Evergreen ILS Website

IRC log for #evergreen, 2018-03-13

| 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
00:42 dbwells joined #evergreen
06:32 pinesol_green News from qatests: Failed Running perl live tests <http://testing.evergreen-ils.org/~live>
08:01 rlefaive_ joined #evergreen
08:33 collum joined #evergreen
08:35 Dyrcona joined #evergreen
08:49 littlet joined #evergreen
08:56 dwgreen joined #evergreen
09:17 jvwoolf joined #evergreen
09:34 jvwoolf1 joined #evergreen
09:36 rlefaive joined #evergreen
09:38 kmlussier joined #evergreen
09:42 terran joined #evergreen
09:58 kmlussier Are bug 1751318 and bug 1754894 the same bug?
09:58 pinesol_green Launchpad bug 1751318 in Evergreen "live_t/24-offline-all-assets.t failure" [Medium,Confirmed] https://launchpad.net/bugs/1751318
09:58 pinesol_green Launchpad bug 1754894 in Evergreen "24-offline-all-assets.t broken with webpack dev environment" [Undecided,New] https://launchpad.net/bugs/1754894
09:58 kmlussier They have two different fixes.
09:59 Dyrcona I didn't think they both had fixes.
10:00 Dyrcona The latter is happening after the previous, unless it was never pushed to master.
10:00 * Dyrcona checks.
10:00 Dyrcona Oh, duh.
10:01 Dyrcona Maybe they are the same bug.
10:01 Dyrcona I'll test the fix for the former.
10:02 kmlussier Dyrcona++
10:04 rlefaive joined #evergreen
10:08 Dyrcona Wouldn't ya know, the test just doesn't want to fail for me, now.
10:13 mmorgan joined #evergreen
10:23 rlefaive joined #evergreen
10:26 Dyrcona And, it looks like offline just doesn't work in Chromium when you're really offline. It apparently works in Firefox and Chrome.
10:27 beanjammin joined #evergreen
10:27 rlefaive joined #evergreen
10:33 Christineb joined #evergreen
10:37 pinesol_green [evergreen|Kevin Tran] Docs: Update copy_statuses.adoc for web client - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=ee0e56f>
10:37 pinesol_green [evergreen|Kevin Tran] Docs: Update SMS_messaging.adoc for web client - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=b914da6>
10:37 Dyrcona Well, this is interesting....
10:38 Dyrcona If you've been to the site and had offline mode open before.
10:39 Dyrcona If you need to accept a new "invalid" certificate in Firefox when you go back, you get offline mode rather than the confirm the security exception.
10:39 Dyrcona You have to go to the OPAC and confirm the except before you can login to the staff client.
10:47 kmlussier The recent test failures for 19-lp1306666-abort-transit-copy-status.t appear to be caused by bug 1755502.
10:47 pinesol_green Launchpad bug 1755502 in Evergreen "Alternate hold pickup popup displays when checking out item to hold patron" [High,New] https://launchpad.net/bugs/1755502
10:47 csharp is there an EG db function that strips out marc fields? for instance, I have a large group of records and I want to strip a 9XX field out indiscriminately
10:48 bshum csharp: Kind of like https://wiki.evergreen-ils.org/doku.​php?id=scratchpad:random_magic_spell​s#how_to_prune_a_tag_under_the_hood ?
10:48 Dyrcona csharp: No there isn't, but we're working on adding that feature to marc export.
10:49 csharp Dyrcona: oh cool
10:49 csharp bshum: I'll look - I forget that page exists :-)
10:50 bshum Yeah it's an oldie, but I just always remember using that SQL to strip unwanted tags from our bibs too
10:50 Dyrcona csharp: It wouldn't be too difficult to add code like that to a function, if you just want to prune tags that match a regex.
10:51 csharp Dyrcona: yeah, I was thinking about doing that, then thought "surely someone else has had this need before" :-)
10:51 Dyrcona BTW, the marc_export only strips the tags in the output. They'll still be in the database records.
10:52 Dyrcona Just in case it wasn't obvious.
