| Time |
Nick |
Message |
| 06:00 |
pinesol |
News from qatests: Testing Success <http://testing.evergreen-ils.org/~live> |
| 07:11 |
|
Dyrcona joined #evergreen |
| 07:20 |
|
rjackson_isl_hom joined #evergreen |
| 07:27 |
|
mantis joined #evergreen |
| 07:39 |
|
dickreckard joined #evergreen |
| 08:17 |
|
mmorgan joined #evergreen |
| 08:18 |
|
rfrasur joined #evergreen |
| 08:34 |
|
Dyrcona joined #evergreen |
| 08:43 |
|
dguarrac joined #evergreen |
| 09:00 |
|
alynn26 joined #evergreen |
| 09:09 |
miker |
berick: when you have a sec ... is there any reason the angular staff cat wouldn't show some facets? like, it doesn't restict to some hard-coded subset, or require more than cmf.facet_field be true or something, right? |
| 09:32 |
|
Keith-isl joined #evergreen |
| 09:37 |
|
dickreckard joined #evergreen |
| 09:39 |
|
ShulaBee joined #evergreen |
| 09:46 |
mantis |
I was following mmorgan's slides for creating a test environment via VB using Ansibile but when trying to fire it up, there is an ejabberd error. |
| 09:47 |
mantis |
"Unable to restart service ejabberd: Job for ejabberd.service failed because the control process exited with error code" |
| 09:48 |
mantis |
The network option has been changed to "Bridged Adapter" in the VB settings if that helps |
| 09:49 |
Dyrcona |
mantis: It's probably a misconfiguration of ejabberd. There should be an ejabberd error log on the VM that will tell you more. |
| 09:50 |
mmorgan |
mantis: Which Ubuntu .iso did you use to build the VM? |
| 09:54 |
berick |
miker: there are limitations to the facet display. i just copied the logic from the tpac's config.tt2 |
| 09:55 |
berick |
miker: https://git.evergreen-ils.org/?p=Evergreen.git;a=blob;f=Open-ILS/src/eg2/src/app/staff/catalog/result/facets.component.ts;h=aacb27c3bcb78e3515a6a97d80a742b0315fe782;hb=HEAD |
| 09:55 |
miker |
berick: hrm... is there some way, without recompiling the staff client, to add facets? |
| 09:56 |
* miker |
wonders why we have that as the default instead of just showing all... sorting, I guess? |
| 09:57 |
miker |
so... it's not the same in that we can't show more/all facets without rebuilding the release tarball, right? (I want to be missing something here...) |
| 09:59 |
berick |
miker: the angular bits would have to be recompiled, yes |
| 10:01 |
miker |
that's fine for a developer, of course ... but, one more to toss on the pile for angstcat regressions for The Normies :( |
| 10:01 |
rfrasur |
Normies |
| 10:03 |
|
jvwoolf joined #evergreen |
| 10:04 |
mantis |
mmorgan: I'm using 20.04.2 |
| 10:04 |
mantis |
Logs are telling me "invalid value of option modules" |
| 10:05 |
berick |
mantis: using https://github.com/berick/evergreen-ansible-installer ? |
| 10:08 |
mantis |
berick: Yes! |
| 10:08 |
mantis |
"Unknown: mod_irc. Did you mean mod_muc?" |
| 10:09 |
berick |
huh |
| 10:09 |
berick |
oddly, mod_irc is commented out in the ejabberd config |
| 10:10 |
mantis |
is that the playbook.yml file? |
| 10:10 |
* mmorgan |
just built a vm using the ansible installer yesterday, and didn't run into any issues. |
| 10:10 |
berick |
mantis: sudo grep mod_irc /etc/ejabberd/ejabberd.yml -- what does that show? |
| 10:11 |
mantis |
mod_irc: {} |
| 10:11 |
csharp_ |
@blame IRC |
| 10:11 |
pinesol |
csharp_: IRC stole bshum's tux doll! |
| 10:12 |
berick |
mantis: in both the ansible code and my installed version, it shows as #mod_irc -- somehow the ejabberd config template wasn't installed.. or something |
| 10:13 |
berick |
though also odd the default config would have an unsupported option |
| 10:13 |
mantis |
I could comment it out and try running it again if that's advisible |
| 10:14 |
berick |
mantis: what is shown for max_user_sessions ? |
| 10:15 |
mantis |
10,000 |
| 10:15 |
berick |
what the.. |
| 10:15 |
* mmorgan |
