Evergreen ILS Website

Search in #evergreen

Channels | #evergreen index




Results

Result pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148

Results for 2015-08-17

10:34 Dyrcona With BOM, of course.
10:36 RoganH joined #evergreen
10:38 jboyer-laptaupe Dyrcona, I'm trying to output something that won't make yaz-marcdump throw up, I don't think that would do it. ;)
10:39 Stompro testing multiple bugs display, sorry for the spam,  bugs 1485240, 1485240
10:39 pinesol_green Launchpad bug 1485240 in Evergreen "Remove last vestiges of script-based circ rules" (affected: 1, heat: 6) [Undecided,Confirmed] https://launchpad.net/bugs/1485240
10:39 Stompro testing multiple bugs display, sorry for the spam,  bugs 1485240, 1484655
10:39 pinesol_green Launchpad bug 1484655 in Evergreen 2.8 "ftp://ftp.mozilla.org has moved to http://archive.mozilla.org" (affected: 1, heat: 6) [High,Fix committed] https://launchpad.net/bugs/1484655
10:40 Stompro Hmm, that didn't work.
10:40 Stompro Unless it watches what it just did, and doesn't repeat bugs that were recently displayed.
10:43 Dyrcona jboyer-laptaupe: Fun thing, is a lot of MARC are actually ISO-8859-1 or some MS encoding when they should be MARC-8.
10:44 Dyrcona jboyer-laptaupe: Good luck! :)
10:44 bshum Stompro: pretty sure it doesn't repeat bugs for X time where X is beyond my recollection.
10:45 Stompro I'm testing over at #openils-evergreen, and the multiple bug display worked fine there, so it was just refusing to display the same bug again immediatly.
10:46 bshum Indeed
10:48 jboyer-laptaupe Dyrcona: Hah! looks like what I'm seeing isn't my problem anyway. Just trying to use yaz-marcdump to break records into line mode and then rebuild them as marc doesn't work (at least not with these Overdrive records...)
10:50 mmorgan Anyone seen anything like this:
11:08 Dyrcona I have a local ticket that I made about it in June of last year, but never made a launchpad bug.
11:08 jeff and indeed, this bucket has four items, but one of the items is there 15 times and the other is there 41 times.
11:09 Dyrcona I started working on a script to clean them up and a possible solution in the UI, but other things came up.
11:10 Stompro testing git integration, looking at commit 0f8ec1
11:10 pinesol_green [evergreen|Bill Ott] LP#1394989: Do not include deleted users when retrieving for Collections - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=0f8ec14>
11:10 ericar_ joined #evergreen
11:10 Dyrcona By "started working on," I probably mean just thinking about an implementation.
11:14 * jeff nods
11:14 Dyrcona There is a launchpad bug about the lack of feedback on that interface.
11:14 Dyrcona Fixing that would go some way toward helping.
11:15 Dyrcona My testing didn't reveal any scenarios other than a patron clicking the button 41 times that would cause it to get added 41 times.
11:16 Dyrcona The dates on the bucket entries also ruled out an older bug.
11:17 jeff looks like about three dupes per second on this example here.
11:18 Dyrcona That sounds a bit fast for a human, but not impossible.
11:40 Dyrcona jeff: I'll see if I can condense the stuff from the links about into a lp bug today.
11:40 Dyrcona s/about/above/
11:41 Dyrcona "kilometers" of paper....I'm curious: how many kilometers has it printed?
11:42 jeff 0.044 km, 1249 cuts.
11:42 jeff this particular model has mostly been used for testing.
11:42 jeff and printing names for a hat for summer reading drawings.
11:42 jeff so, lots of short jobs.
11:42 Dyrcona ok. not as interesting as one from a circ. desk.
11:43 jeff right.
11:43 jeff now i wonder about the printer embedded in the self checkout kiosks.
16:53 mrpeters ah yes, it is -- config.org_unit_setting_type
16:56 mrpeters so i'd propose a simple UPDATE config.org_unit_setting_type SET description = 'Regular expression defining the required password format when resetting passwords.  Note:  Adjust description in config.tt2 to suit your requirements.' WHERE name='global.password_regex'; if that kind of database change is permitted at this stage in the game
16:58 Stompro yboston, the subject of your email to the dig list about the irc practice has a different date & time than the body.  Subject says 26th 3pm eastern, body says June 24th 2pm est.
17:01 yboston Stompro: that is what I get for copy pasting
17:06 yboston Stompro: I also has the date wrong from the text I pasted
17:07 yboston Stompro: BTW, we shoudl coordinate to sign-off on each other's PGTAP tests at soem point this week
17:08 mmorgan left #evergreen
17:10 pinesol_green Incoming from qatests: Test Failure - http://testing.evergreen-ils.org/~live/test.html <http://testing.evergreen-ils.org/~live/test.html>
17:12 jeff Additional error details:

Results for 2015-08-16

03:05 b_bonner_ joined #evergreen
03:05 csharp_ joined #evergreen
03:10 rashma_away joined #evergreen
05:12 pinesol_green Incoming from qatests: Test Success - http://testing.evergreen-ils.org/~live/test.html <http://testing.evergreen-ils.org/~live/test.html>
08:11 rlefaive joined #evergreen
09:31 Dyrcona joined #evergreen
09:43 Dyrcona joined #evergreen
10:40 Dyrcona joined #evergreen
11:14 Dyrcona joined #evergreen
11:43 Dyrcona @dunno
11:43 pinesol_green Dyrcona: What do you mean? An African or European swallow?
16:54 pinesol_green Incoming from qatests: Test Success - http://testing.evergreen-ils.org/~live/test.html <http://testing.evergreen-ils.org/~live/test.html>
21:10 jonadab joined #evergreen
22:01 dluch joined #evergreen
22:16 dluch joined #evergreen

Results for 2015-08-15

04:56 pinesol_green Incoming from qatests: Test Success - http://testing.evergreen-ils.org/~live/test.html <http://testing.evergreen-ils.org/~live/test.html>
10:37 Dyrcona joined #evergreen
15:00 Dyrcona Whee! Found some last vestiges of legacy circ scripts to get rid of.
15:01 Dyrcona Seventeen fewer lines in Circulate.pm.

