Evergreen ILS Website

IRC log for #evergreen, 2022-03-16

| 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
01:54 dbs joined #evergreen
06:00 pinesol News from qatests: Testing Success <http://testing.evergreen-ils.org/~live>
07:15 rjackson_isl_hom joined #evergreen
07:48 rjackson_isl_hom joined #evergreen
08:01 rfrasur joined #evergreen
08:04 rjackson_isl_hom joined #evergreen
08:10 collum joined #evergreen
08:34 mmorgan joined #evergreen
08:35 Dyrcona joined #evergreen
08:50 Dyrcona joined #evergreen
08:59 jvwoolf joined #evergreen
09:01 terranm joined #evergreen
09:03 mantis1 joined #evergreen
09:04 mmorgan Bmagic: Re: the dropdown and progress bar difference on the pull list - those are in 3.8 with the angular pull list. The system that shows the strangeness is 3.7.2.
09:09 terranm joined #evergreen
09:11 Keith-isl joined #evergreen
09:17 rjackson_isl_hom joined #evergreen
09:35 Dyrcona So, we want to sort parts when placing a hold in the Angular staff catalog similar to how they are sorted in the TTOPAC and BOOPAC when placing a hold. Right now, it's looking like the "best" place to do that is in the open-ils.circ.hold.get_metadata backend call.
09:46 JBoyer joined #evergreen
09:48 miker Dyrcona: seems reasonable, I can't imagine the order coming out of that is being depended upon today
09:53 Dyrcona miker: Yeah, I was about to say that the call used in the OPAC sorts the holds in the backend.
09:54 Dyrcona I'll try it on our training server, then Lp it, and push a working branch.
10:05 Bmagic This bug continues. The query that Evergreen creates still can create duplicate rows. This time due to multiple hold notes. Consider this query: https://pastebin.com/rFfzJa5D
10:07 Bmagic If a hold has more than one note, the query returns more than one copy of the hold for the pull list. Which breaks pagination. So, I'm thinking that the pull list code needs to DISTINCT or GROUP BY. However, there are 49 columns there, that would be a "fun" GROUP BY clause. I've gone ahead and tested the query with DISTINCT and again with GROUP BY
10:08 Bmagic There is an error, because the ORDER BY  clause is ordering by the shelving location, which isn't in the column selection. And that's not been an issue apparently. But it is an issue as soon as you want PG to dedupe the rows
10:09 Bmagic My question is: can I just edit the Evergreen code so that it does ask PG for the shelving location column?
10:15 Bmagic ahopl
10:26 mmorgan Bmagic: Shelving locaton is vital for the Holds Pull List. Our libraries routinely sort the pull list by Shelving Location, cn prefix, cn label, cn suffix. Sorting in this manner works in current production with no issues.
10:26 miker Bmagic: the request note data is never returned by that query. I think we can probably drop the note from the join list, but checking
10:27 Bmagic sure, I'm not wanting to change that behavior. It's the query that has the bug. And the ORDER BY should remain intact. But it's ordering by a column that it's not asking for. Therefore, the ORDER BY clause produces an error when you want to dedupe the results
10:29 * mmorgan is unsure of where the data on the Holds Pull List in the client actally comes from. Is it the ahopl query in the fm_IDL.xml, or has that been superceded somewhere else in the code?
10:30 Bmagic miker: that would fix it for the duplicate hold note issue, but I was thinking that this query could have the bug if other joined tables had many-to-one in the future. My conclusion was that this query needs to be GROUPed or DISTINCTed?
10:31 Bmagic mmorgan: I think is generated by grd.components.ts->HoldsGridComponent->fetchHolds but I'm not sure
10:33 mmorgan Bmagic: So that would be in the 3.8 angularized version, right?
10:33 miker Bmagic: IMO, if it's asking for things that it doesn't need, and that ask is causing a problem, we need it to stop asking
10:33 Bmagic I'm looking at 3.7 code
10:34 Bmagic miker: fair enough, I think* this code is multi-purpose for the pull list and the patron hold interface? And maybe others?
10:35 miker ahopl? no, that's only used for the pull list
10:36 Bmagic yeah, for sure... I'm just looking at all the JS code around it
10:46 Dyrcona Oof. Haven't built master for a while. I have to update Node.js.