also sees mod_irc {} for that grep, and system is running fine |
| 10:15 |
Dyrcona |
There's no mod_irc directive on my Ubuntu 20.04 vm. |
| 10:15 |
berick |
mantis: worth a shot to comment it out |
| 10:16 |
jeffdavis |
not sure about the ansible part, but I've had to comment out mod_irc for things to work on 20.04 |
| 10:17 |
csharp_ |
mantis: looking at the Debian changes to ejabberd.yml (https://packages.ubuntu.com/focal/ejabberd), not seeing that they add mod_irc or that it is in the original upstream version |
| 10:17 |
csharp_ |
jeffdavis: weird |
| 10:18 |
berick |
yeah, it looks like the template is getting installed via the ansible installer, but for some reason mod_irc is not commented out |
| 10:18 |
berick |
which i can't explain |
| 10:18 |
Dyrcona |
And, neither of my Ubuntu 20.04 VMs have mod_irc in ejabberd.yml, not even commented out. |
| 10:19 |
berick |
Dyrcona: yeah, a holdover from older versions of the template |
| 10:21 |
mantis |
berick: I commented it out and tried running but no luck. Same error in the logs. |
| 10:21 |
|
collum joined #evergreen |
| 10:22 |
mantis |
Dyrcona: I'm running on 20.04.2 |
| 10:22 |
mantis |
idk if that makes a difference |
| 10:23 |
Dyrcona |
Doesn't make a difference. The template is installed by ansible so the problem is on the ansible side of things. |
| 10:24 |
* Dyrcona |
is having fun with templates this morning. Has one that is broken but only turns up when I add a new "page" to the OPAC. One of the included or processed templates is missing an END. I don't know why this doesn't blow up elsewhere, though. |
| 10:26 |
miker |
berick: can you think of any way that angular could pull in, a la xhr or something, some equivalent config blobs run-time? I think if there were a general framework for that we'd be able to stem the tide of some angst from angstcat ... there are several areas where that would at least let us make things work quickly without waiting for some release a year from now with an in-db config interface |
| 10:27 |
berick |
miker: of course! |
| 10:28 |
berick |
could get fancy an offer a UI to select the ones you want to see and store in WS setting |
| 10:30 |
berick |
or on the other end, just a JS file somewhere |
| 10:34 |
|
dickreckard joined #evergreen |
| 10:34 |
berick |
mantis: i'm adding support to the ansible installer for the boopac node bits, so I'm about to run the installer again on a new VM. i'll let you know what happens |
| 10:35 |
miker |
I'd be happy if there were a url that could be grabbed to override config chunks |
| 10:36 |
Dyrcona |
Does anyone have any suggestions for finding where the actual error is in a deeply nested set of TT2 templates, like in the OPAC? |
| 10:37 |
berick |
Dyrcona: apache error logs usually help |
| 10:38 |
Dyrcona |
Yeah, it just says "egweb: template error: file error - parse error - opac/ecard/form.tt2 line 33: unexpected end of input." |
| 10:39 |
Dyrcona |
Which tells me there is a missing END somewhere, but not where. |
| 10:39 |
berick |
ah, yeah, that's the best you can get I think |
| 10:39 |
Dyrcona |
I'm 99.999999% sure it is not my template, as it has 1 IF/ELSE/END. |
| 10:42 |
|
devted joined #evergreen |
| 10:44 |
berick |
miker: a general purpose .js file with config info for the angular bits makes sense |
| 10:49 |
|
jvwoolf left #evergreen |
| 10:51 |
* Dyrcona |
sighs. It's definitely in one of the included/processed templates or one included/processed by those. |
| 10:51 |
Dyrcona |
Well, there goes my day..... |
| 10:54 |
Dyrcona |
I suppose I could try this against a stock branch to see if it is a problem with our customization, which I highly suspect. |
| 10:54 |
berick |
Dyrcona: i seem to recall it tells you the file that has issues, even if it's an INCLUDEd file. i would expect form.tt2 to be the location of the issue |
