| Time |
Nick |
Message |
| 00:34 |
|
Bmagic joined #evergreen |
| 05:17 |
|
book` joined #evergreen |
| 07:21 |
|
collum joined #evergreen |
| 08:36 |
|
mmorgan joined #evergreen |
| 08:47 |
|
mantis joined #evergreen |
| 09:33 |
|
Dyrcona joined #evergreen |
| 09:34 |
Dyrcona |
Bmagic: Don't ask me what future me remembers. I'll have to get back to you last week. :) |
| 09:34 |
Bmagic |
Dyrcona++ |
| 09:43 |
csharp_ |
upgrading the evergreen email list server, FYI - should be short downtime |
| 09:44 |
Dyrcona |
csharp_++ |
| 09:44 |
* Dyrcona |
checks the git server. |
| 09:44 |
|
dguarrac joined #evergreen |
| 09:45 |
Dyrcona |
Imagonna reboot the git server after updates install. It'll be down 5 minutes tops. |
| 09:46 |
Dyrcona |
I usually announce it, but this is spur of the moment. |
| 09:46 |
Dyrcona |
Tell ya what. I'll pick a time and reboot it then. |
| 09:51 |
Dyrcona |
"at not installed." OK. No biggie. |
| 09:52 |
Dyrcona |
Four thirty pm on a Friday...Best time ever to do anything..... |
| 10:13 |
Dyrcona |
Hmm. I thought that there was a bug for the emergency closing handler timing out/being really slow under some circumstances, but I can't find it in Lp. |
| 10:16 |
Dyrcona |
Nothing quite like arbitrarily closing old tickets on a Friday. :) |
| 10:22 |
csharp_ |
well sheeeeeitt |
| 10:22 |
csharp_ |
mailman is, like not even a package in 22.04 |
| 10:22 |
csharp_ |
and it depends on a lot of python2 stuff - seeing what's possible |
| 10:29 |
Dyrcona |
csharp_: Mailman 3 is Python 3. |
| 10:30 |
csharp_ |
yeah |
| 10:31 |
|
redavis joined #evergreen |
| 10:33 |
csharp_ |
gonna plow ahead for a bit and see if I can get mailman3 working on 24.04 |
| 10:33 |
csharp_ |
if not I can revert to 20.04 |
| 10:33 |
csharp_ |
upgrades-- |
| 10:33 |
Dyrcona |
I've been saying that Ubuntu upgrades lately have been feeling more like downgrades. |
| 10:34 |
csharp_ |
yeah :-( |
| 10:34 |
csharp_ |
if 20.04 weren't EOL soon I would have left it |
| 10:35 |
Dyrcona |
20.04 had itssues, too. |
| 10:35 |
Dyrcona |
Trusty Tahr forever! :) |
| 10:36 |
csharp_ |
ah, that puts your comment the other day in context |
| 10:36 |
csharp_ |
and yes, that was a good release |
| 10:37 |
csharp_ |
but yeah, if you're running mailman, 22.04 is a no-(mail)-man's land |
| 10:37 |
csharp_ |
it's either 20.04 or 24.04 if you run LTSs |
| 10:38 |
csharp_ |
24.04 = the new 14.04? |
| 10:39 |
csharp_ |
my first Ubuntu install was 7.10 in early 2008 - ah that was fun |
| 10:39 |
csharp_ |
gutsy gibbon |
| 10:41 |
Dyrcona |
24.04 on desktop is crap. |
| 10:41 |
Dyrcona |
I'm using it. |
| 10:41 |
csharp_ |
heh |
| 10:41 |
csharp_ |
I haven't used Ubuntu as my daily driver for a long time |
| 10:42 |
berick |
24.10 gives me what i need |
| 11:16 |
Dyrcona |
Well, disabling and enabling the user service for emacs with systemd did not resolve the issues, so I'm going to stop trying to run it as a service on Ubuntu 24.04. |
| 11:21 |
Dyrcona |
berick: Do you use the XML SMS notices with message bee for curbside pickup? |
| 11:22 |
berick |
we use them for locker pickup. not using the curbside evergreen pieces. |
| 11:22 |
berick |
guessing it's about the same |
| 11:22 |
Dyrcona |
I wonder if I have to contact UMS to let them know we're adding it? |
| 11:23 |
berick |
i think you have to contact them to add pretty much anything, setup email templates, etc. |
| 11:26 |
Dyrcona |
Yeah. That's what I think, too. |