10:53 csharp Dyrcona: I figured ;-)
10:53 Dyrcona It was as much for the logs as for you. :)
10:53 csharp thanks
10:53 Dyrcona I wouldn't want to disappoint anyone down the road. :)
11:13 pinesol_green [evergreen|Bill Erickson] LP#1751318 Remove direct tablesort import - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=4d7851e>
11:23 rlefaive joined #evergreen
11:41 pinesol_green [evergreen|blake] LP1411422 Copy details repeated in search results when item/volume moved with parts attached - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=01e21f8>
11:41 pinesol_green [evergreen|blake] LP1411422 Copy details repeated in search results when item/volume moved with parts attached - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=f0ba9f6>
11:41 pinesol_green [evergreen|Jason Stephenson] Lp#1411422: Clean up the code a little bit. - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=70699c5>
11:41 pinesol_green [evergreen|Jason Stephenson] Lp 1411422: Fix & rename test file. - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=8537883>
11:53 kmlussier Bmagic++ Dyrcona++
11:55 khuckins joined #evergreen
11:59 ngf42 joined #evergreen
12:09 beanjammin joined #evergreen
12:30 yboston joined #evergreen
12:33 jihpringle joined #evergreen
12:43 Dyrcona jeffdavis: If you're around, I have a question about overdrive integration.
12:43 jeffdavis I'm around. :)
12:44 Dyrcona What triggers the check out feature being available? I'm trying to test it with someone at Overdrive on their test site and nothing shows a way to check it out.
12:45 jeffdavis let me take a look
12:54 jeffdavis Dyrcona: you need to have ebook_api.enabled = 'true' in config.tt2 and the args.ebook object needs to exist. The latter is created at the end of misc_util.tt2 if the record contains an identifier belonging to one of your enabled ebook vendors (like an Overdrive identifier in the 037 tag).
12:55 jeffdavis Dyrcona: If both those things are the case, there will be hidden "Check Out E-Item" and "Place Hold on E-Item" links on the page. JS will unhide one of them based on whether the external API reports available copies.
12:55 jeffdavis So you get a checkout link if there's an available copy, and a place hold link otherwise.
12:56 yboston joined #evergreen
12:58 jeffdavis If you're not seeing one of those links on search results/record summary, you could check to see whether the hidden elements exist or not; if not, EG is not detecting that the record is an Overdrive record. Could be the identifier isn't in the expected location/format or (for older-style records without the ID in 037) your Overdrive URL pattern in config.tt2 doesn't match the URL in 856.
12:59 jeffdavis If the hidden elements do exist, the issue is likely with the availability API lookup, in which case you could see what osrfsys logs report about the API call.
13:05 Dyrcona I have multiple patterns in config.tt2. Does that work?
13:05 Dyrcona And what would I be looking for for the hidden elements?
13:06 * Dyrcona is, as typical, trying to do 3 things at once.
13:07 jeffdavis Multiple patterns should work. Note that they are matched on the 856 using the TT2 match function, i.e. they are treated as regexes, so forward slashes might need to be escaped and so on.
13:07 jeffdavis We use the 037 so I'm not sure about that bit.
13:08 Dyrcona I think you mean reverse solidus, aka backslash. ;)
13:08 Dyrcona Should be any reason to escape a slash.
13:08 Dyrcona Shouldn't.
13:08 rlefaive joined #evergreen
13:09 jeffdavis Well, I was thinking if match expects a pattern of the form "/something/" then it might be misinterpreting the forward slashes in "http://example.com/"
13:10 jeffdavis The hidden elements contain a link to /myopac/ebook_checkout or /myopac/ebook_place_hold. You could also check your browser console, if it says 'Found 0 ebooks on this page' that's a sign that EG is not picking up that your records are ebook records.
13:10 Dyrcona I'm getting lots of errors from ebook API, too.
13:11 Dyrcona found 1 ebooks on this page
13:11 Dyrcona So, that's working.
13:12 jeffdavis What sort of API errors?
13:17 Dyrcona EbookAPI: OverDrive Library Account API request failed
13:17 Dyrcona EbookAPI: unable to set HTTP Authorization header without token
13:18 Dyrcona EbookAPI: failed request: $VAR1 = {
13:18 Dyrcona EbookAPI: attempted an HTTP request but no URI was provided
13:18 Dyrcona EbookAPI: failed to retrieve OverDrive holdings counts for title 9816FFF8-FF33-4C2F-BE39-E87488F685E4
13:19 Dyrcona I should probably truncate the log and give it another go.
13:19 Dyrcona The log is 256MB.
13:19 yboston joined #evergreen
13:21 jeffdavis "unable to set HTTP Authorization header without token" sounds like your system is not authenticating properly with Overdrive's test environment - could be "OverDrive Basic Token" is not set properly in org settings or your account isn't set up properly on their end.