| 10:57 |
|
jvwoolf joined #evergreen |
| 10:59 |
Dyrcona |
berick: It doesn't look like it to me. The error does go away if I add and extra END at the bottom, though. I'm going to test with stock to see what happens. |
| 10:59 |
berick |
k, i could be misremembering |
| 10:59 |
Dyrcona |
I could be wrong, too. |
| 11:00 |
Dyrcona |
I'll paste my template if I get really stumped. Maybe it's something obvious that I'm too blind to see. |
| 11:03 |
berick |
mantis: 20.04.2 install worked for me (now with boopac support) -- only difference is I'm using LXD instead of VB |
| 11:06 |
Dyrcona |
berick: I'm also getting an error for EGWeb.pm at line 132. In the version that I have checked out, that's where the maketext module is loaded, and my template does use l(), so maybe that's my actual problem. |
| 11:11 |
miker |
berick: sorry, meeting! yes, that would definitely be a good bandaid until some future release ... I'm not sure how runtime JS can be accessed from within angular TS, though. in the particular facet case, how would we overwrite the definition of 'export const FACET_CONFIG = ...'? |
| 11:11 |
|
jwoodard joined #evergreen |
| 11:12 |
miker |
and I'm using this as a specific example of a general problem, obv |
| 11:14 |
mantis |
berick++ |
| 11:14 |
mantis |
Thanks for the help I'll let you know if I progress any further |
| 11:14 |
berick |
miker: similar to how it interacts with the OpenSRF js. just have to 'declare' a few variables to trick the compiler. i could do a quick facet example... |
| 11:14 |
Dyrcona |
Yeah....It's in one of the other templates. When I add "END;" after the hunk where the process/includes are, the page displays without a 500. |
| 11:15 |
Dyrcona |
It's also an issue in stock, not my customization. I did that change with the stock branch. |
| 11:15 |
Dyrcona |
This is 3.5.3, fwiw. |
| 11:16 |
miker |
berick: that would be much appreciated! |
| 11:16 |
miker |
Dyrcona: boopac, perchance? |
| 11:16 |
miker |
oh, 3.5, probably not |
| 11:16 |
Dyrcona |
miker: No. I'm a dumbass. :) |
| 11:17 |
miker |
Dyrcona: ah, well, "know thyself" I guess? ;) |
| 11:17 |
Dyrcona |
I noticed that the footer appears above my text with the end at the top. |
| 11:17 |
miker |
was it no END to a WRAPPER? (been there, done that) |
| 11:17 |
Dyrcona |
Yeah, "That a bingo!" |
| 11:18 |
Dyrcona |
But I don't think we wan the END to the wrapper. |
| 11:18 |
Dyrcona |
want..... |
| 11:18 |
|
Guest23 joined #evergreen |
| 11:18 |
* Dyrcona |
smacks his fingers. |
| 11:18 |
Dyrcona |
When I put the END at the end, everything looks correct. |
| 11:19 |
Dyrcona |
Well, basically, it's my fault for forgetting how WRAPPER works. |
| 11:21 |
miker |
IMO, wrapper's semantics can be confusing. it's an inside-out include, where the named block sees the result of the WRAPPER as its content var |
| 11:21 |
miker |
like I said, been there, done that |
| 11:22 |
Dyrcona |
Yeah. |
| 11:22 |
Dyrcona |
Probably won't go there again... :) |
| 11:28 |
mantis |
Does anyone know which file the change operator code is in? Having some trouble trying to locate it. I know the authenicator is in opernsrf.xml |
| 11:30 |
miker |
mantis: change-operator in the staff client? if so, there are two places, one angularjs and one angular |
| 11:32 |
mantis |
miker: yes! |
| 11:34 |
miker |
mantis: for the angularjs interfaces (/eg/staff/blah) the service is in Open-ILS/web/js/ui/default/staff/services/op_change.js in the repo |
| 11:34 |
mantis |
miker++ thank you! |
| 11:35 |
miker |
for angular (/eg2/staff/blah) look in Open-ILS/src/eg2/src/app/staff/share/op-change |
| 11:35 |
miker |
several files in that directory |
| 11:36 |
jeff |
this software supports MARC exports in all the useful character sets: Windows Standard (ANSI), Mac Standard (ASCII), AND MARC-8! |