Results for 2015-08-14

00:59 book` joined #evergreen
04:23 mtj_ joined #evergreen
04:41 pinesol_green Incoming from qatests: Test Failure - http://testing.evergreen-ils.org/~live/test.html <http://testing.evergreen-ils.org/~live/test.html>
07:23 mrpeters joined #evergreen
07:34 collum joined #evergreen
07:44 rjackson_isl joined #evergreen
08:29 dkyle joined #evergreen
08:40 mmorgan joined #evergreen
08:42 maryj joined #evergreen
08:44 Stompro Happy test writing day everyone, Woohoo
08:44 jboyer-isl miker: In that situation, wouldn’t the logs that would be backed up just be lost? That’s obviously preferable to knocking app servers offline, but still not ideal I would think.
08:55 ericar joined #evergreen
08:56 * csharp confirms that PINES is using UDP to ship logs
09:34 miker jboyer-isl: not necessarily. depends on buffer settings at various levels
09:36 jboyer-isl Ah, ok.
09:51 rfrasur Soooo, does this mean no one is volunteering to set up the festival canopy?
09:56 berick i think you can also configure rsyslog to drop new messages when its outbound queueu fills up instead of throttling the message submitter
09:57 berick as a last layer of defense
10:01 * berick puts on his test-writing pants
10:04 ldw How do we want to work on tests today? Do people want to take a bug and work on it idependently?  Do we need to discuss pgTap or test writing in general?
10:04 ldw s/idependently/independently/
10:10 Christineb joined #evergreen
10:13 Shae joined #evergreen
10:14 Stompro I think I understand what needs to be done for the evergreen.lowercase test, I just have to get things setup so I can test it.  I'll let you know if I run into anything.
10:18 berick i'm going off-script a little bit and doing a perl live test.  will return to pgtap after that.
10:18 jboyer-isl I’ll be pretty in and out as far as test writing day goes ( :(  ) but I would like to recommend that tests include things expected to fail. I don’t mean tests that fail, I mean writing the tests so that they basically say “This sql/function/etc. should fail and that’s ok. If it doesn’t fail, fail this test.”
10:20 ldw jboyer-isl: do you mean pass in data to the function that should fail it and check for the fail?
10:21 jeff that's how i interpreted his statement.
10:22 jboyer-isl ldw: Yes. It depends on what’s be tested of course, it doesn’t make sense for every test. The thinking is that if the test starts failing, something has broken in the function. (it’s not validating arguments correctly, it’s breaking a contract that it will never return NULL, etc.)
10:36 Stompro Debian Wheeze has a package for pgtap 0.90 released in 2011... I wonder if that will work?  I'm getting errors " ERROR:  function isnt_empty(unknown, unknown) does not exist"
10:37 ldw Stompro: you need to run CREATE EXTENSION pgtap; on your database.
10:37 ldw I am running into issues with pgTap and Postgres 9.4.  I am going to fall back to 9.1 for today.
10:40 Stompro I'm just running pg_prove,  "pg_prove -U evergreen -d egdb -h localhost t"
10:45 Stompro Do you install the pgtap extension for just your Evergreen db, or globally, for template1?
10:46 ldw Stompro: your own database
10:46 Stompro Running \dx shows "pgtap     | 0.90.0  | evergreen  | Unit testing for PostgreSQL
10:46 Stompro "
10:47 jboyer-isl Stompro: I believe things added to template1 only affect databases created after template1 is changed. Existing dbs are unaffected.
10:47 Stompro jboyer-isl, makes sense, thanks.
10:48 jboyer-isl The answer appears to be that pgtap .90 does not include isnt_empty, only is_empty: https://github.com/theory/pgtap​/blob/v0.90.0/sql/pgtap.sql.in
10:48 jboyer-isl That’s unfortunate. :(
10:50 Stompro I'll switch to a debian jessie test system with a newer version, thanks for finding the answer jboyer-isl++
10:50 Dyrcona Stompro: You could try installing a more recent pgtap from source, just sayin'.
10:52 Stompro I know, I just want to get going on writing the tests and not mucking with building pgtap today.
11:01 Stompro Sigh, no debian package for pgtap in Jessie, I guess I will be building it.
11:06 jeff looks like the pgtap package in debian is not well maintained. as such, it was removed from testing back in 2013.
11:06 * berick grabs 1483508
11:06 dbs http://apt.postgresql.org/pub​/repos/apt/pool/main/p/pgtap/ may be an option
11:07 csharp I don't mean to distract from all the test-writing awesomeness, but can someone confirm this behavior on a 2.7+ system using built-in selfcheck?...
11:07 * jeff waits for csharp's question to be about usernames/barcodes
11:07 csharp jeff: yep ;-)
11:07 dbs but then you're into the whole "use postgresql's apt repos instead of debian/ubuntu's" thing