13:22 jeffdavis curl -v -k -H "Content-Type: application/x-www-form-urlencoded;charset=UTF-8" -H "Authorization: Basic ${BASIC_TOKEN}" -X POST -d 'grant_type=client_credentials' https://oauth.overdrive.com/token
13:23 jeffdavis Dyrcona: try that with $BASIC_TOKEN = your base64-encoded client key+secret?
13:24 Dyrcona That works.
13:25 Dyrcona I get back a JSON object with an access_token and other fields.
13:25 jeffdavis hmm
13:26 Dyrcona Do they use the same oauth server for testing? They must 'cause that's the token that I calculated for testing.
13:26 jeffdavis Yes, same auth endpoint for prod and testing.
13:30 Dyrcona It's apparently getting a 200 OK from the oauth service but failing to get the token.
13:31 Dyrcona jeffdavis: https://pastebin.com/WmkH0n4N
13:36 jeffdavis Don't you love it when HTTP status codes don't actually match the nature of the response? :(
13:36 jeffdavis "EbookAPI: failed request: $VAR1 = {" is the beginning of Dumper output of the response I think, do the lines after that in your logs have any more information?
13:37 Dyrcona Well, yes.
13:39 Dyrcona https://pastebin.com/9qirtYCk
13:39 Dyrcona I believe the headers are empty because it is not getting the token.
13:45 yboston joined #evergreen
13:49 Dyrcona So, just looking at the code involved, this looks like it should be working.
13:54 csharp whee! perl is fun - trying to access the values within an array of arrays
13:56 Dyrcona pfft... easy...
13:57 csharp Data::Dumper shows the values I'm after, but when I try to print/manipulate them, I only see the likes of ARRAY(0x353b1a0)
13:57 Dyrcona They're array references. enclose it with @{}
13:57 Dyrcona Modern perl should just print arrays.
13:58 Bmagic Dyrcona++ # bib merge parts bug merge
13:58 Dyrcona What's fun is I've got an Emacs window open that I can't see....
13:58 csharp https://pastebin.com/rCZ81U75
13:58 jeffdavis Dyrcona: can you share your values for the org settings ebook_api.overdrive.discovery_base_uri and ebook_api.overdrive.circulation_base_uri (if any)?
13:59 * csharp is trying to fix bug 1666987
13:59 pinesol_green Launchpad bug 1666987 in Evergreen "build/tools/relator_map no longer fetching relator codes" [Low,Confirmed] https://launchpad.net/bugs/1666987 - Assigned to Chris Sharp (chrissharp123)
14:00 Dyrcona csharp: I looked into some of that and thought part of the problem was that LoC was blocking bots.
14:00 Dyrcona But, maybe I'm thinking of something else, and I should concentrate on my problem at hand.
14:00 Dyrcona jeffdavis: Maybe in a bit. I'm trying to log the request and my tools are fighting me.
14:01 jeffdavis no worries, I'm going afk for a bit now anyway :)
14:01 csharp Dyrcona: yeah it's blocking bots (understand about divided attention - thanks for responding anyway :-) )
14:01 abowling joined #evergreen
14:01 csharp LWP::UserAgent solve that issue
14:02 csharp *solves
14:02 jeffdavis unrelated, but I think I see a possible loop in the overdrive code that I should fix: do_client_auth > get_library_info > handle_http_request > 401 response from API > do_client_auth
14:06 Bmagic Is there a database function that can quickly tell if any given date lands on a day that the library is either closed for a holiday or is closed on that weekday?
14:07 Dyrcona Gnome notifications can be a pain, particularly when they pop up over the browser tab you want to click on just as you're about to click on it.
14:07 berick Bmagic: don't think so.  the code uses open-ils.storage.actor.org​_unit.closed_date.overlap
14:07 Bmagic cool, thanks
14:08 Bmagic berick++
14:10 Dyrcona So, the request looks good, and I'm getting a response.
14:10 littlet joined #evergreen
14:11 Dyrcona Somewhere between getting the token and trying to use it, the token is getting lost.
14:12 csharp @who lost Dyrcona's token?
14:12 pinesol_green Bmagic lost Dyrcona's token.
14:12 csharp @blame [someone] for losing Dyrcona's token
14:12 pinesol_green It's all maryj's fault! for losing Dyrcona's token
14:13 Dyrcona Oh! They gave me the same URI for circulation and discovery.
14:13 Dyrcona So, it's looking for a patron token, but I ain't got one.
14:17 Dyrcona So, I delete the circulation URI setting, and the errors disappear, but still no way to checkout.