| 11:36 |
miker |
mantis: what are you thinking about changing in op-change? |
| 11:39 |
Dyrcona |
jeff: ASCII stupid question, get a stupid ANSI. :) |
| 11:40 |
Dyrcona |
Mac Roman would be "fun" to have deal with.... ;) |
| 11:42 |
jeff |
AppleDouble encoded MARC data, with holdings information contained in the resource fork. |
| 11:43 |
Dyrcona |
Ha! |
| 11:44 |
jeff |
245 0 $a NO TITLE |
| 11:44 |
Dyrcona |
Cataloging via HyperCard. |
| 11:45 |
miker |
Dyrcona: https://eric.ed.gov/?id=EJ476284 |
| 11:46 |
Dyrcona |
:) |
| 11:46 |
berick |
miker: https://git.evergreen-ils.org/?p=working/Evergreen.git;a=shortlog;h=refs/heads/user/berick/lpxxx-angular-dynamic-config |
| 11:46 |
* Dyrcona |
sheds a tear for ResEdit.... |
| 11:47 |
jeff |
https://archive.org/details/hc_book-keeper-20㼅最㼀:䴠剁ⵃℸ |
| 11:47 |
|
Stareagle80 joined #evergreen |
| 11:47 |
|
Stareagle80 left #evergreen |
| 11:47 |
|
Stareagle joined #evergreen |
| 11:49 |
Stareagle |
Test post to see if |
| 11:49 |
Stareagle |
I got things right |
| 11:49 |
miker |
berick: you are a gentleman and a scholar, sir! that does seem simple and I plan to steal it. (and also LP it, and point to your branch. so, not steal so much as thank you) |
| 11:49 |
Dyrcona |
I see your test. Welcome aboard, Stareagle. |
| 11:53 |
berick |
miker++ # LP |
| 11:58 |
mmorgan |
mantis: miker: Regarding change operator, we've thought about making the order of Temporary, Staff, and Persistent oposite, making Peristent the default. Haven't done it yet, though. |
| 12:00 |
Dyrcona |
mmorgan: Is that because you have a lot of staff with individual logins? |
| 12:00 |
mmorgan |
Dyrcona: Yes, exactly. Moving away from generic logins. |
| 12:02 |
* Dyrcona |
goes looking for some lunch. |
| 12:05 |
jeff |
Why not remove persistent and log out to change users? |
| 12:05 |
jeff |
I guess I could look to see how/why Persistent was added. Seems a strange place for it. |
| 12:10 |
Dyrcona |
Well, people are prone to laziness. |
| 12:12 |
rfrasur |
laziness++ |
| 12:12 |
rfrasur |
er |
| 12:12 |
rfrasur |
yeah |
| 12:13 |
mmorgan |
jeff: Persistent op change seems to have a different behavior than logout. If I have multiple tabs open, Persistent op change seems to preserve them with the new operator with a refresh. Not so much with logout. |
| 12:13 |
mmorgan |
Not sure if that's desirable or intended. |
| 12:14 |
rfrasur |
It's a useful thing for a lot of workflows. Though I tend to use "staff" more than "peristent." But if finding that a task needs a different account/permission set or demonstrating things for libraries, etc. I use both of those more than "temporary." |
| 12:17 |
mmorgan |
In the past we've found "temporary" unpredictable. Not sure if this is still true, but temporary would time out the login entirely, requiring the staff user to login again. |
| 12:19 |
mmorgan |
laziness=~efficiency (sometimes) :) |
| 12:21 |
mantis |
miker: mmorgan: same, we want to change the order, too. There's been complaints about Temp logging out of Evergreen entirely for overrides |
| 12:23 |
|
jihpringle joined #evergreen |
| 12:27 |
Dyrcona |
I don't use the staff client, except when checking bug reports, so I can't say much. |
| 12:27 |
miker |
ok, thanks folks. it wasn't intended to be "change to another user for good" functionality, really, but if that's how it's being used... |
| 12:28 |
jeff |
so is the workaround for "temp doesn't reliably work for overrides" that you "log in persistent as the overriding user", then "log in persistent as the original user"? |
| 12:29 |
miker |
it was really only meant for short-term work. think: manager at the grocery store comes over to a checkout lane with a special key that lets them override some restrictions |
| 12:29 |
* jeff |
nods |
| 12:30 |
