| Time |
Nick |
Message |
| 07:28 |
|
BDorsey joined #evergreen |
| 07:38 |
|
rhamby joined #evergreen |
| 07:39 |
|
akilsdonk joined #evergreen |
| 07:39 |
|
phasefx joined #evergreen |
| 07:41 |
|
mantis1 joined #evergreen |
| 08:35 |
|
mmorgan joined #evergreen |
| 08:46 |
|
dguarrac joined #evergreen |
| 08:48 |
|
kworstell-isl joined #evergreen |
| 09:03 |
|
Dyrcona joined #evergreen |
| 09:09 |
pinesol |
News from commits: LP1999282 Less intense badges for staff interface <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=39c44d2c0a6f5eab4d93b564b079be1cf75d85f2> |
| 09:56 |
Dyrcona |
BTW, if anyone wants to try development and testing on Pg 15 before Lp 1997098 and Lp 1999274 go in, I pushed user/dyrcona/pg15-master-for-dev to the working repo that combines both of those branches into master. I'm using it on my local development VMs with concerto. |
| 09:56 |
pinesol |
Launchpad bug 1997098 in Evergreen "Add Support for PostgreSQL 15" [Medium,Confirmed] https://launchpad.net/bugs/1997098 |
| 09:56 |
pinesol |
Launchpad bug 1999274 in Evergreen 3.10 "Performance of Search on PostgreSQL Versions 12+" [Medium,Confirmed] https://launchpad.net/bugs/1999274 |
| 09:57 |
Bmagic |
Dyrcona++ |
| 09:58 |
Bmagic |
skipping to 15 is the recommended Evergreen path at this point? |
| 10:01 |
Dyrcona |
Bmagic: Not at this time. |
| 10:02 |
Dyrcona |
We still have to decided what version to recommend after those patches go in. (I would recommend 15, but others will have other opinions.) |
| 10:03 |
Dyrcona |
I don't know of any site running greater than Pg 10 in production. |
| 10:03 |
Bmagic |
right |
| 10:13 |
|
kworstell-isl joined #evergreen |
| 10:20 |
Dyrcona |
Couple of database live tests failed. I'd better have a look. |
| 10:36 |
|
Stompro joined #evergreen |
| 10:41 |
|
BDorsey joined #evergreen |
| 10:46 |
|
Christineb joined #evergreen |
| 11:09 |
Dyrcona |
Pro tip: Try to remember which port is which version of PostgreSQL, or you could waste an hour and a half trying to figure out whey pgtap live tests are failing on 1 server but not the others. |
| 11:09 |
Dyrcona |
s/whey/why/ |
| 11:09 |
pinesol |
News from commits: LP#1999274: Add Release Note <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=31e2d9a565a603f0044ce52ff1719e9eaa151dbd> |
| 11:09 |
pinesol |
News from commits: LP#1999274: Improve Search Performance on Pg 12+ <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=aa30c7e7b23a6dc4f10f5f1e7382cba13d787cf5> |
| 11:09 |
Dyrcona |
But, for those who don't like cottage cheese, the result is about the same. |
| 11:11 |
Dyrcona |
I should have realized that something was up when it said that the pgtap extension already existed after I just did eg_db_config with --create-database.... |
| 11:11 |
Dyrcona |
"I'd like to buy a clue." |
| 11:14 |
* Dyrcona |
didn't expect that to get pushed today.... I was hoping for next week, but I'll take it. :) |
| 11:14 |
csharp_ |
Bmagic: we're still on 10, ready to move to 11 - can't believe I hadn't thought of a Spinal Tap joke about that before now |
| 11:15 |
Dyrcona |
Hah! |
| 11:15 |
Bmagic |
csharp_++ |
| 11:15 |
Dyrcona |
csharp_++ |
| 11:16 |
* Dyrcona |
would like to go to 15, but I'd recommend at least 13 or 14 because 11 doesn't offer much improvement over 10 in performance. |
| 11:16 |
abowling |
I haven't had time to really dig into the issue, but the 3.10.0 seems to be incomplete. When performing a clean install from scratch, quite a few of the JS files are missing. |
| 11:17 |
abowling |
This probably merits a launchpad bug, but I haven't yet had time to fully research what's going on. |
| 11:17 |
Dyrcona |
abowling: I haven't run into that. |
| 11:17 |
csharp_ |
abowling: you did npm/ng builds for stuff for Ang/AngJS before make install? |
| 11:17 |
abowling |
FWIW, JBoyer was a witness with me |
| 11:18 |
abowling |
