Posts by zop@7td.org
(DIR) Post #AQkGDdds1oHjtlkxyi by zop@7td.org
2022-12-18T22:33:24.822051Z
0 likes, 0 repeats
@besserwisser @feld Twitter is no longer a public company. As for usage being up, we don’t have any source for that. Nothing Space Karen reports can be taken as truthful as he’s very clearly demonstrated. He has every reason to lie about mDAU numbers, and since he made sure to boot out all the adult supervision there’s likely no one left to actually compile a report that can be trusted. He has a looooong history of promising far more than can be or will be delivered.
(DIR) Post #ARMqXyjsHUMxq2Suhs by zop@7td.org
2023-01-06T02:56:48.828379Z
1 likes, 0 repeats
@feld thanks for fucking with my headspace buddy.
(DIR) Post #AS0TwKh0JO3zEAPUJM by zop@7td.org
2023-01-25T16:15:00.344287Z
0 likes, 0 repeats
@himay @feld @paul correct. 128-bit FlakeID - DB side PostGreSQL UUID - under the hood https://git.pleroma.social/pleroma/elixir-libraries/flake_id
(DIR) Post #AY3hWALUmfM9Bfk4OW by zop@7td.org
2023-07-25T16:43:42.938889Z
0 likes, 0 repeats
Why do so many web sites require two or three clicks during sign in/log in flow even without 2FA. I can not see the sense in having an entirely separate page load just to get the username. Followed by another for the password, and so many times a third to get past whatever other nonsense they want to stuff into the flow.
(DIR) Post #AZ858c7zwl6D7Dm0Q4 by zop@7td.org
2023-08-26T17:19:34.449286Z
1 likes, 0 repeats
@cstross @darrelplant @helenczerski @shiri a part of why cars are getting physically larger (besides the SUV craze started here in the US) are crash safety standards. You can’t absorb energy without giving it some space to dissipate. We’ve crammed the engine compartment about as much as it’s possible to increase interior and cargo volume. And at least here in the US people are on average a lot larger too than decades ago.
(DIR) Post #AZJ7UeAvLwlWTzKpPM by zop@7td.org
2023-09-01T01:06:23.977975Z
1 likes, 0 repeats
@tracketpacer this anodes me greatly :)
(DIR) Post #AbJ73SpMMHrozLracq by zop@7td.org
2023-10-30T20:11:17.272324Z
1 likes, 0 repeats
Completed some major open heart surgery of my blog. Sayonara WordPress, Hello Hugo. Saving grace is despite the site dating back to ~2008 I don't really have that much content, so there wasn't a ton of fixup to do. Still a few minor things bugging me, but in general it's all working now as a static site.
(DIR) Post #AbJ73V51yzahyXD8KW by zop@7td.org
2023-10-30T20:43:57.941956Z
0 likes, 0 repeats
I'm also not sure what exactly happened ... probably just the fact the I've let the site go really stale ... but it looks like Google may have stopped (re)indexing the blog site. I've hit what should be the appropriate endpoints to force it to crawl again. Looks like it's basically fallen out of the search index too...not that it matters, I think the ONLY thing that ever got hit with search was a post I wrote about getting the old Apple Time Capsule's working w/ Windows 7 RC1 and later. That's probably largely non-applicable nowadays.
(DIR) Post #AbKazjJEKTMajLvkci by zop@7td.org
2023-10-31T12:28:10.181201Z
0 likes, 0 repeats
@bruce @FrayJay @arstechnica exactly. Heat or A/C (an option that might not be installed since it’s 10k-20k $USD extra by the way) can really eat up range. Some actually offer a diesel fired heater for that reason. I don’t know if any of them use heat pump type systems instead of resistive electric heat. Tours are quite a lot of stop and go throughout however many miles it is (ours range from 30-60 miles, 60-120/day for the small rural bus co I’m sub for)
(DIR) Post #AbKazl3LqTqm8gW69Q by zop@7td.org
2023-10-31T12:45:38.967199Z
1 likes, 0 repeats
@FrayJay @bruce @arstechnica there’s not any time between multiple tours at the same end of the day. If you’re running two or three tours for a school that’s three back to back. Or you’re starting that first tour a couple of hours earlier. The early route might already be having to have kids up and at their stop for pick up by 0600 All the kids have to be at school before the first bell. Afternoon routes same deal only there has to be paid staff at the school until the last kids leave and the last bus finishes unloading the last kid, at 1700-1800 - real times from when I was a kid my rural route had me on the “early” AM and “late” PM where nearly each bus driver was doing two tours AM/PM. You can’t just go back to the bus barn to charge it.
(DIR) Post #AbQsaTfnMOznwuATK4 by zop@7td.org
2023-11-03T13:46:45.423578Z
0 likes, 0 repeats
@feld I really hadn’t seen that before and that’s pretty funny IMO.
(DIR) Post #AbR9wh8bg5rzIhp9U0 by zop@7td.org
2023-11-03T17:31:34.113498Z
1 likes, 0 repeats
@feld on Charter/Spectrum@here if you don’t request the expected size via DHCPv6 it issues nothing. You’ll get the interface for the firewall by idk SLAAC or whatever it is but no PD, DNS, anything, if your PD request isn’t the right number of prefix bits (I think it’s /56? I’d have to look at my pfSense) … pfSense was also and may also so be broken, you can’t have more than one IPv6 upstream unless they’re static.
(DIR) Post #AbXFRaxqLXrbql6RRA by zop@7td.org
2023-11-06T14:27:09.299351Z
1 likes, 0 repeats
@feld hey at least your teeth are strong and your breath is fresh even if your brain is rotting, right?