| 11:26 |
Dyrcona |
Thanks! |
| 11:30 |
Dyrcona |
I found an email where we sent them language for these last year, so I guess I need to check with Unique to let them know that we want to start sending them, since we haven't so far. |
| 11:35 |
csharp_ |
ok - gonna move back to 20.04 :-( - mailman3 won't install because the python/django configurations are utterly broken |
| 11:35 |
Dyrcona |
csharp_: Sounds 'bout right. |
| 11:35 |
csharp_ |
this is the point where if I were smarter, I'd just revert to the snapshot I took before upgrading |
| 11:36 |
csharp_ |
<padme>You took a snapshot, right? RIGHT?</padme> |
| 11:36 |
berick |
csharp_++ # fightin the good fight |
| 11:36 |
csharp_ |
thanks for the pats on the back, y'all |
| 11:37 |
bshum |
csharp_++ # you got this! |
| 11:37 |
csharp_ |
one crazy-pants option I thought about was trying to run a 20.04 docker on top of the 24.04 I have now, but then I'd have (lost count) uh, 4 problems |
| 11:37 |
csharp_ |
bshum: :-) |
| 11:39 |
csharp_ |
docker config + mailman + apache + network + postfix = NEVER MIND |
| 11:39 |
Bmagic |
lol |
| 11:40 |
Bmagic |
I wouldn't do it for the sole reason, it's still 20.04. You can't polish a turd (wrap it in docker) and make it any better :) |
| 11:40 |
csharp_ |
agreed |
| 11:40 |
abneiman |
@quote add < Dyrcona > Don't ask me what future me remembers. I'll have to get back to you last week. |
| 11:40 |
pinesol |
abneiman: Error: You must be registered to use this command. If you are already registered, you must either identify (using the identify command) or add a hostmask matching your current hostmask (using the "hostmask add" command). |
| 11:40 |
abneiman |
bite me, pinesol |
| 11:40 |
csharp_ |
@quote add < Dyrcona > Don't ask me what future me remembers. I'll have to get back to you last week. |
| 11:40 |
pinesol |
csharp_: Error: You must be registered to use this command. If you are already registered, you must either identify (using the identify command) or add a hostmask matching your current hostmask (using the "hostmask add" command). |
| 11:40 |
csharp_ |
pinesol: hey! |
| 11:40 |
pinesol |
csharp_: I eat more coconut cream pie before breakfast than most people eat all day |
| 11:41 |
abneiman |
oh snap, pinesol has some sauce today |
| 11:41 |
Bmagic |
@quote add <Dyrcona> Don't ask me what future me remembers. I'll have to get back to you last week. |
| 11:41 |
pinesol |
Bmagic: Error: You must be registered to use this command. If you are already registered, you must either identify (using the identify command) or add a hostmask matching your current hostmask (using the "hostmask add" command). |
| 11:41 |
csharp_ |
pinesol: you break-a my heart |
| 11:41 |
pinesol |
csharp_: Message root @ server God....Universe going down for reboot.... |
| 11:41 |
Bmagic |
lol |
| 11:41 |
|
Christineb joined #evergreen |
| 11:41 |
abneiman |
hahaha |
| 11:41 |
abneiman |
Bmagic++ csharp_++ # valiant attempts |
| 11:41 |
csharp_ |
pinesol: jackweed |
| 11:41 |
pinesol |
csharp_: is your coffee overgranularized? |
| 11:41 |
bshum |
pinesol: whoami |
| 11:41 |
pinesol |
bshum: bshum |
| 11:41 |
berick |
pinesol is 1 step away from I know you are but what am I? |
| 11:41 |
pinesol |
berick: Mr. Spock: Something fascinating just happened. |
| 11:41 |
csharp_ |
pinesol: blame pinesol |
| 11:41 |
pinesol |
csharp_: itself stole bshum's tux doll! |
| 11:41 |
bshum |
It knows who I am at least ;) |