10:47 miker so, the pull list already "fixed" (by using the wide hold api) in 3.8 (just to make sure we're on the same page)
10:48 Bmagic I see, Evergreen doesn't call back to the IDL ahopl after 3.8?
10:49 Bmagic I was just reading through grid.component.ts around line 263... which performs a request to open-ils.circ.hold.wide_hash.stream (EG 3_7)
10:49 mmorgan miker: wide hold api isn't used until 3.8?
10:49 miker Bmagic: well, I'm not saying that. I'm saying that the angular pull list uses the wide hold api
10:50 miker Bmagic: ah, well, then 3.7 uses that as well. but that's not the query you pasted
10:50 Bmagic I get offtrack greping through the code. So where is the appropriate place I need to be looking that generates this query?
10:50 Dyrcona BTW: When I updated packages, I got a ncurses window with a lot of text about Pg 9.6 being obsolete.
10:51 mmorgan So back to my question - has the ahopl in the fm_IDL.xml been entirely superceded (at least in 3.7+)
10:51 mmorgan The sql, that is.
10:52 miker mmorgan: there are 2 mentions of it in master, both in older pull list UIs. I think you can get to those, but they're not the ones offered by menus IIUC
10:52 miker so "sorta"? :)
10:53 mmorgan Ok, thanks! Sorta is good enough for me!
10:57 miker one is the dojo-based OLD "printable pull list", and the other is AngJS pull list. both are made hard (or impossible) to get to in 3.8, AFAICT
11:02 Dyrcona I'm trying to build latest master after doing a git clean -x -f -d and updating prerequisites.
11:03 Dyrcona npm install hangs for me at extract:rxjs on Ubuntu Focal.
11:03 Dyrcona Oh, that's in Open-ILS/src/eg2/
11:05 berick Dyrcona: rxjs hanging is a known issue.  just have to wait it out :\
11:05 miker Bmagic: what's the URL path for the UI you're getting that query from, because I don't think it's /eg2/en-US/staff/circ/holds/pull-list ... that's not involving ahopl AFAICT, either in the main interface or the print link. but that's where grid.component.ts with the fetchHolds you mention is used.
11:06 berick haven't looked recently for a fix, could be one
11:08 Dyrcona breick: Ok. Thanks....
11:09 Dyrcona grr. berick ^^
11:11 Dyrcona Well, that puts the kibosh  on the testing I was about to do.
11:15 csharp_ @decide breick or berick
11:15 pinesol csharp_: go with berick
11:15 csharp_ pinesol: good bot
11:15 pinesol csharp_: I just wanted to let you know that you mean a lot to me
11:15 mmorgan Awww.
11:20 terranm lol
11:29 mmorgan miker: Don't want to speak for Bmagic, but I'm seeing the URL /eg/staff/circ/holds/pull in 3.7.2
11:57 Keith-isl joined #evergreen
12:04 jihpringle joined #evergreen
12:34 collum joined #evergreen
13:02 collum joined #evergreen
13:19 pinesol News from commits: LP1916600 Add Create Date as a column in Holdings View <https://git.evergreen-ils.org/?p=E​vergreen.git;a=commitdiff;h=dad5e6​f01f18f4eb7dff0b088eb9b6cb5922c655>
14:17 Bmagic miker: eg/staff/circ/holds/pull
14:18 Bmagic miker: like I was saying, I got off track using grep. I knew it didn't look quite like the code that could produce that query
14:19 Bmagic js/ui/defaul/staff/circ/holds/app.js is more like it?
14:46 berick grabbing 1312
14:50 pinesol News from commits: LP1960956 Stamping DB upgrade / usr message index <https://git.evergreen-ils.org/?p=E​vergreen.git;a=commitdiff;h=af5a40​94349360083e9ab116693d013d2ec78c9f>
14:50 pinesol News from commits: LP#1960956: Fix slow user merges/deletes <https://git.evergreen-ils.org/?p=E​vergreen.git;a=commitdiff;h=951a51​dcd395dbea0bb3918641d82646b4812919>
14:51 terranm New Devs Working Group is starting in a few minutes - login info at https://wiki.evergreen-ils.org/doku.php​?id=newdevs:meetings:agenda-2022-03-16
14:58 JBoyer I knew that looked familiar. Bmagic, the ahopl base query is defined in fm_IDL.xml. There's no ORDER BY in there so it's possible it's altered where used. But that's where the ausp join is coming from.