miker |
then when the manager walks away they take their "key" with them by switching back to the original login |
| 12:31 |
miker |
if they op-change as temp, and then that times out, yeah, it fires a "log 'em all out" callback. but that's intended so that a circ clerk doesn't get to keep the admin login just because both folks walked away from the WS for 10 mins (in the "designed" workflow) |
| 12:33 |
miker |
anyway, if it's mostly about moving to another user at shift change rather than logging out and back in ... maybe that deserves a different entry point. or maybe temp really isn't useful? sounds like community surveying is in order! :) |
| 12:34 |
|
collum joined #evergreen |
| 12:41 |
|
plux joined #evergreen |
| 12:50 |
mmorgan |
miker: So what was the intent for Persistent if not to change user for good? There's no option to Restore Operator. |
| 12:51 |
mmorgan |
jeff: the "staff" option allows switching to another user without timeout, then restoring to the original user. |
| 12:54 |
berick |
mmorgan: "there's no option to Restore Operator" ? |
| 12:54 |
miker |
mmorgan: I'd have to look at the addition context -- I don't think that was an original option (I think it was originally /only/ temp) but I could be misremembering. I suspect it was "persistent exists, provides flexibility" |
| 12:55 |
jeff |
looks like there wasn't much context. commit 9600f30 |
| 12:55 |
pinesol |
jeff: [evergreen|Mike Rylander] webstaff: Allow persistent op change - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=9600f30> |
| 12:55 |
mmorgan |
berick: No, not when choosing Persistent. |
| 12:55 |
* miker |
shakes fist at former self |
| 12:55 |
jeff |
hrm. no, that's not the whole change. |
| 12:55 |
rfrasur |
the persistent maintains some settings that might get changed from a full log out/log in. Staff does as mmorgan says. And temp as miker describes - overriding something, though many things have an override log-in that provides for that. |
| 12:55 |
miker |
also, I don't recall if it was in xul or not originally |
| 12:56 |
rfrasur |
It wasn't in xul IIRC |
| 12:56 |
mmorgan |
Persistent wasn't an original option, originally, it was only temporary |
| 12:56 |
miker |
rfrasur: the override function actually /is/ the change-op feature, btw (or, was in first webstaff) ;) |
| 12:57 |
jeff |
persistent in the staff client seems odd, as opposed to staff. |
| 12:57 |
jeff |
persistent is the same as "stay logged in" in the public catalog. |
| 12:57 |
miker |
right |
| 12:57 |
berick |
mmorgan: hm, the option's there when I do a persistent change op |
| 12:58 |
rfrasur |
miker, I'm talking specifically about the "change op" menu as opposed to override pop-ups that occur. Are we talking about the same thing(s)? |
| 12:59 |
rfrasur |
And I'm also speaking exclusively from UI experience. Not behind the scenes functionality. |
| 12:59 |
miker |
rfrasur: they use the same underlying code ... might be different html, though? |
| 12:59 |
berick |
mmorgan: correction, it's there if you change operator in Angular (/eg2/) |
| 12:59 |
rfrasur |
Right, okay. That makes sense. |
| 13:00 |
rfrasur |
miker, and it uses the "temp" change op for the override pop-ups? (just clarifying) |
| 13:00 |
mmorgan |
berick: Ah ok, made me look ;-) |
| 13:01 |
miker |
rfrasur: yes, IIRC |
| 13:02 |
rfrasur |
Makes sense. |
| 13:03 |
|
collum joined #evergreen |
| 13:03 |
mmorgan |
berick: So Angular also doesn't have the Operator Changed banner across the top of the screen. (which appears when using temp, or staff in Angularjs) |
| 13:05 |
rfrasur |
And "Staff" and "Persistent" DO provide different levels of flexibility moving between log-ins. I, personally, HATE having the temp option in the change op menu, but it probably is useful where the override popup doesn't appear (if that's a thing). But my two cents don't come out as worth much when doing a currency exchange. |
