

19:36

lol

19:36

Well, something to do call/response

19:37

they can certainly say hello but I doubt that's the issue really

19:38

and at any rate, it's not like all networks are invited to dump all their issues for whaack to have what to address.

19:48

diana_coman: The query "select message, networkname from irclog where irc_message_type='CL-IRC:IRC-KICK-MESSAGE';" gave me the following results from vm1 http://paste.deedbot.org/?id=VvJp and from vm2 http://paste.deedbot.org/?id=Igiq

19:50

ahahaha

19:50

"You have been found co-channeling in one or more blacklisted channels"

19:50

oh the lulz

19:53

but not that much variety, hm.

19:53

I notice that it was only from 3 networks that I received kicks. But my fleetlogs show that on one VM i have 4354 disconnects and 4440 joins

19:53

so there is likely something else causing the disconnect/reconnect dance. or i'm getting kicked from channels without receiving an irc-kick-message.

19:55

whaack: well, overall it sounds like you should first understand a bit better just what exactly your bots are doing there and how, really

19:55

I mean: there's so much fun to be had there but you do need to be better at it than their lame efforts, seriously.

19:57

btw, i didn't implement the reconnect-on-kick feature intentionally. with cl-irc you set a nick to have a list of channels, and i found out cl-irc automatically rejoins the nick to a network when kicked. I will have to explicitly remove the channel from the nick's list when it gets kicked.

19:58

diana_coman: what do you mean by I need to be better at it than their lame efforts?

19:58

anyway, fix so it doesn't reconnect on kick for starters; get to the bottom of that 4354 disconnects vs 4440 joins, if at all possible; add whatever additional logging you need to have full information on hand and preferably timely delivered too; maybe start with a smaller trial that you can monitor more closely to figure out what's going on, if needed

19:59

whaack: well, their "fighting the oh-noez-spam-not-ours" is from what I see a combination of "oh, you won't answer me???" and that co-channeling lulz; the rest is just repetition really.

19:59

well, unintentional implementation is ....

20:00

and anyways, their bans are for some number of days, can even rotate etc.

20:00

alright I'll take BingoBoingo's advice and see if adding an Eliza-light will help

20:00

but it's not yet even clear it's worth the effort

20:01

whaack: first of all get to the bottom of the connect/reconnect and kick/not kick, anyway; eliza-light comes afterwards and only if really justified.

20:01

diana_coman: btw do you have a way you'd like me to select the channels? I sorted by the most popular channels and joined those first...that was not because I thought that was a particularly good idea in the general case - but because I was mostly testing networks that had ~0 activity in anything but the more popular channels

20:02

it makes sense since otherwise you'd see empty lines anyway so not much gain; that's not a problem there

20:03

^ alright i'll publish a plan to address the above this evening / tomorrow midday.

20:04

whaack: is the whole code clear to you now how and why it works and what is going on in there?

20:06

diana_coman: my code and trinque's irclogger are clear to me. (btw trinque did you look into the reconnect bug I pointed out? I will confirm there's a problem and publish a vpatch for it later this week if I get the time.) how cl-irc handles certain things is a bit of a blackbox.

20:06

ossabot

Logged on 2020-01-04 22:56:08 whaack: trinque: I think I may understand this if you never got around to it. The thread that normally calls ircbot-reconnect is the ircbot-ping-thread. However the ping-thread commits suicide because ircbot-disconnect calls (sb-thread:terminate-thread (ircbot-ping-thread bot)) . So when the ping thread calls ircbot-reconnect after calling ircbot-disconnect it never

20:08

whaack: is cl-irc that huge/gnarly that it's a huge project to look and find out how it handles those exact certain things that you don't know about but need?

20:10

diana_coman: no it is comprehensible. 4.5k LOC and there is a lot of handling of portions of the irc protocol i don't need to look into.

20:11

whaack: so then look into it, since you are relying on it for the whole thing.

20:12

(4.5k LOC _and_ from my understanding it does not import any external libraries)

20:12

let eliza be for the moment; it might be a good idea but it will get evaluated when the problem and its extent are clear, not just slapped on now because "it seems like a good idea" , ugh.