| 11:42 |
csharp_ |
pinesol: whoami |
| 11:42 |
pinesol |
csharp_: I don't recognize you. You can message me either of these two commands: "user identify <username> <password>" to log in or "user register <username> <password>" to register. |
| 11:42 |
Bmagic |
I'm rubber and your glue.... |
| 11:42 |
csharp_ |
pinesol: I recognize you, mf |
| 11:42 |
pinesol |
csharp_: Try restarting apache. |
| 11:42 |
bshum |
@quote add <berick> pinesol is 1 step away from I know you are but what am I? |
| 11:42 |
pinesol |
bshum: The operation succeeded. Quote #251 added. |
| 11:42 |
csharp_ |
bshum: showoff! |
| 11:42 |
Bmagic |
did somone plug pinesol into deepseek? |
| 11:42 |
csharp_ |
@band add DeapSeak |
| 11:42 |
pinesol |
csharp_: Band 'DeapSeak' added to list |
| 11:43 |
csharp_ |
huh - it lets me do that |
| 11:43 |
bshum |
That's probably a less secure plugin I guess |
| 11:43 |
Bmagic |
computers suck |
| 11:43 |
csharp_ |
pinesol: especially YOU |
| 11:43 |
pinesol |
csharp_: GitHub is my copilot |
| 11:45 |
abneiman |
pinesol: I need that on a tshirt pls |
| 11:45 |
pinesol |
abneiman: Yeah, well, you know, that's just, like, your opinion, man. |
| 11:54 |
csharp_ |
pinesol: whoami |
| 11:54 |
pinesol |
csharp_: csharp |
| 11:54 |
csharp_ |
you know my REAL NAME? |
| 11:55 |
Dyrcona |
pinesol: whoami |
| 11:55 |
pinesol |
Dyrcona: Dyrcona |
| 11:57 |
|
redavis joined #evergreen |
| 12:00 |
|
jihpringle joined #evergreen |
| 12:27 |
mmorgan |
pinesol: whoru |
| 12:27 |
pinesol |
mmorgan: Try running autogen |
| 13:28 |
Dyrcona |
"Who are you? Who? Who? |
| 13:29 |
berick |
i really wanna know |
| 13:29 |
Dyrcona |
And, now, I have to play it. |
| 13:30 |
berick |
:) |
| 13:30 |
* Dyrcona |
increases the volume. |
| 13:32 |
Dyrcona |
Whoooooo arrree youoooo? |
| 13:45 |
jeffdavis |
I'm still having trouble with search slowness due to a bad query plan for asset.record_has_holdable_copy in 3.14. |
| 13:46 |
jeffdavis |
I poked a little bit at statistics and such, but I can't figure out how to get Postgres to use a better plan. |
| 13:50 |
jeffdavis |
This is the EXPLAIN output (for a slow query that's equivalent to asset.record_has_holdable_copy): https://explain.depesz.com/s/8xly |
| 13:53 |
jeffdavis |
If I run the same query in a 3.11 database on the same Postgres server, it's quite fast: https://explain.depesz.com/s/bUX8 |
| 13:54 |
Dyrcona |
Does not deleted need to be in there so much? |
| 13:54 |
Dyrcona |
I'm specifically looking at the bitmap heap scan on serial.unit. |
| 13:55 |
Dyrcona |
Same thing for asset.copy. I wonder if that's throwing the optimizer off? |
| 13:55 |
jeffdavis |
I think that's due to this index on asset.copy introduced in 3.12: CREATE INDEX cp_extant_by_circ_lib_idx ON asset.copy(circ_lib) WHERE deleted = FALSE OR deleted IS FALSE; |
| 13:56 |
jeffdavis |
combined with the use of the actor.org_unit_descendants function |
| 13:56 |
Dyrcona |
It should probably just be the latter. Try recreating the index without "deleted = FALSE OR". |
| 13:57 |
* Dyrcona |
tries something. |
| 13:58 |
Bmagic |
jeffdavis: have you VACUUM ANALYZE; ed recently? |
| 13:59 |
Dyrcona |
Well, "false = false" is true. I guess my hypotheses was proven incorrect. |
| 13:59 |
jeffdavis |
yep, there's been lots of vacuum analyze along the way :) |
| 13:59 |
Bmagic |
that's bitten me more than once, had to check |
| 14:00 |
jeffdavis |