11:08 csharp with the Patron Barcode Format YAOUS set to ^[0-9]+$, when I enter a barcode or username, it appears to accept it, but doesn't accept the password after it's typed
11:08 dbs http://wiki.postgresql.org/wiki/Apt fwiw
11:08 csharp the web console shows that the regex works
11:08 csharp testing <barcode#> ... barcode
11:08 berick did we decide we can stop adding all of the \pset boilerplate at the top of these tests, since it's not needed with pg_prove ?
11:09 csharp I see the ****** (dots) appear for the password, but Enter doesn't work
11:10 csharp if I remove the regex from the YAOUS, it works fine
11:10 Stompro Ok, building pgtap is about as simple as installing the package...  I'm going to add a note to the qa wiki page.
11:59 bmills joined #evergreen
12:03 jihpringle joined #evergreen
12:07 Stompro Is there a trick to pasting unicode characters into putty?  I can view them fine, but I cannot copy them.
12:07 pastebot "yboston" at 64.57.241.14 pasted "first pgTAP test attempt" (42 lines) at http://paste.evergreen-ils.org/24
12:08 yboston I was wndering if I could get some feedback on my first pgtap test. I am testing a function called public.first_agg from Open-ILS/src/sql/Pg/002.functions.aggregate.sql
12:08 yboston LP 1483506
12:08 pinesol_green Launchpad bug 1483506 in Evergreen "public.first_agg needs test" (affected: 1, heat: 6) [Undecided,New] https://launchpad.net/bugs/1483506 - Assigned to Yamil (ysuarez)
12:08 yboston what file name should I use?
12:10 berick yboston: file name for the test?
12:10 yboston berick: yes
12:11 yboston also, not sure how the actual function is being used
12:11 berick yboston: something like Open-ILS/src/sql/Pg/t/lp1483506-first-agg.pg should do
12:11 yboston to help me test it better
12:11 berick see other examples in Open-ILS/src/sql/Pg/t/
12:11 yboston OK, was not sure if I shuld put the LP number or not
12:12 berick doesn't really matter, but may as well
12:13 yboston the LP bug was created for todays "test day," the LP bug does not have any significant content
12:13 berick so, first_agg() is a very simple function.  pass in 2 things, it returns the first one that's not NULL.  if they're both null, it returns NULL
12:14 yboston sorry, I meant I was havign a hard time finding it used int he code. Might have grepped wrong
12:14 yboston I am not sure if I shuld be testing with more than simple data types. not sure if I should test with passing arrays(?)
12:14 berick hm, don't think it is used in the code.
12:14 berick see no references to it
12:15 yboston berick: thanks, I feel less crazy now
12:15 yboston maybe in the past it was used, but with all the cleanign that has been done it is no longer needed
12:16 berick maybe
12:17 yboston berick: thanks for the feedback, I feel less crazy now.
12:21 dbs berick: yboston: maybe no lp number in the filename, so that we can add more tests to the same file later?
12:21 yboston dbs: makes sense
12:22 yboston what about file name structure, shoudl I prepend the SQL file name then add the fucntion name?
12:22 berick yeah, that makes sense.  perhaps we should clean up the existing ones too
12:22 berick re: no lp number
12:23 berick yboston: hm, i would just do the function name
12:23 yboston berick: good idea
12:24 yboston can we use subfolders>
12:24 yboston ?
12:24 yboston berick: I think first_agg gets used by calls to first()
12:24 yboston if so, I have more ideas of how to test it
12:25 berick yes, in fact it does
12:25 yboston may add a test for first() too
12:25 berick good catch
12:27 Stompro Hmm, evergreen.lowercase doesn't seem to handle the two turkish I's correctly.
12:28 yboston so the two ideas are soemthing like  1) 002.functions.aggregate.sql_first_agg.pg  2) first_agg.pg
12:30 ldw I like subfolders.  But pg_prove needs to be run with --recurse for that to work.
12:30 yboston berick: makes sense
12:30 yboston ldw: good to know
12:30 berick i mean, if we're going to use the file name, we just need to also change the test name, if we're going to include the file name in the test
12:30 berick should a function move
12:31 berick maybe that's very rare, not worth worrying about
12:31 berick anyway, it popped in my head, and i blurted sans filter :)
12:31 yboston I think we don't have to come up with a solid decision right now, I just wanted to run it by soem folks to get the cnversation started
12:32 berick hm, i should add an aggregate test to my text_concat test as well
12:33 csharp okay - I can confirm the same "set Patron Barcode Format setting and bork selfcheck" issue on a test server running master
12:33 * csharp runs off to open the bug report
12:34 yboston BTW, since pg_proves needs --recurse for subfolders to work, I will avoid using a subfolder