(DIR) Post #Abu1qTqwFXMFwoNYMy by zop@7td.org
2023-11-17T15:53:59.659866Z
1 likes, 0 repeats
Had a bit of an outage from yesterday. Unclear what happened because the VM was completely locked up when I found it, on reboot Pg did not come up, probably a systemd problem because it tried to start the Pg service before it finished mounting filesystems. All back to normal now. ISh. Still catching up on queued federations.
(DIR) Post #AdDee7NjMhHXr71uEq by zop@7td.org
2023-12-27T01:19:07.204091Z
1 likes, 0 repeats
@feld @mWare I’ve half started writing an Elixir MSA a few times due to my frustrations with existing stuff…
(DIR) Post #AjuGk5ayxvhVkAQvBY by zop@7td.org
2024-07-14T01:58:45.421732Z
0 likes, 0 repeats
@p @stux good for you you found the ONE fucking whataboutism! How about the people mowed down in Charlottesville? How about Jan 6? How about the dead in Pulse nightclub? Club Q? The MAGAts are already conducting warfare against the United States and its citizens.
(DIR) Post #B1yw8xVIc0xFRtZKtc by zop@7td.org
2026-01-05T02:02:44.465244Z
1 likes, 0 repeats
It went poorly! I figured out the rebar3 pc dependency problem while I was offline at least (old cached version...rebar3 not updating if cache existed).As for the database, still on the old PostgreSQL version. The dump/reload process stopped with the backend workers not apparently making any progress (re)creating an index. It might just be my maintenance_work_mem setting being way too low to recreate the particular index efficiently but...I'll have to defer upgrading it and look into it some other time.
(DIR) Post #B1yw92PoLqXGgK2WXI by zop@7td.org
2026-01-05T02:12:46.840709Z
1 likes, 0 repeats
For anyone curios, it was apparently hung (CPUs spinning on two workers, no real progress...WAL was being created though, slowly) here during the load:CREATE INDEX "activities_visibility_index" ON "public"."activities" USING "btree" ("public"."activity_visibility"("actor", "recipients", "data"), "id" DESC NULLS LAST) WHERE (("data" ->> 'type'::"text") = 'Create'::"text");It was given hours to complete, and it should've been able to in that time. Overall size isn't that big on the source, 2.8GB. ~3.75M rowspleroma=# \d+ List of relations Schema | Name | Type | Owner | Persistence | Access method | Size | Description--------+---------------------------------------------------+----------+----------+-------------+---------------+------------+------------- public | activities | table | pleroma | permanent | heap | 2857 MB |....pleroma=# select count(1) from activities; count--------- 3768228(1 row)Time: 2314.623 ms (00:02.315)pleroma=#I did abort once, and bumped the maintenance_work_mem way up (256MB -> 1GB) but it may need more. I'll play around with reloading the data into a Pg17 instance asynchronously though. This might be some wack artifact from the OS upgrade today, because the underlying pgdg postgresql package version didn't change, the Pg17 build may not actually be running the correct package. So I'm going to first just do an apt-get install --reinstall on the Pg17 packages, and just give it another go using the dump/load part manually and not as an offline pg_upgradecluster to see if it works.
(DIR) Post #B20zjwsGbVYOODYs1g by zop@7td.org
2026-01-06T16:23:31.470685Z
1 likes, 0 repeats
For anyone backing up a pleroma instance DB, if you're using an sql backup (all variants of pg_dump) - YA DON'T HAVE A WORKABLE RESTORE WITHOUT SOME MASSAGING! You MUST have indexes pre-created to load data. Post creation of indexes effectively won't work...unless your DB is small.Soooo I think I know what causes the restores to basically never complete, maybe...I'm near the limits of my psql debugging capability because I've never hit this so I actually don't exactly know how to see what the CREATE INDEX is actively doing... But tl;dr; 1.8T rows. One Point Eight TRILLION rows. But...HOW?! The dump+load upgrade method I tend to use causes this. In the default mode it creates the schema (sequences setup, all the functions, that sort of thing, BUT NO INDEXES), loads the data, then creates the indexes.The index it's never finishing creating is calling this function:CREATE OR REPLACE FUNCTION public.activity_visibility(actor character varying, recipients character varying[], data jsonb) RETURNS character varying LANGUAGE plpgsql IMMUTABLE PARALLEL SAFE SECURITY DEFINERAS $function$DECLARE fa varchar; public varchar := 'https://www.w3.org/ns/activitystreams#Public';BEGIN SELECT COALESCE(users.follower_address, '') into fa from public.users where users.ap_id = actor; IF data->'to' ? public THEN RETURN 'public'; ELSIF data->'cc' ? public THEN RETURN 'unlisted'; ELSIF ARRAY[fa] && recipients THEN RETURN 'private'; ELSIF not(ARRAY[fa, public] && recipients) THEN RETURN 'direct'; ELSE RETURN 'unknown'; END IF;END;$function$Well. users is ~500k rows, and this function would be called for the ~3.7M activities rows....at a point in time where users has no indexes! So each invocation of activity_visibility would scan ~500k rows, times 3.7M rows in activities, and .... that's ~1.8T rows.Why did I footgun myself like this when pg_upgrade exists? Well, binary incompatibilities, index collation differences, in the past causing Major Heartburn. I'm not going to dive into how it might be possible to fix the dump+load procedure, since, well, yeah. But for anyone backing up a pleroma DB be warned. Either take a binary copy (f/ex pg_basebackup) or dump schema and data seperately so you can load the data with indexes. Or find another workaround!cc @feld
(DIR) Post #B20zk2AWv28KoJ91ge by zop@7td.org
2026-01-06T16:24:42.148024Z
0 likes, 0 repeats
I should probably turn this into a blog post.