| 13:08 |
miker |
mmorgan: so persist in angurlarjs doesn't maintain the prior login, I see that in the code now, sorry for the confusion! it looks like it is intended for "change user, for good" ... I think we should get kmlussier in here to discuss the history that I don't recall, too! :) |
| 13:08 |
mmorgan |
:) |
| 13:15 |
berick |
mmorgan: ah, an oversight on the banner. |
| 13:19 |
* mmorgan |
can open a bug about the lack of banner and the "restore" option after a persistent op change in eg2. |
| 13:20 |
berick |
mmorgan++ # would be appreciated |
| 14:08 |
|
nfBurton joined #evergreen |
| 14:53 |
|
Christineb joined #evergreen |
| 14:54 |
|
jonadab joined #evergreen |
| 14:58 |
jvwoolf |
Has anybody been able to set up notification that go to the patrons using the stgu.created action trigger? |
| 15:02 |
jvwoolf |
I'm trying it now, but I can't figure out how to get the patron info to come up. I thought defining [%- user = target -%] at the top and using user.email, user.family_name at the top would work but it doesn't. |
| 15:02 |
Dyrcona |
jvwoolf: Have put au in the environment? |
| 15:02 |
Dyrcona |
Have ^you put.... |
| 15:03 |
jvwoolf |
Dyrcona: I wasn't sure what should go there. I had assumed it would be pulling directly from the staging.user_stage table |
| 15:04 |
jvwoolf |
Oh, I see what you're saying |
| 15:04 |
jvwoolf |
There won't be anything in au though because these would be still pending |
| 15:04 |
Dyrcona |
jvwoolf: Oh, right.. staging user.... See if it is in the IDL and what class it ihas. If it's not in the IDL, you'll likely have to add it. |
| 15:05 |
jvwoolf |
Ah, got it. Thank you. |
| 15:05 |
jvwoolf |
Dyrcona++ |
| 15:05 |
Dyrcona |
The trigger environment has to have the classes int he environment in order for the trigger to work. |
| 15:11 |
jvwoolf |
Hm. It seems like stgu has everything I'm calling in it. |
| 15:13 |
mmorgan |
jvwoolf: Dyrcona: So stgu would need to be in the environment, right? |
| 15:16 |
mmorgan |
nm, I see the backscroll :) |
| 15:21 |
jeffdavis |
With three new parallel-requests bugs reported this week, I wonder if we need to take another look at client-side request throttling? |
| 15:24 |
jeffdavis |
i.e. bug 1912834 |
| 15:24 |
pinesol |
Launchpad bug 1912834 in OpenSRF "Browser client should limit the number of parallel requests" [High,New] https://launchpad.net/bugs/1912834 |
| 15:30 |
jvwoolf |
jeffdavis: I actually need to submit a bug related to that. We had an instance of a library adding too many members to a group, which we think created too many parallel requests. |
| 15:31 |
jvwoolf |
I've been meaning to ask for help with verifying this. I'm not sure how to view open-ils calls in real time. |
| 15:37 |
jvwoolf |
Dyrcona: mmorgan: I got it to work! |
| 15:37 |
mmorgan |
jvwoolf++ |
| 15:38 |
mmorgan |
What was the trick? |
| 15:38 |
jvwoolf |
Needed to be [%- user = target.0 -%] at the top |
| 15:38 |
mmorgan |
Ok, gotcha! |
| 15:38 |
jeff |
sounds like you're grouping in the event def? |
| 15:39 |
jeff |
if so, what field are you grouping on? |
| 15:40 |
jvwoolf |
jeff: I was working off the original template, which creates a single email listing multiple patrons for staff |
| 15:41 |
Dyrcona |
jvwoolf: You might want to drop the grouping on the event definition and then you should be able to drop the .0. |
| 15:41 |
jvwoolf |
Aha! |
| 15:41 |
jeff |
what value is in your event definition's "Processing Group Context Field"? |
| 15:42 |
jeff |
I doubt that you want to group in this case, but I suppose you could group on id if you want to force target to be an array. |
| 15:42 |
jvwoolf |
jeff: home_ou |
| 15:43 |
jvwoolf |