JIT is disabled too (it wasn't when I started looking, turning it off didn't help though). |
| 14:00 |
Dyrcona |
Vacuum full analyze; -- When in doubt, defrag out. |
| 14:00 |
Dyrcona |
You restarted after turning off JIT? |
| 14:00 |
jeffdavis |
reloaded config, not a full restart |
| 14:01 |
Dyrcona |
I'd try a full restart. |
| 14:04 |
|
jihpringle joined #evergreen |
| 14:05 |
jeff |
I haven't looked at the explain output, but since you mentioned actor.org_unit_descendants(): would your plan benefit any if that function (whichever version is being used here -- there are at least two versions with different arguments) were marked as STABLE instead of VOLATILE? |
| 14:06 |
jeffdavis |
Changing it to STABLE didn't improve the performance. |
| 14:06 |
jeff |
ok, worth a try. sorry that didn't help! |
| 14:06 |
jeff |
I think that function might be safe to be marked as stable, but I may be incorrect there also. |
| 14:08 |
jeffdavis |
That's my understanding too - sounds like any function that does not modify data (i.e. is SELECT only) should be safe to mark as stable? |
| 14:09 |
Dyrcona |
jeffdavis: That's my understanding. |
| 14:10 |
jeffdavis |
Modifying that index to get rid of the redundant deleted condition does fix the problem, which is better than my previous workaround of omitting the deleted condition altogether. |
| 14:10 |
Dyrcona |
jeffdavis++ |
| 14:10 |
jeffdavis |
Dyrcona++ |
| 14:11 |
jeffdavis |
I still don't understand *why* it produces a bad query plan in some environments though. |
| 14:11 |
Dyrcona |
I think we might have done that and never open a Lp bug about it. |
| 14:11 |
* Dyrcona |
checks. |
| 14:12 |
Dyrcona |
Nope. Not in production anyway. |
| 14:12 |
Dyrcona |
jeffdavis: Did you mention the Pg release? |
| 14:12 |
Dyrcona |
Oh. I see the version didn't seem to matter. |
| 14:12 |
jeff |
I think (at least in the past) it was common to need two versions of a similar index, like deleted = FALSE and deleted IS FALSE. I don't know the reason for an index combining both. It might cause it to be used by neither style query, or it might depend on PostgreSQL version. |
| 14:12 |
jeffdavis |
oh, this is PG 14 |
| 14:14 |
Dyrcona |
Maybe it was the OR causing a heapscan sometimes? |
| 14:15 |
jeff |
I'm not looking at the index though, so I should stop taking stabs in the dark. :-) |
| 14:15 |
Dyrcona |
jeff: That sounds like the kind of thing one hears on the Internet. :) |
| 14:16 |
jeffdavis |
I see a random Stack Overflow comment from 2017 that says "Predicates in partial indexes' WHERE clause *must* match your queries' predicates in order to be used. WHERE ispublish, WHERE ispublish = TRUE and WHERE ispublish IS TRUE are 3 different predicates." |
| 14:18 |
Dyrcona |
"random StackOverflow comment" Who is the random commenter? |
| 14:18 |
|
jonadab joined #evergreen |
| 14:18 |
* Dyrcona |
has seen a lot of stuff on StackOverflow lately that he knows to be just plain wrong. (Not saying that this is wrong, but suspects it is out of date.) |
| 14:19 |
jeffdavis |
https://stackoverflow.com/questions/42972726/postgresql-create-index-for-boolean-column#comment73038823_42972726 |
| 14:19 |
Dyrcona |
@who is the random commenter. |
| 14:19 |
pinesol |
jeffdavis is the random commenter. |
| 14:19 |
Dyrcona |
Ha! |
| 14:19 |
jeffdavis |
oh god |
| 14:21 |
jeffdavis |
https://www.postgresql.org/docs/14/indexes-partial.html |
| 14:22 |
Dyrcona |
Ok. I buy it. It's not just "things you read on the Internet." |
