Evergreen ILS Website

IRC log for #evergreen, 2019-12-30

| Channels | #evergreen index | Today | | Search | Google Search | Plain-Text | summary | Join Webchat

All times shown according to the server's local time.

Time Nick Message
06:01 pinesol News from qatests: Testing Success <http://testing.evergreen-ils.org/~live>
07:07 rjackson_isl joined #evergreen
08:07 stephengwills joined #evergreen
08:22 stephengwills joined #evergreen
08:34 mantis1 joined #evergreen
08:34 mmorgan joined #evergreen
08:54 Dyrcona joined #evergreen
09:43 jvwoolf joined #evergreen
09:53 sandbergja joined #evergreen
09:56 stephengwills joined #evergreen
10:07 Dyrcona JBoyer: Dunno if you're really around today, but I've discovered something about my 92 collected auto-renewal notices from last Thursday. They're all for the same user!
10:08 JBoyer This is the 1 day I am around this week. :)
10:08 JBoyer And yeah, that does sound promising.
10:08 JBoyer Anything weird about the user account?
10:10 Dyrcona Well, they had 92 items checked out that had no renewals left.... :)
10:11 JBoyer Seems perfectly normal to me, heh.
10:11 Dyrcona :)
10:11 Dyrcona They also don't have an email address, so they wouldn't have gotten the notices anyway.
10:13 Dyrcona This patron also has 2x as many notices as the next patron, who has 46. I wonder if it's a timeout issue or......max stanza size?
10:13 * Dyrcona checks max stanza size on this utility server.
10:14 Dyrcona max_stanza_size is set to 2,000,000
10:16 Dyrcona I suppose that I could dump the template output from the next patron and assume that this one would be double....
10:18 Dyrcona This user also tops the list of courtesy collected courtesy notices from 2 days before.
10:20 pinesol [evergreen|Bill Erickson] LP1857350 Org selector sorts by display value - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=1bed2a7>
10:20 pinesol [evergreen|Galen Charlton] LP#1857350: (follow-up) fix lint issue - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=0f46da2>
10:20 Dyrcona I have a lot of them that failed with the 2-day courtesy notices.
10:21 Dyrcona Well, more users and failed notices, that is.
10:21 lstratton joined #evergreen
10:31 JBoyer I don't suppose those circ ids show up anywhere in your logs since their renewals failed?
10:31 JBoyer I know you said you didn't think they were complete, but it's really weird that a bunch of failed renewals would knock out an AT runner.
10:34 Dyrcona Well, he auto-renew events (124) are all OK. It's the notices. There was also a problem with the 2-day courtesy notices, 2 days before. I JabberClient is not connected messages at both times, so I'm operating on the assumption that this user's email notices break something.
10:35 Dyrcona The 124 is the event id. There are only 92 of them.
10:36 Dyrcona Did anyone come up with a filter for these that would exclude patrons with no email on file? I seem to recall mmorgan asking about/working on that.
10:38 JBoyer Not that I know of. Would be helpful though. I can't recall if this is the sort of thing that could easily be done in action_trigger_filter.json or if it would need to be a new validator.
10:39 Dyrcona Yeah, I'm not sure, either. I'd have to check the environment, but I think it could.
10:40 Dyrcona ...be done in the filter.
10:41 * mmorgan looks for notes about filters to exclude patrons without email
10:43 stephengwills joined #evergreen
10:43 Dyrcona So, the email template output for the patron with 46 items is only 9,769 bytes, and does not have an email address, either. It's not the lack of an email address and it's not the size of the template output.
10:44 Dyrcona It could still be the amount of data, depending on what is fleshed, or it could be a timeout.
10:45 Dyrcona I suppose I could look at the trigger code to see if chunking/bundling is present and if not, add it.
10:45 Dyrcona mmorgan: Thanks! I thought you had done it.
11:32 khuckins joined #evergreen
11:42 mmorgan Dyrcona: I located my notes, but realize I never actually got it into production :-( Might have a go at it this afternoon.
11:42 alynn26 joined #evergreen
12:12 Dyrcona mmorgan: OK. Thanks, anyway!
12:23 Christineb joined #evergreen
13:05 Dyrcona Bleh... CSV is too big for LibreOffice.
13:59 Dyrcona I thought I had detected a pattern, but it doesn't really hold up.
14:00 Dyrcona I have noticed that old_due_date appears to not get poplulated, even for successful autorenewals.
14:01 Dyrcona Most of the autorenewal events that get stuck in collected state, seem to start with 1 patron who has a pile of things that have no renewals remaining, like 50+.
14:02 Dyrcona Then a bunch that are started after remain collected.
14:05 mmorgan Dyrcona: could there be some piece of data from item info that's breaking it? I'm always suspicious about punctuation causing problems.
14:08 Dyrcona mmorgan: Maybe.
14:10 Dyrcona But, I've got one user whose 85 items renewed and got a status of complete for the notice, then got a collected status 3 weeks later when there were no renewals remaining. I'll check if they're the same copies.
14:11 mmorgan Dyrcona: I only see old_due_date populated when the autorenewal is unsuccessful.
14:12 Dyrcona I don't think I've seen it poplulated, yet, but I'm only looking at the events of the users who had the most things to renew. I made a CSV and sorted it by a count.