csharp_: no, straight install from the tarball |
| 11:18 |
csharp_ |
ah |
| 11:18 |
Dyrcona |
So, tarball is busted. |
| 11:18 |
abowling |
Dyrcona: yes, that's our operating theory |
| 11:18 |
* Dyrcona |
doesn't use the tarballs, except for testing a release before release. |
| 11:18 |
abowling |
Git build likely works |
| 11:18 |
abowling |
tarball doesn't |
| 11:19 |
Dyrcona |
So that's bug worthy: broken tarball. |
| 11:19 |
abowling |
and i'm not 100% sure of this, but like i said, I wasn't the only one to witness ;) |
| 11:19 |
Dyrcona |
That's all supposed to be automated, but something likely went wrong with one of the npm steps. |
| 11:20 |
abowling |
Dyrcona: agreed; I'm hoping someone can test and disprove my theory...or confirm it |
| 11:20 |
csharp_ |
confirmed: no eg2 directory in the 3.10.0 tarball |
| 11:20 |
Dyrcona |
csharp_++ |
| 11:20 |
abowling |
csharp_: there's a ton missing...seemingly |
| 11:21 |
abowling |
for example, .../web/js/ui/default/build/common doesn't get created, for one |
| 11:21 |
csharp_ |
no node_modules in the AngJS dir either |
| 11:22 |
csharp_ |
abowling: jeez - you've had some back luck |
| 11:22 |
csharp_ |
don't install PG 13 for god's sake! |
| 11:22 |
csharp_ |
hey, PG-13 - lots of punny Postgres stuff flying around today |
| 11:22 |
abowling |
csharp_: yep, but all tied together once we figured it out. |
| 11:22 |
csharp_ |
or maybe I'm getting punchy - fasting today for Ash Wednesday |
| 11:23 |
Dyrcona |
:) |
| 11:23 |
abowling |
installing 3.9.1 THEN 3.10.0 seemed to take care of a lot of it |
| 11:23 |
Dyrcona |
I solved my issues with ports and testing by doing apt purge postgresql-14. :) |
| 11:23 |
abowling |
we jumped from 3.7.1 to 3.10.0 |
| 11:24 |
Dyrcona |
abowling: We were going to do the same, but decided to wait until more 3.10 bugs are fixed. |
| 11:24 |
abowling |
Dyrcona: you decided to wait because you're smarter than I am |
| 11:25 |
csharp_ |
@iqtest abowling vs. Dyrcona |
| 11:25 |
pinesol |
csharp_: Go away, or I'll replace you with a very small shell script! |
| 11:25 |
csharp_ |
pinesol: I know you couldn't do it |
| 11:25 |
pinesol |
csharp_: zarro boogs found |
| 11:25 |
* abowling |
forfeits preemptively |
| 11:26 |
Bmagic |
you broke my heart Alfredo |
| 11:26 |
Dyrcona |
:) |
| 11:26 |
Dyrcona |
Well, it was not entirely my decision. We have a team here that made the decision. |
| 11:28 |
Dyrcona |
We do plan to backport a few things in April: https://git.evergreen-ils.org/?p=working/Evergreen.git;a=shortlog;h=refs/heads/user/dyrcona/cwmars_custom_rel_3_7_4 |
| 11:29 |
csharp_ |
@blame 22 Bmagic |
| 11:29 |
pinesol |
csharp_: I know it was you, Bmagic. You broke csharp_'s heart. You broke csharp_'s heart. |
| 11:30 |
Bmagic |
lol |
| 11:30 |
Bmagic |
nice pull |
| 11:52 |
* Dyrcona |
doesn't remember installing Pg 14 on that particular VM and doesn't know why it was there, but now it is gone. |
| 12:18 |
|
jihpringle joined #evergreen |
| 12:29 |
|
collum joined #evergreen |
| 12:32 |
abowling |
another: Can't use an undefined value as an ARRAY reference at /usr/local/share/perl/5.30.0/OpenILS/Application/Actor.pm line 3315 |
| 12:33 |
abowling |
Can't use an undefined value as an ARRAY reference at /usr/local/share/perl/5.30.0/OpenILS/Application/Actor.pm line 3315 |
| 12:33 |
abowling |
open-ils.actor.patron.search.advanced.fleshed : 500 : *** Call to [open-ils.actor.patron.search.advanced.fleshed] failed for session |
| 12:36 |
Bmagic |
fm_IDL mismatch? |
| 12:52 |
Stompro |
NCIPserver++ drycona++ I'm looking at lib/NCIP/ILS/Evergreen.pm , very well commented and easy to understand. |
| 12:56 |
Dyrcona |
Stompro: Thank you! I'd like to drop the requirement for Plack and add it to Evergreen as a mod_perl module some day. |
| 12:58 |
Bmagic |
Stompro++ # reading |
| 13:00 |
Stompro |
