Evergreen ILS Website

IRC log for #evergreen, 2016-11-22

| 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
00:00 * bshum didn't test the feature, but wonders idly how it aligns with the action_trigger_runner script and various timings
00:00 bshum It seems to have a 5 min delay after the checkout, so I would expect it to take that long before it did anything, and also only happen when the actiontrigger runner script runs
00:00 * bshum might have to play and learn how it's supposed to work :)
05:00 pinesol_green News from qatests: Test Success <http://testing.evergreen-ils.org/~live>
07:30 rjackson_isl joined #evergreen
08:14 collum joined #evergreen
08:22 kmlussier joined #evergreen
08:26 agoben joined #evergreen
08:37 mmorgan joined #evergreen
09:05 Dyrcona joined #evergreen
09:30 Bmagic bshum: thanks for the late night support!
09:30 Bmagic bshum++
09:31 yboston joined #evergreen
09:36 Bmagic I think it's interesting and figured I would share. It seems when memcached runs out of space, the resulting behavior is interesting. Some people said that checking in items will result in a wrong transits.
09:36 Bmagic Some people were prompted to login with every screen load. And sometimes, the staff client would suddenly behave like they were logged into a different workstation at a different branch!
09:36 Bmagic Almost like the auth tokens were playing musical chairs
09:43 jeff In the way that we currently use memcached, if you get in a situation where memcached is evicting, you've lost.
09:44 jeff I'd be interested in hearing more details about the "client would suddenly behave like they were logged into a different workstation at a different branch" part.
10:00 tsbere That almost sounds like either corruption of memcache's memory space or something isn't acting properly when things can't be returned
10:05 jeff or a re-prompt for auth using au.home_ou instead of workstation ou, etc.
10:34 miker Bmagic: where (and how many times) are you running memcache?
10:35 miker kmlussier: I've tested rebasing the sprint4 branch against master ... it applies cleanly. do you have any desire to merge some of the webstaff bug fixes into the collab branch, and then merge the whole shebang into master soon-ish?
10:36 pinesol_green [evergreen|Kyle Huckins] LP1537214 Staff Initials in Patron Notes - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=1fbaa1d>
10:38 kmlussier miker: I could test and merge what's there now. Did that last commit mess up your rebase?
10:38 miker no, it didn't. it happened after my rebase test :)
10:39 kmlussier OK, I'll take a look at it now then. I've already tested most of what's in the collab branch, so it shouldn't take too long. As long as I don't get distracted.
10:40 miker I actually was thinking more about the outstanding LP pullrequests
10:40 miker fwiw
10:41 miker pulling those into the collab branch, and then shoving the lot into master
10:41 kmlussier miker: Is there an advantage to doing it that way?
10:41 miker but, if you're doing the committing, you define the process :)
10:43 miker it fits in my head better ... and I think of the pullrequest'd code as further from master than the collab branch. the collab branch is "all the things working together" rather than the individual, segregated silos of the small branches
10:48 Bmagic miker: memcache is running on a single machine with 4096 as the limit. 3 bricks.
10:56 sandbergja joined #evergreen
10:59 miker kmlussier: but either way is fine. we may just find we're fixing collab code instead of pullrequest code if there's conflict
11:00 kmlussier miker: OK, I'm testing a couple of small things to possibly put in the collab branch before I look at the larger branch.
11:01 miker kmlussier: one situation that might go the other way would be obviously backportable fixes ...
11:04 kmlussier miker: Yeah, I haven't been backporting as much based on brief discussion with berick last week http://irc.evergreen-ils.org/​evergreen/2016-11-16#i_276134
11:04 kmlussier It seemed best to focus efforts on trying to get web client code into 2.12 sooner.
11:19 brahmina joined #evergreen
11:39 bmills joined #evergreen
11:52 miker kmlussier: ahh, I missed much of that. yeah, I agree with the sentiment -- we're 3 months from a "production" release, let's focus there
12:00 jihpringle joined #evergreen
12:12 berick jeff: when you have a sec, remind me what we were looking at that would allow a browser extension to be externally_accessible (w/ perm click-through) based on the content of a page (node ID, css class, etc.) instead of the domain.  thought I bookmarked it, but no.
12:37 collum joined #evergreen
12:39 kmlussier If anyone has spare time and is looking for something 'fun' to do, it looks like 2.11.1 and 2.10.8 milestones could be updated in Launchpad.
12:46 dbwells I'll do 2.11.1 if it isn't done when I return from my meeting this afternoon (~3:00pm).