20:13

(ah nvm it does use three - :split-sequence :usocket and :flexi-streams)

20:13

ahaha, it did sound rather too good to be true, lol.

20:13

alright, no eliza

20:15

whaack: enough work there to keep you busy ? clear enough so you know what you need to do?

20:16

certainly the former, and yes I think I have a good idea of what to tackle next and how to go about it.

20:17

cool then.

20:17

the other major item in the pipeline today is the plan for my computer

20:17

anyone else in need of work around here? speak up, there's plenty to pass around!

20:18

whaack: wasn't there some article too?

20:18

diana_coman: yup that is done, just needs to do the final review

20:18

i just need*

20:19

good then; any trouble re plan for your computer?

20:19

i don't think there should be, but i haven't made a meta plan lol

20:20

whaack: ahaha, no, no need for that or you'll loop forever in the meta-metas; I simply meant if you needed to clarify anything on that front.

20:21

diana_coman: no, although I am not sure how much detail I need to go into for how I go about building the computer. Do I need to spell out how all the cables are going to be laid in what directions + what gets zip tied to what?

20:23

Currently I'm thinking I list all the parts I need to install, what order I install them in, and what items i will need / have available in order to do their install. I also will include any important installation notes for various pieces.

20:24

whaack: no need for cable-level of detail, no; whenever in such doubt, focus on what the goal of the plan (or whatever it is you are working on) really is

20:24

that's where everything else should come from; so: what's the goal of this plan of yours?

20:25

diana_coman: The goal is to ensure I have what I need and know what to do in order to execute a timely, safe, and proper build of my computer.

20:26

safe means I don't open up the possibility of breaking something, and proper means I don't make a mistake that requires me to do undo a bunch of work to fix.

20:28

whaack: so then, does that require a list of what gets zip tied and cable directions? (it might or it might not, but it's your decision since you defined the goal of what you need, so ..up to you)

20:30

diana_coman: right...that's why I asked, I'm not so sure. There are certain cables that must be laid out before installing other parts. So I think yes I will have to consider some of the cable layout in my plan.

20:31

whaack: so then consider it, sure; it's *your* plan and it should serve you best, first and foremost.

20:31

*

BingoBoingo recommends saving zip ties for last. Otherwise there's opportunities for friction as you plug and unplug cables in search of the cleanest routes, the friction being cutting and placing new zip ties.

20:31

^ noted

20:32

aha, not a bad idea; and in general, tidying stuff neatly up tends to be at the end precisely because otherwise it might get untidy again before everything is done.

20:33

You also don't want to be putting any cables under unnecessary tension. Once you're satisfied all cables are plugged in with a comfortable amount of slack, then zip tying can begin.

~ 1 hours 42 minutes ~

22:15

oh hey, nice photos, whaack ! and you basically showed Panama City more than jfw and dorion so far combined, ha! takes a non-local to publish photos of a place, it would seem.

22:16

diana_coman: aha that's how it goes I guess. I have ~no photos of NYC.

22:22

So much more glass and height than Montevideo has.

22:22

BingoBoingo: more recent former money place I guess.

22:23

diana_coman: Likely.

22:26

I suspect Uruguay's generally less accomodating to height variance requests leading to 10 to 12 storey canyons on a number of streets.

22:30

Anyways, my experience here is that the visitors have always come with fresher eyes. It doesn't seem to take much time for the visual landscape anywhere to start feeling normal.

22:33

BingoBoingo: if by normal you mean "perpendicular" (to my disposition at times if to nothing else), I'll fully agree to that - it doesn't take much time at all!! lolz

22:36

diana_coman: By normal I mean, the brain starts applying a "this place is familiar" filter over what they eyes are sending it. There seems to be a window first arriving in a new place where the brain's "Everything in visual range is new!" gets the shutter button on the camera moving.

22:39

BingoBoingo: I got what you meant, just being amused otherwise because despite knowing what you mean, I am - if anything - even less able to take a decent picture of something new than of something more familiar.

22:42

Well, sometimes things are both new and familiar. The local Afro cultists might leave a dead chicken in a new place, but once you've seen enough chickens left in odd locations...

22:46

lol, that's one sort of new and familiar, can be.

22:51