jeff: seems unlikely that a patron would request more than one card |
| 15:43 |
jeff |
Yeah, I would suggest changing that. Probably easiest/best to drop it and do as Dyrcona suggests and drop the target.0 array syntax. |
| 15:43 |
jvwoolf |
jeff: Will do. Thank you! |
| 15:43 |
jvwoolf |
jeff++ Dyrcona++ |
| 15:44 |
jeff |
as it is, if there are multiple patrons at a given home_ou that have a staging user, you'll only be sending email to one of them (but you'll be saying in the db that the one event handled all of them) |
| 15:44 |
jeff |
and I'm pretty sure you don't need to put stgu in an event def's environment if the core type of the hook is stgu. |
| 15:45 |
jeff |
also, make sure you have the stgu suggested filter in your filter defs, though I suspect if you didn't you wouldn't be having the success you had already. |
| 15:46 |
jeff |
(I think without the filter in place the passive hook will never create any events) |
| 15:46 |
jeff |
also, you might want/need a different max event validity delay. |
| 15:46 |
jvwoolf |
jeff: Yeah, I got it to work without adding it to the enviroment. And yes, I also had to go back and add the filter to the action_trigger_filters.json |
| 15:46 |
jeff |
depending on how many not-created staging users you have and how often you purge and how far back you want to send emails. |
| 15:47 |
jvwoolf |
It's been a very educational day :) |
| 15:50 |
jvwoolf |
The Max Validity Delay is 1 day, which seems pretty reasonable |
| 15:53 |
* mmorgan |
is following all this good info! |
| 15:53 |
mmorgan |
jeff++ |
| 15:57 |
* jvwoolf |
is creating a pending patron for every fictional character today |
| 15:58 |
gmcharlt |
berick: et al. interested in your opinion on bug 1932363 |
| 15:58 |
pinesol |
Launchpad bug 1932363 in Evergreen "Angular: eg-org-family-select with limitPerms set can include ancestors that the user doesn't have access to" [Medium,New] https://launchpad.net/bugs/1932363 |
| 15:59 |
gmcharlt |
@later tell sandbergja interested in your opinion on bug 1932363 |
| 15:59 |
pinesol |
gmcharlt: The operation succeeded. |
| 16:13 |
jeffdavis |
jvwoolf: I don't have a way to monitor API calls in real time, but we do monitor # of drones for each service, so I get alerts when we hit max_children. I also have a cronjob that looks for "no children available" warnings in the logs. |
| 16:13 |
jeffdavis |
When that happens I manually review the logs looking for clusters of similar-looking calls to that service leading up to the alert. |
| 16:14 |
jeffdavis |
Then I look at the staff client to see where that particular call happens, and see if I can reproduce on a test server. |
| 16:14 |
jvwoolf |
jeffdavis++ |
| 16:14 |
jvwoolf |
Thanks for the info |
| 16:40 |
mmorgan |
We've had a few reports over the past week of white screens and the screen just refreshing when attempting to login. Anyone else heard similar? |
| 16:41 |
mmorgan |
Usually it's related to hatch or the Chrome extension. |
| 16:48 |
Shula_L |
We had that happening, and in our case it was a failed redirect to /staff/eg2/, and staff trying to log-in on /staff/eg/. |
| 16:55 |
mmorgan |
Shula_L: Interesting, that doesn't seem to be our issue, though. I always want to blame updates, Windows updates have been going around. |
| 16:56 |
Shula_L |
mmorgan: I wouldn't be surprised at all if it's a case of the updates. |
| 17:44 |
mmorgan |
berick: fyi, bug 1932369 |
| 17:44 |
pinesol |
Launchpad bug 1932369 in Evergreen "Angular: Change Operator behavior differs from AngularJS" [Undecided,New] https://launchpad.net/bugs/1932369 |
| 17:45 |
|
mmorgan left #evergreen |
| 18:01 |
pinesol |
News from qatests: Testing Success <http://testing.evergreen-ils.org/~live> |
| 20:01 |
|
kco joined #evergreen |
| 20:21 |
|
jihpringle joined #evergreen |
| 20:24 |
|
cmalm joined #evergreen |