14:18 Dyrcona I threw the ball back to Overdrive because I need to know if the test environment has a different URI for circulation and if so, what it is.
14:20 abowling1 joined #evergreen
14:23 abowling2 joined #evergreen
14:33 jeffdavis You probably want https://integration-patron.api.overdrive.com/ for circulation base URI and https://integration.api.overdrive.com/ for discovery base URI, but Overdrive will be able to confirm
14:35 Dyrcona They told me to use: "http://overdriveapi.lib.overdrive.com/
14:35 Dyrcona But, is that the production URI?
14:41 rlefaive joined #evergreen
14:41 Dyrcona Using those URIs, I get 404 responses.
14:43 jeff hopefully not http
14:44 Dyrcona That's what they said to use in an email.
14:46 yboston joined #evergreen
14:49 Dyrcona On the bright side, it works with the https protocol. Still doesn't show the checkout links, they're hidden.
14:49 Dyrcona I'll need to find out what to use for circulation.
14:56 Bmagic when action.circulation receives a new row, the due_date is pushed to 11:59:59. Now that we are passed Daylight savings time, I would expect all of the due dates to be at that time of day. But I am still finding things that are getting 10:59:59 for due date.
14:57 rlefaive_ joined #evergreen
14:58 csharp Bmagic: servers getting updated with NTP or something?
14:59 Dyrcona Bmagic: 1. Does that hour really matter? 2. Is the timezone set correctly so that DST takes effect automatically? (There are often versions of the timezone where it doesn't.) 3. It could be a bug in the new timezone-awareness code, but I think we've gone through a couple of changes already.
14:59 rlefaive__ joined #evergreen
15:01 Dyrcona @blame docker. :)
15:01 pinesol_green Dyrcona: docker. :) is the SPY!
15:01 Bmagic they are all using NTP
15:02 berick Bmagic: curious if the 10:59 due times use the same timezone offset as NOW().
15:02 Bmagic the docker machines are reporting the right timezone, the database machines are reporting the right timezone
15:03 Bmagic it matters because occationally (not always) the due date is off by a whole day. I believe the circulations that are wrong are the ones that are due on 10:59:59. All of the due dates of 11:59:59 have the correct day
15:04 Bmagic occationally / occasionally  #gees I suck at english
15:05 berick Bmagic: what't the full time stamp of the due date in the DB?
15:05 Bmagic Ex: "2018-04-11 22:59:59-05"
15:05 berick what does 'select now()' show?
15:05 Bmagic "2018-03-13 14:05:52.24003-05"
15:06 berick ok, that answers my question.
15:06 berick and that's for a circ created after this Sunday morning?
15:07 Bmagic I started my session with my DB client by issuing 'set session time zone 'EST';' - but now() is reporting the wrong time for that zone.... what's up with that?
15:07 Bmagic correct, I am looking at circs created post DST
15:07 berick we're currently DST
15:08 Bmagic I would expect now() to tell me that it's 3:08 EST
15:09 csharp CST, no?
15:09 csharp or CDT, actually
15:09 Dyrcona Bmagic: Yeah, EDT or CDT, not EST or CST.
15:09 berick ignore my dst comment, i don't know what PG wants.
15:09 Dyrcona -05 is just the offset which is either EST or CDT.
15:10 csharp -04 is EDT
15:10 Dyrcona EDT offset is -04
15:10 Bmagic I see, EDT vs EST.
15:10 Bmagic EDT    -14400 D  # Eastern Daylight Saving Time
15:10 Bmagic EST    -18000    # Eastern Standard Time (America)
15:10 Bmagic another gotcha!
15:11 Bmagic So, the config needs to be EDT? My gosh I am confused
15:12 Dyrcona Well, are you in Eastern or Central timezone?
15:12 Bmagic Eastern
15:12 Dyrcona sudo dkpg-reconfigure tzdata
15:13 Dyrcona dpkg not dkpg....
15:14 Bmagic I type it backwards all the time
15:14 Dyrcona Then pick America, then New York, and done.
15:14 Dyrcona Then, you may have to restart everything. :)
15:14 Bmagic those were the options already selected
15:15 csharp npm run build-prod fun: https://pastebin.com/cF6wVVVh
15:15 csharp Module not found: Error: Can't resolve 'angular-tree-control/context-menu.js' in '/home/opensrf/Evergreen/Open​-ILS/web/js/ui/default/staff'
15:15 berick show timezone; reports 'localtime' for me fwiw.