12:34 yboston I look forward tpo having so many tests that we need to start using subfolders
12:35 berick yboston++ # fixing problems before they become problems
12:37 yboston berick for your feedback
12:37 yboston berick++
13:53 dbwells Right, it only affects circs where there is an ungenerated fine at checkin, and the balance was zero before checkin.
13:53 bshum Oh nevermind, I see you assigned yourself.
13:53 bshum You're on top of it
14:39 phasefx dbwells: unrelated, but in the same area, it looks like that having that lost/zero setting could prevent fines from being generated at all during checkin
14:40 * phasefx hasn't actually tested that yet though
14:41 dbwells phasefx: Thanks.  I've got a fix for the first part (just writing a test), but I'll need to think more on that.  The idea there is to purposely not generate fines if it is a lost checkin with currently zero fines, as that is what I understand the setting to mean.
14:42 bshum dbwells: yeah I think that's the new setting that miker added for keeping stuff at zero.  Stop gap for negative balance stuff :P
14:43 phasefx dbwells: I did put some logging statements in testing the reported bug, and the handle_fines that gets called is the one inside if (!$dont_change_lost_zero) {
14:44 * phasefx squints, okay, it's probably safe, since !$dont_change_lost_zero isn't just based on the setting, but looks like it considers the actual copy status, etc
14:45 buzzy joined #evergreen
14:45 dbwells right.  It tries to just short-circuit the whole thing, and I think it's right, or at least close ;)
14:53 bshum Okay...
15:00 bshum Now of course, I have more users than the list i was given, but I can work with that.
15:00 bshum dbwells++ # thanks!
15:01 dbwells money.billable_xact is the "parent" table of action.circulation, money.grocery, and booking.reservation, so they all share a pool of ids.
15:04 bshum After we remove or void those entries, part of me thinks I should do something to refresh the user summaries
15:05 bshum Ah, there's a trigger on money.billing for that sort of thing, maybe...
15:06 * bshum checks
15:06 bshum (and tests)
15:14 bshum Oh happy day.
15:21 jboyer-isl Bmagic: if you are still looking into ways to relieve rsyslog congestion, look up the $ActionQueueType, $ActionQueueFileName, $SystemLogRateLimitInterval, and $SystemLogRateLimitBurst directives.
15:22 jboyer-isl On a related note, fulling turning on that firehose for our database is yielding about 20%+ larger logs than letting the rate limiter eat thousands of what are likely 0ms queries.
15:25 ldw jboyer-isl: I am writing negative tests, but I am using the isnt function to make sure they pass.  This results in all tests passing.  Does this seem reasonable to you, or is there a specific reason you had for wanting a test to fail?
15:26 jboyer-isl ldw: that’s exactly what I meant. if you just run pg_prove it should say all tests pass. A test that passes because it’s expecting a function to fail still passes. :)
15:26 ldw ok cool :)
15:39 Stompro ldw, I'm looking at the pgtap test for change_db_setting , which calls alter database... but it looks like changes to alter database only take effect for new sessions.  Which would be outside the testing transaction, so I don't think that is testable.   What do you think?
15:39 Bmagic jboyer-isl: thanks! for now, I am going to leave at local logging. I am afraid of bringing it all down again. Makes for an unpleasant day!
15:39 ldw Stompro: you should enclose your tests in a BEGIN and ROLLBACK block, so the changes take effect for the duration of the test, but are not permanent.
15:40 jlitrell joined #evergreen
15:40 ldw Stompro: I see what you are saying
15:40 ldw hmmm
15:40 jeff evergreen.change_db_setting is involved in the search_path mess.
15:40 Stompro jeff, that is the only thing it is used for that I can see.
15:41 jeff i believe it is only used to change the db search_path, and potentially in a way that doesn't account for the fact that such changes are not immediate. i haven't dug into it further/recently.
15:41 jeff but i remember it from my search_path digging.
15:41 Stompro So I was trying to change the search path, and then show the search path, but if you want to change the search_path for the current session you use 'set' , not alter db.
15:41 * jeff nods
15:42 jeff if pg_tap supports disconnecting and reconnecting between test steps, that might be useful. if not, then this might not be a great function to test Right Now.
15:43 Stompro But would the transaction persist across connections?
15:43 jeff no.
15:44 ldw could it be SELECTed from the SCHEMA instead of specifying a value to expect, specify a SQL statement?  I am not sure if you can do that.