Since we are only using acceptitem I'm wondering if I can skip setting up a dedicated system and branch org unit? We have an ILL org unit already. |
| 13:03 |
Stompro |
I'll give it a try and see if it works. |
| 13:06 |
abowling |
BMagic: it's not getting to an osrfsys.log error, so maybe |
| 13:06 |
|
kworstell_isl joined #evergreen |
| 13:16 |
abowling |
BMagic++ |
| 13:16 |
Bmagic |
oh? It was fm_IDL? |
| 13:16 |
abowling |
BMagic: the view actor.usr_message_penalty didn't get created in the 3.7.1-3.8.0 SQL upgrade |
| 13:16 |
abowling |
so, kinda |
| 13:16 |
Bmagic |
nice |
| 13:17 |
Bmagic |
abowling++ |
| 13:18 |
|
dmoore joined #evergreen |
| 13:24 |
|
kworstell_isl joined #evergreen |
| 13:40 |
|
jihpringle joined #evergreen |
| 14:02 |
Dyrcona |
That sounds like a potential bug in the upgrade scripts. |
| 15:06 |
csharp_ |
sometimes it's "we already have that update, DROP DEAD!" |
| 15:06 |
csharp_ |
which has always been annoying |
| 15:08 |
Dyrcona |
Well, that's why i made this: https://gist.github.com/Dyrcona/00bd6b6290b6fbbb579c7f93b360ab0d |
| 15:09 |
Dyrcona |
I think we should split the tarball script up into smaller scripts, and have the main one run the others. |
| 15:19 |
Stompro |
Dyrcona++ JBoyer++ I just placed my first ncip request, so exciting. |
| 15:19 |
JBoyer |
Stompro++ good to hear it's already working! |
| 15:21 |
Stompro |
That sample request was very very helpful. |
| 15:21 |
JBoyer |
yeah, I keep a few of the samples in my back pocket because logging isn't great, as Dyrcona has mentioned on LP. |
| 15:22 |
Stompro |
I think I'll be able to add in a few things like setting a circ_mod and adding a default price to the item. And maybe some extra bib info, we have a default bibicon for our ILL items. |
| 15:23 |
Dyrcona |
Stompro++ |
| 15:24 |
Dyrcona |
The NCIPServer project was originally envisioned to work with different ILL products and to be mostly the same between Koha and Evergreen. The latter disappeared in a hurry, but you can make your modifications in a copy of the current Evergreen driver and submit them in a working branch if you like. |
| 15:25 |
Dyrcona |
The idea is that there could be different Evergreen drivers for different ILL vendors' products. |
| 15:25 |
JBoyer |
I also put a db script together recently (seems like everyone is getting in on the fun.): https://gist.github.com/HitScan/2655f0486ca84ecf658131e2999f2845 |
| 15:26 |
JBoyer |
I had envisioned a step where it also checks the db in order to build the list of scripts to run but gave up on that and decided to make it check each upgrade number before trying to run it instead. |
| 15:26 |
Stompro |
Makes sense, I'll look at adding oils_ncip.xml config options for things I want to add. |
| 15:27 |
Dyrcona |
JBoyer++ I see yours applies them. I was looking through my MVLC scripts recently for something else and stumbled past a Perl script I had that did that. |
| 15:27 |
JBoyer |
run it anywhere inside an Eg repo, but the resulting script currently expects to be run in Open-ILS/src/sql/Pg/upgrades |
| 15:27 |
JBoyer |
Not quite, it just spits out a script that you run later. |
| 15:28 |
JBoyer |
writing a script that writes scripts is a fun exercise in quoting |
| 15:29 |
Dyrcona |
Oh! I should have looked more closely. |
| 15:30 |
Dyrcona |
I write programs that write programs all the time. It's faster than using macros to edit a 10,000 line csv into SQL. |
| 15:30 |
Dyrcona |
Well, sometimes it's faster. It's certainly more interesting. :) |
| 15:31 |
Dyrcona |
When I saw the psql, I thought you were running the upgrades but I see it does what you said above about checking for the upgrades in the database. |
| 16:10 |
pinesol |
News from commits: LP2006513 Angular login redirect repair <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=922a71d07d270c6c29c506d40eed6c276ba43dd9> |
| 16:22 |
|
mantis1 joined #evergreen |
| 17:04 |
|
mmorgan left #evergreen |