| 14:22 |
Dyrcona |
"I'll take 'things you read on the Internet' for 400, Alex." |
| 14:31 |
jeffdavis |
specifically it says "The system can recognize simple inequality implications, for example “x < 1” implies “x < 2”; otherwise the predicate condition must exactly match part of the query's WHERE condition or the index will not be recognized as usable." |
| 14:31 |
jeffdavis |
...which, I don't know if deleted = false is a "simple inequality implication" or not |
| 14:33 |
csharp_ |
well, in the interest of Learning Something New™ I have gone a bit down the road of installing mailman3 from source |
| 14:33 |
csharp_ |
the Debian/Ubuntu packagers have *failed* miserably to create an installable package |
| 14:33 |
Dyrcona |
jeffdavis: Well, "deleted is false" and "deleted = false" *should* be treated the same. Doesn't mean that they are. |
| 14:34 |
csharp_ |
the bad news is that the lists are down until I've figured this out - and there's no way to announce that on the lists, heh |
| 14:34 |
Dyrcona |
IDK, though. There may be some semantic difference between "=" and "is" that's not obvious. |
| 14:35 |
Dyrcona |
csharp_++ ubuntu-- |
| 14:35 |
csharp_ |
I feel confident that I can get it going over the weekend - not sure if it will be this afternoon |
| 14:35 |
csharp_ |
kind of having fun using pip/python3 to do stuff |
| 14:36 |
csharp_ |
and the new version is 1) modular and 2) database-backed (using PG) |
| 14:36 |
csharp_ |
theoretically, mailman3 core is already running, but the lists aren't configured yet - there are methods to migrate |
| 15:34 |
|
mantis left #evergreen |
| 16:01 |
Dyrcona |
csharp__++ |
| 16:12 |
csharp_ |
oh! it works |
| 16:12 |
csharp_ |
I'm running around looking for stuff |
| 16:13 |
csharp_ |
the web stuff isn't working yet - untangling old apache stuff, trying to get Let's Encrypt working, etc. |
| 16:13 |
csharp_ |
but if mail can flow that's awesome |
| 16:16 |
Dyrcona |
csharp__++ |
| 16:17 |
csharp_ |
damn - I don't think I've breathed all day |
| 16:18 |
Dyrcona |
Well, take a breath. |
| 16:18 |
* csharp_ |
collapses in relief |
| 16:18 |
csharp_ |
I need to go play some guitar about this |
| 16:19 |
Dyrcona |
I think I hear either a Beatles or a Taylor Switch parody. |
| 16:19 |
Dyrcona |
"Yeah-eah, I'm the mail man." |
| 16:20 |
Dyrcona |
Heh.... "Switch...." |
| 16:23 |
csharp_ |
I slipped up and called her Taylor Smith recently and my daughter can't get over it |
| 16:23 |
abneiman |
csharp_: tell her to shake it off ;-) |
| 16:24 |
JBoyer |
Like a polaroid picture; really add some flavor to this artist mix. |
| 16:25 |
berick |
that would leave me with some Bad Blood |
| 16:26 |
berick |
(i do like that song) |
| 16:26 |
csharp_ |
abneiman++ |
| 16:27 |
csharp_ |
too... many... songs... now in my head |
| 16:35 |
pinesol |
News from commits: LP#2068740 Followup - Leave admin account password alone in concerto <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=7b18b3bd51fc72e1b4fba68391085c02d9b92bad> |
| 16:48 |
|
jihpringle joined #evergreen |
| 16:51 |
abneiman |
shake it off like a polaroid prooooobleeeeem |
| 16:56 |
berick |
Taylor Smith is crushing it |
| 17:04 |
abneiman |
Taylor 3000 |
| 17:11 |
|
mmorgan left #evergreen |
| 17:24 |
berick |
TayTay3K |
| 17:25 |
berick |
@band add The Plot Chickens |
| 17:25 |
pinesol |
berick: Band 'The Plot Chickens' added to list |
| 17:52 |
|
jihpringle joined #evergreen |
