Hide Idle (>14 d.) Chans
← 2021-12-21 | 2021-12-29 → ↓
16:31 billymg test
16:31 billymg http://logs.bitdash.io/asciilifeform/2021-12-28#1069879 << yup, looks like it lost the db connection again. just restarted
↖
16:31 bitbot (asciilifeform) 2021-12-28 asciilifeform: btw billymg's pestlogger appears to be down atm
~ 4 hours 4 minutes ~
20:36 signpost db connection issue?
↖
20:36 signpost and howdy billymg, will send connection deets shortly
20:50 signpost billymg: http://paste.deedbot.org/?id=wTj3
20:50 signpost lmk if it works
20:57 signpost awt: possibly minor bug in the wot command. I've got both billymg and whaack added, but only see billymg. both are in the state where no packets have been exchanged yet
20:58 signpost curious whether whaack's will work
20:58 signpost interestingly whaack does show up in weechat as a joined nick
20:58 * signpost will peek at the coad for `wot` in a sec
21:08 signpost same for shinohai, not showing him in wot yet, only billymg
21:08 signpost (and others already with *active* peerings)
21:12 awt signpost: there should be a JOIN message sent to your client after you /peer someone. Possibly missed it somehow. If you reconnect, you should get every peer.
21:12 signpost yeah, the irc side shows everybody
21:12 signpost I mean that those two folks are missing in the response of the wot command
21:14 awt signpost: you can also run sqlite3 blatta.db and check out the wot table.
21:14 signpost http://paste.deedbot.org/?id=p_BV
21:14 signpost response of `wot` command, with addresses of active peers removed
21:14 * signpost checks sqlite deeb
21:15 signpost awt: wot has IDs for everybody, and handles has the expected content
21:15 signpost keys also
21:18 awt Can also confirm that the line for whaack is not being sent in the debug log.
21:18 signpost newp, just those three. I'll play with the query a bit and see if I can reproduce more usefully for ya
21:19 awt ok
21:21 signpost wot knows about the missing names if I specify them individually, so I wager there's something with self.state.get_peers()
21:23 signpost awt: ah yeah, I think there's a bug in is_duplicate.
21:23 signpost in python None == None
21:24 signpost so it's considering those gents a duplicate of billymg, I believe.
21:25 awt aha!
21:25 awt I likely have never had two peers w/o addresses in my db before.
21:28 awt ty for digging into that signpost
21:29 signpost happy to
21:30 * signpost bbl
~ 48 minutes ~
22:19 asciilifeform ftr anuther dupe:
22:19 asciilifeform awt | ty for digging into that signpost
22:19 asciilifeform awt[signpost] | ty for digging into that signpost
22:19 asciilifeform oh and lulzily, a selfdupe: asciilifeform[awt] | awt | ty for digging into that signpost
22:19 asciilifeform ^ these oughta be trivial to abolish mechanically
22:26 shinohai hello signpost
22:27 asciilifeform hm asciilifeform was defo peered with shinohai , but in AT sees last update in nov. and currently seeing his broadcasts relayed via awt & billymg
22:27 asciilifeform and lol anuther selfdupe
22:27 asciilifeform shinohai: what do you have in yer AT for asciilifeform ?
22:27 * signpost doesn't see direct peering from shinohai yet
22:28 signpost only via awt
22:28 * signpost brb, acquiring calories
~ 36 minutes ~
23:05 billymg having some issues peering and keying currently (tried with signpost and whaack), might need to restart my blatta instance.
23:05 billymg strangely was able to at though
23:05 billymg error simply e.g.: "error attempting to add peer signpost" (and nothing more in logs)
23:08 billymg hrm, same error after restarting the blatta instance
23:11 billymg awt: any idea what the error might be or where i could look to try to debug?
23:17 awt billymg you could remove the exception handling around the code causing that error and get a stack trace
23:31 billymg awt: http://paste.deedbot.org/?id=jXXJ
23:36 awt billymg: are you on 9983?
23:36 billymg 9985
23:37 awt Oh ok.
23:38 billymg could that be it?
23:39 awt billymg there may have been a bug in 9985 - anyway the broken code is updated in 9983.
23:41 awt also difficult for me to debug 9985 since I don't have that pressed at the moment.
23:47 billymg awt: ah, cool. ok i'll try out 9983 then, though it won't be right away since i'm away from my vtron box
23:49 awt billymg: cool. the issue is I think just that it should be cursor() instead of cursor, or cursor is out of scope or something.
23:50 signpost awt: I'm curious if the bug from earlier is preventing my peering with shinohai
23:50 signpost since I have 3 different edges that are not active yet
23:50 signpost btw just got a dupe from you for my first message just now
23:50 signpost so far no dupe for other two.
23:50 awt yeah. there will be dupes until the next release.
23:51 signpost nbd, just giving ya signal
23:52 awt signpost: do you mean the bug related to is_duplicate?
23:52 signpost yeah
23:52 signpost would shinohai not showing up in get_peers prevent him from ending up in my AT?
23:53 signpost basically wondering if I'm blocked on billymg since he's first in line.
23:53 * signpost tries haxing in a check for None in is_duplicate
↖
23:53 awt signpost: yeah that could be a problem
23:54 signpost yep, hai shinohai
23:55 signpost "inserted new at entry for shinohai" in log
23:55 awt hah I know shinohai got that because I got a dupe via shinohai lol
23:56 signpost got whaack also
23:56 whaack heyo
23:56 signpost hello!
23:56 awt noice
23:56 whaack wow
23:57 whaack why is there signpost and awt[signpost]
↖
23:58 signpost whaack: slight b00g awt is aware of
23:58 whaack ah okay
23:58 awt there's a bug that I think modifes messages occasionally that causes messages to slip through the dupe filter
← 2021-12-21 | 2021-12-29 → ↑

Random(pest) | Download hourly DB snapshot | Get Source Code

Theme: Light | Dark | System Default