12:47 dbwells Actually, I'll do both at that time.
12:48 kmlussier dbwells++ Thanks!
13:11 genpaku joined #evergreen
13:38 bmills joined #evergreen
13:38 terran joined #evergreen
13:50 terran Hi gang - I'm trying to figure out how to enable Spanish on a 2.11.1 test server. I've updated eg_vhost.conf so the dropdown appears, but nothing happens when I try to use it - can someone push me in the right direction?
13:51 bshum terran: Hi Terran!  Spanish you say?  Fascinating :)
13:51 terran bshum: I saw you were working on that :)
13:52 bshum terran: So when you say you have a dropdown for the language picker, what languages show up there?  And you select and hit the change button and nothing?
13:52 bshum What did you throw in the eg_vhost.conf btw
13:52 bshum Oooh, also better question
13:53 bshum TPAC or webstaff?
13:53 terran bshum:     PerlAddVar OILSWebLocale "es_es"     PerlAddVar OILSWebLocale "/openils/var/data/locale/opac/es-ES.po"
13:53 terran TPAC
13:53 bshum Just checking, webstaff is glitchy (read, BROKEN)
13:54 terran bshum: I added them right after the sample French lines, line 712
13:54 JBoyer I'm guessing in the dark here, but I would think that the 2 vars would have 2 different names since one is a locale and the other a file full of strings. Not so?
13:55 bshum terran: Did you check to make sure you actually have a file at /openils/var/data/locale/opac/es-ES.po  ?
13:55 terran errr....
13:56 * bshum grabs the 2.11.1 tarball to see what's in it...
13:57 Christineb joined #evergreen
13:57 bshum Well it's in the source file, so that's good
13:57 bshum So like Evergreen-ILS-2.11.1/Open-ILS/​src/data/locale/opac/es-ES.po
13:58 JBoyer Also, for the logs: I have shown a light upon my ignorance and no, you do use the same var name in pairs. What a *cough* delightful perl-ism.
14:01 jeff berick: I don't think it required using externally_connectable at all. The idea was to use chrome.declarativeContent.RequestContentScript to inject Javascript into pages matching criteria (like, url matches /eg/staff and/or page contains an element with this magic ID).
14:02 bshum terran: A quick reveal would be to do something like "nano /openils/var/data/locale/opac/es-ES.po " on the server and see if you see a file with strings and translations, otherwise if it comes up empty or not found, you're missing something
14:03 terran bshum: I'm not even seeing a data folder, so I'm guessing it didn't installed.
14:03 jeff berick: lots of unanswered questions there, though. i don't know what the prompt looks like, or if the permission grant persists across browser sessions, etc.
14:03 bshum o.O
14:03 jeff berick: i'm going to play around with it a little bit.
14:03 bshum Well there should definitely be an /openils/var/data folder
14:04 bshum That's where things like <ils_events>/openils/var/data​/ils_events.xml</ils_events> gets put
14:04 bshum So, if you're missing that, then I think something is super odd with your installation......
14:05 jeff berick: to summarize the chain: by using the declarativeContent api to inject a content script, the content script can message its parent extension without needing to hardcode anything. that extension in turn can be on the allowed_origins list in the Native Messaging Host's manifest.
14:06 jeff (because content scripts can message their parent extension without requiring special permissions)
14:07 jeff there are still some other options if that turns out to be unsuitable, including "ask for scary permissions" and "build your own extension and customize your native messaging host manifest to include your custom extension's ID"
14:07 jeff but i at least like the promise of this possibility.
14:07 kmlussier OK, I should be merging the sprint4 branch very soon. Now's not a good time to add to the collab branch if anyone was planning to do so.
14:07 kmlussier Sorry to interrupt the regularly-scheduled chat.
14:08 terran bshum: Okay, I see the open/ils/var/data folder, but it doesn't have a locale folder in it (it does have the ils_events.xml file in it though)
14:09 jeff kmlussier++ a worthy interruption! :-)
14:10 kmlussier @love git quickpick
14:10 pinesol_green kmlussier: The operation succeeded.  kmlussier loves git quickpick.
14:10 terran bshum: So yes, that seems to be the problem. I'll talk to Chris about the install - thank you for pointing me in the right direction!
14:10 kmlussier bshum++
14:10 terran bshum++
14:11 terran And JBoyer++ for eyeballing it, too!
14:12 JBoyer terran++ # for not giving up in the midst of my nonsense. ;)
14:13 pinesol_green Showing latest 5 of 44 commits to Evergreen...