14:14 Dyrcona Yeah, the renewal notices that succeeded are for the same copies where it didn't succeed when there were no renewals remaining.
14:18 lbarry joined #evergreen
14:22 Dyrcona The autorenewal events work. It seems to be the notices that get hung up.
14:26 mmorgan Did you say the old_due_date doesn't get populated for these failed autorenewals?
14:28 Dyrcona I have not seen an autorenewal notice event successful or not where old_due_date is anything other than "" in the user_data.
14:29 dbs Belated update to my Firefox / MARC Batch Edit struggles: tried running Firefox on Linux with a clean profile and didn't encounter the same problem (of the au.id being replaced with the aou.id and thus not loading a list of containers)
14:29 dbs So it's Ublock Origin, or Cookie Autodelete, or Zotero, or my browser privacy settings, or ... something
14:31 Dyrcona Forget it, dbs. It's Firefox....
14:32 * dbs ignores the provocation
14:32 Dyrcona It's a famous movie "quote."
14:35 Dyrcona Have you disabled Ublock Origin? I use it, too.
14:35 Dyrcona I use it in Chromium, too, and haven't had any trouble with Evergreen yet.
14:37 Dyrcona Guess I'll just have to set a new checkout limit of 50. :)
14:45 dbs Well, I just installed the 8 different add-ons I use in Firefox into a clean FF profile and can't trigger the error
14:46 dbs (Bitwarden, Cookie Autodelete, EZProxy Redirect Foxified, Firefox Multi-Account Containers, GNOME Shell integration, Grammalecte[fr], uBlock Origin, and Zotero Connector)
14:48 Dyrcona Of those, I use GNOME Shell Integration and uBlock Origin.
14:50 dbs was inspecting the stored user preferences under Administration -> Workstation but I don't really see how that would make a difference
14:51 dbs ah well, hopefully it's just something weird & truly unique with this workstation
14:53 khuckins joined #evergreen
15:18 dbs oh man. so I'm using Container Tabs in Firefox, and had been testing in a Work container tab
15:18 dbs for "fun", I created a new container tab called "Testing" and used that to register a new workstation and test MARC Batch Edit
15:19 dbs and of course it shows all of my record buckets (also called containers, just to confuse things further with overlapping terminology)
15:21 dbs so... somehow the workstation inside one particular category of container tabs gets that XHR response wrong unless I use network tools to disable caching
15:21 dbs and I can't reproduce it even within the same browser (as Firefox containers separate cookies & localstorage, etc, from other containers)
15:22 Dyrcona So, don't do that?
15:24 jeff i remain intrigued. it sounds like there isn't a reliable way to reproduce the issue?
15:25 jeff but the nature of the oddness is... puzzling.
15:25 jeff it did give me cause to read up on the "raced" bit in firefox dev tools.
15:29 mantis1 left #evergreen
15:37 dbs Dyrcona: the problem is, I don't know what "that" is that caused the problem in the first place
15:38 Dyrcona OK. Fair enough. I was thinking maybe don't use the container tabs.
15:38 dbs jeff: yeah, I've tried matching up as many configuration bits as I can and can't reproduce it outside of this one container
16:21 sandbergja joined #evergreen
16:47 pinesol News from qatests: Failed Installing OpenSRF pre-requisites <http://testing.evergreen-ils.org/~live//arch​ive/2019-12/2019-12-30_16:00:03/test.7.html>
17:06 mmorgan left #evergreen
17:15 pinesol [evergreen|Mike Risher] lp1855931 wrap text for wide Angular eg-grid column headers - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=fe8a465>
17:15 pinesol [evergreen|Galen Charlton] LP#1855931: (follow-up) make grid filter control cells wrap as well - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=9077cbc>
17:17 berick joined #evergreen
17:48 Bmagic for small Evergreen patches that introduce a new row into a table, is a pgtap test required? To test that the row exists?
17:49 Bmagic (Example: AutoRenew patch did not have a pgTAP to test AT hook row)
17:51 pinesol [evergreen|lfloyd] DOCS: LP 1767378 Work Log documentation - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=8296e8c>
17:51 pinesol [evergreen|lfloyd] Docs: fixed a spacing issue - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=593a93a>
17:51 pinesol [evergreen|Jane Sandberg] Docs: LP1767378 follow up: adding manual anchor - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=7ac78d5>
17:51 pinesol [evergreen|Jane Sandberg] Docs: Fixing asciidoc syntax so fop doesn't complain about staff client admin manual - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=2cd4053>
18:07 Dyrcona Bmagic: Well, obviously, we don't require a test for that, but adding a test would be highly encouraged, at least by me.
18:08 Dyrcona Retroactively adding tests for the auto-renew feature would make a decent, bite sized bug. Perhaps I'll open one tomorrow.
18:16 Dyrcona Anyway, I should have signed out a while ago. Catch you all again, tomorrow!
18:41 abowling1 joined #evergreen
18:59 abowling joined #evergreen
19:26 sandbergja joined #evergreen
20:13 sandbergja joined #evergreen
21:57 sandbergja joined #evergreen

| Channels | #evergreen index | Today | | Search | Google Search | Plain-Text | summary | Join Webchat