15:20 pinesol News from commits: LP1843970: Refactor Coded Value Map grid to use eg-admin-page <https://git.evergreen-ils.org/?p=E​vergreen.git;a=commitdiff;h=ff3db8​2ca0022234d46cc9c8f1cd1e91862f7696>
15:20 pinesol News from commits: LP1838580: Holds Shelf detail view record summary doesn't populate <https://git.evergreen-ils.org/?p=E​vergreen.git;a=commitdiff;h=224814​0fe3af54901d8702f0ee5974e0f6c0cb22>
15:59 Guest87 joined #evergreen
16:00 Bmagic JBoyer: right, I submitted a patch to that query (in the IDL) - that produces duplicate rows on it's own when patrons have more then one non-CAPTURE-blocking penatly. But now, I've encountered the same bug but for a different reason. That query gets wrapped inside of another one with all those LEFT JOINS
16:02 JBoyer Ah, ok. I may not have read close enough to tell where you were running into trouble.
16:14 berick grabbing 1313
16:20 pinesol News from commits: LP1956003 Stamping DB upgrade / hold group grids <https://git.evergreen-ils.org/?p=E​vergreen.git;a=commitdiff;h=91fa2e​e5f2ad75a5ad480d92c74fe91f3ddfbbeb>
16:20 pinesol News from commits: LP1956003 Hold Group Workstation Settings to Server Settings. <https://git.evergreen-ils.org/?p=E​vergreen.git;a=commitdiff;h=4d667a​a09ff61496c9f60a0cfae34c13566a5a61>
16:20 pinesol News from commits: LP1960956 Stamping DB upgrade / usr message index (fix) <https://git.evergreen-ils.org/?p=E​vergreen.git;a=commitdiff;h=64803f​f1fbca0fc24d649eb10b31cf7f26a19019>
16:20 pinesol News from commits: LP1922975 Build scripts python3 minor fixes <https://git.evergreen-ils.org/?p=E​vergreen.git;a=commitdiff;h=9ab97e​b8d80d4f1d5852ea30ee64dbffc32766e1>
16:20 pinesol News from commits: LP#1922975: install python3 dependencies, rather than python2 <https://git.evergreen-ils.org/?p=E​vergreen.git;a=commitdiff;h=e4a07d​8f825c986774a11bf52a7d20a23062163e>
16:20 pinesol News from commits: LP#1922975: update i18n scripts for Python 3 <https://git.evergreen-ils.org/?p=E​vergreen.git;a=commitdiff;h=b49185​fd713b3ea27a023cf30f0b1fc4d1940f71>
16:50 pinesol News from commits: LP1846552 Shelving location Order handle new locations <https://git.evergreen-ils.org/?p=E​vergreen.git;a=commitdiff;h=48771a​f5ae7d6292eef763f9a5677acc93ecacc9>
16:50 pinesol News from commits: LP1846552 Shelving Location Order Angular UI <https://git.evergreen-ils.org/?p=E​vergreen.git;a=commitdiff;h=651665​9d05dc80d3169462ee91a62e5cfd134292>
16:50 pinesol News from commits: LP1965161: Sort Monograph Parts in Holds Metadata <https://git.evergreen-ils.org/?p=E​vergreen.git;a=commitdiff;h=f61f1e​147b1d4ed288554c88ff72d040c924a0f1>
17:17 mmorgan left #evergreen
17:20 pinesol News from commits: LP1847827 - Evergreen Web Based Self Check - Use prefered first name in header. <https://git.evergreen-ils.org/?p=E​vergreen.git;a=commitdiff;h=fddcef​b2fb0db58ff5357f205034bc62f20dbac8>
17:25 jvwoolf left #evergreen
18:00 pinesol News from qatests: Testing Success <http://testing.evergreen-ils.org/~live>
18:20 pinesol News from commits: LP1863196-Add series title to holds pull list. <https://git.evergreen-ils.org/?p=E​vergreen.git;a=commitdiff;h=93646c​82129e6399251f7c9d795877cf86b0bb55>
22:28 JBoyer joined #evergreen

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