15:16 * csharp shows 'localtime' too
15:16 Bmagic yeah, postgres is configured with localtime, so it's using /etc/timezone or maybe /etc/localtime
15:16 Dyrcona csharp: How did you install Node.js?
15:17 berick just tried 'npm run build-prod' (and 'build') on current master w/o errors.
15:18 * Dyrcona suspects an old version of Node.js, like from packages.
15:19 Dyrcona Reconfiguring tzdata should take care of /etc/timezone and /etc/localtime.
15:19 Dyrcona Assuming you're on a Debian-based distro, which is all we support.
15:21 Bmagic Dyrcona: yep
15:21 berick tested with npm 3.10.10 / node v6.11.3
15:21 jeff what makes you think that your timezone settings are off?
15:22 csharp Dyrcona: I'll check it out - this is a server with previous nodejs stuff on it
15:24 csharp looks like I have an older npm installed
15:25 csharp hmm - this is npm  5.7.1 and I can't seem to update it
15:27 Dyrcona Now, it sounds like you too recent a version of Node.
15:28 Dyrcona How'd you get npm 5, I wonder.
15:29 Dyrcona Theoretically, it should work.
15:30 berick csharp: npm -v OR node -v ?
15:33 Dyrcona berick: npm 5.7.1 is the latest, unless you count 5.8.0-next.0
15:33 berick yeah, was just checking on that...
15:33 Dyrcona npm install npm@latest -g
15:33 Dyrcona that would install it.
15:35 jvwoolf joined #evergreen
15:37 khuckins joined #evergreen
15:59 kmlussier @quote random
15:59 pinesol_green kmlussier: Quote #167: "<berick> the morning sun is shining like one big.. ball of circs" (added by gmcharlt at 01:02 PM, June 01, 2017)
15:59 * gmcharlt checks the archive, wondering what the heck the context was for THAT
16:00 berick let me know if you find it
16:01 gmcharlt heh, it seems I set it up for you, and you knocked it down
16:01 gmcharlt http://irc.evergreen-ils.org/​evergreen/2017-06-01#i_308303
16:02 gmcharlt "action.all_circulation_slim" seems so disapointing as a name in comparison ;)
16:04 berick heh, true
16:05 berick @tea [band]
16:05 * pinesol_green brews and pours a pot of English Breakfast, and sends it sliding down the bar to Ah Well, Nevermind (http://ratetea.com/tea/harn​ey/english-breakfast/3525/)
16:08 yboston joined #evergreen
16:10 yboston joined #evergreen
16:10 Dyrcona @coffe [band]
16:10 pinesol_green Dyrcona: I see nothing, I know nothing!
16:10 Dyrcona @coffee [band]
16:11 * pinesol_green brews and pours a cup of El Salvador Finca Kilimanjaro, and sends it sliding down the bar to Judgmental Roomba
16:13 kmlussier @blame [dessert]
16:13 * pinesol_green I know it was you, grabs some pineapple coconut cheesecake. You broke kmlussier's heart. You broke kmlussier's heart. for kmlussier
16:14 berick dizzying ;)
16:16 gmcharlt ok, I guess /that's/ off the menu ;)
16:18 yboston joined #evergreen
16:28 pinesol_green [evergreen|Morkor Quarshie] Docs: Add directions for configuring columns - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=3a118d6>
16:56 yboston joined #evergreen
17:01 jeffdavis What's the right way in Launchpad to target a bug to multiple milestones?
17:04 gmcharlt jeffdavis: first, select each of the relevant series (3.0, 2.12, etc.), then the appropriate milestone within each one
17:06 jeffdavis Thanks. I get confused when I have the option to have a single milestone for "Evergreen" but then separate milestones for targeted series (3.0, 3.1 etc).
17:27 remingtron_ joined #evergreen
17:29 abowling joined #evergreen
17:29 yboston joined #evergreen
17:34 yboston joined #evergreen
17:43 abowling1 joined #evergreen
17:48 hbrennan joined #evergreen
18:00 abowling joined #evergreen
18:14 mmorgan left #evergreen
18:15 beanjammin joined #evergreen
18:31 pinesol_green News from qatests: Failed Running perl live tests <http://testing.evergreen-ils.org/~live>
18:40 yboston joined #evergreen
18:47 beanjammin joined #evergreen
21:41 yboston joined #evergreen
23:45 yboston joined #evergreen
23:54 yboston joined #evergreen

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