14:13 pinesol_green [evergreen|Galen Charlton] webstaff: teach egJsonExporter about generators - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=df49728>
14:13 pinesol_green [evergreen|Galen Charlton] webstaff: add import and export of print templates - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=ea4d93e>
14:13 pinesol_green [evergreen|Jason Etheridge] webstaff: language tweak in Bill History receipt - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=8894286>
14:13 pinesol_green [evergreen|Mike Rylander] webstaff: Allow ACQ catalog and link-to-record, er, links to work in a browser - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=b6ff1f4>
14:13 pinesol_green [evergreen|Billy Horn] LP#1522599 add visual indicator for an expired patron - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=46c2449>
14:14 terran JBoyer: Not nonsense at all! I was guessing a bit based on contextual clues anyway.
14:18 berick jeff++ # thanks!
14:18 berick kmlussier++
14:18 bshum terran: I'd be curious what csharp ends up discovering
14:19 bshum To be honest, I haven't used a tarball in years
14:19 * berick has some rebasing to do
14:19 bshum So maybe it's not copying over the locale files right during the install phase
14:19 terran bshum: sure thing, I'll let you know what he says
14:19 bshum terran++ # testing things
14:19 * csharp appears just at that moment
14:20 csharp oh - spanish, I see
14:21 bshum To get things rolling, I think it should be as simple as "mkdir /openils/var/data/locale" and Then "cp /home/wherever/Evergreen-ILS-2.11.1/O​pen-ILS/src/data/locale/opac/es-ES.po /openils/var/data/locale/es-ES.po"
14:21 bshum Followed by an apache restart
14:21 bshum But there might be more issues at stake I guess
14:22 bshum And if you want other languages, well... lol
14:22 csharp bshum++
14:27 * jeff closes browser tabs for three hours worth of phone calls
14:27 jamesrf joined #evergreen
14:56 Dyrcona So, I did a git pull on master this morning, and just about 3 hours later, I'm 44 commits behind.... kmlussier++
15:01 csharp bshum: your suggestion to just copy the es-ES.po file works (partially), and is revealing where we might have some hard-coded strings in our customizations
15:01 mmorgan kmlussier++
15:01 csharp kmlussier++
15:01 csharp bshum++
15:02 bshum Cool!  And yeah.... that gets weird
15:02 bshum csharp++ # magic wand!
15:04 terran bshum++ csharp++
15:05 bshum In theory, one could create an "english" PO file, and drop customized strings into it to replace the existing english with other words instead
15:05 bshum but if it differs too much, then the difference between English and Spanish might drift too much for comfort
15:06 Dyrcona Yes, that is one way to customize strings without changing the code.
15:12 hbrennan joined #evergreen
15:15 terran To make sure I understand - there isn't a separate English po file right now, it just displays whatever is in the tt2? (Like [% l('Acquisitions') %] ) --- so if I created a separate English PO file, would I only need to have translation lines in it for the ones that would vary from the original?
15:16 bshum terran: I want to say yes.
15:17 terran bshum: I'll try that out
15:17 Dyrcona terran: I believe that is how it would work.
15:18 Dyrcona gettext, at least, falls back to using the "key" as the display string, which is the string from the l() wrapper or whatever.
15:24 kmlussier I just want to remind folks to add any potential 2.12 new features to the Roadmap at https://wiki.evergreen-ils.org/doku​.php?id=faqs:evergreen_roadmap:2.12.
15:24 kmlussier I'll probably poke some people directly after Thanksgiving.
15:33 * berick adds one
15:35 jeffdavis re "Add support for jQuery to the public catalog" on the roadmap: does that mean converting all OPAC use of Dojo to jQuery?
15:35 jeffdavis or will we support both for a period?
15:38 kmlussier jeffdavis: The existing catalog features using Dojo will not move right away, but the plan is that they should all eventually move to jQuery.
15:49 bmills joined #evergreen
16:07 bmills joined #evergreen
16:19 gmcharlt kmlussier: FYI, I've now merged master back onto collab/miker/sprint4; the sprint4-merge-nov22 tag identifies the starting point the next time a merge shoudl be done
16:23 kmlussier gmcharlt: OK, thanks!
16:23 kmlussier gmcharlt++
16:31 JBoyer Happy Thanksgiving #evergreen!
16:34 kmlussier Yes, Happy Thanksgiving all!
16:34 mmorgan Happy Thanksgiving!
17:00 pinesol_green News from qatests: Test Success <http://testing.evergreen-ils.org/~live>
17:03 mmorgan left #evergreen
19:52 StomproJ joined #evergreen

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