15:44 Stompro and the change would rollback at the disconnect.  So lets skip that one.
15:45 Stompro Maybe, I'll dig and see where those values are stored.
15:47 ldw Stompro: I am not sure if the second paramter from the tests can be a SQL statement.  I might be wasting your time if you are digging through the schema.  I am trying to see if I can get a SELECT statement to run as the second paramter.  But, no luck so far.
15:47 Stompro I was using results_eq  Which takes a query and what the results should match.
15:48 ldw you could run a \set after the alter to set a variable to the value of a select statement and use that value in your test.
15:59 ldw Stompro: hmm, I cannot get \set to equal a value of a SELECT statement either.
16:00 ldw Stompro: you can run a query as the first paramter to a pgTap test if it is enclosed in ().
16:01 ldw So, if you know where that value is stored, you could check it that way.
16:01 Stompro All I can find is the pg_settings virtual table, but that doesn't reflect changes made by alter database, since they only take effect for new connections after the change.
16:08 ldw Stompro: you could try this SELECT pg_reload_conf();
16:09 ldw We need some no-op command for pgTap though
16:10 Stompro pg_reload_conf cannot be run in a transaction.
16:10 jeff i don't think that pg_reload_conf is what you want here.
16:11 jeff ``pg_reload_conf sends a SIGHUP signal to the server, causing configuration files to be reloaded by all server processes.''
16:11 Stompro Do we really need to test to make sure Alter Database works... can't we leave that to the postgresql guys.
16:11 ldw Stompro: no
16:11 ldw no we do not need to test it
16:12 jeff i repeat my earlier: "this might not be a great function to test Right Now" :-)
16:13 ldw I like to battle with the computer some days :/.
16:16 Stompro jeff++, got it, I'm moving on.  ldw++ thanks for the suggestions.
16:23 dkyle left #evergreen
16:28 ldw How should we go about including these tests in a release?
16:31 Stompro From reading the pgtap docs, we could include the pgtap.sql file with Evergreen, and the tests could run during eg_db_config.
16:34 Stompro Are the pgtap tests part of the current QA run?
16:34 ldw Stompro: I am more thinking about the EG release process.  Presumably we will have a number of tests attached to LP bugs, I should then target them to a release, but I am not sure which release.
16:34 ldw Stompro: I do not know about the current QA run.
16:36 Stompro A LP targeted test would be proving that the fix for that bug works, correct?  So the test would need the fix applied before it would work?
16:37 Dyrcona Stompro: Usually. I could also see a test that could reveal the problem before it is fixed being useful, but that could be the same thing.
16:38 Dyrcona Most of the pgTAP tests I've written for LP bugs have been to make sure new indexes and fields exist after the upgrade script is run, for instance.
16:40 ldw In this case though, we are writing tests for functions tha alread exist.  To prevent against regression.  But we still need the tests to get commited to the master branch at somepoint.  Dyrcona: is a pull request enough for that?
16:41 Dyrcona Yes it is. Someone will merge it, usually after running the tests, themselves.
16:42 ldw Stompro: do you know how to add a pullrequest tag to LP bugs?
16:42 Stompro Yep, I added it to the one that I completed today.
16:42 ldw awesome. I have added one to mine as well.
16:43 Stompro yboston, did you get yours done today?
16:44 Stompro Just checked bug log, yes he did.
16:47 ldw I have 6 tests listed with test-writing-day-0 and a pullrequest tag.  Not a bad first showing.  Hopefully, we can get more participation during the next one.  Does a test writing day every three months sound too onerous?
16:48 Stompro Not to me, that sounds just fine.
16:54 dbwells ldw: I did write a test today for bug #1484989, if that counts :)  Would have liked to do more.
16:54 pinesol_green Launchpad bug 1484989 in Evergreen "Fines are not calculating until after circulation is closed" (affected: 1, heat: 6) [Medium,Incomplete] https://launchpad.net/bugs/1484989
16:54 ldw dbwells: that counts! Thank you.
17:00 yboston lwd & Stompro there was a question earlier about running a select on the second parameter, if you mean the second param of is(), I am doing it in my test.
17:01 yboston just had to wrap it in parens
17:01 yboston see here... http://git.evergreen-ils.org/?p=working/Evergre​en.git;a=blob;f=Open-ILS/src/sql/Pg/t/public.fi​rst.pg;h=31941b3b21f2c69d8e3125fd6e34b8f74954d4​37;hb=3bcef83d5438fee4355c7eaa3ea304f271cac9b5
17:04 Stompro Thanks yboston, have a good weekend.
17:10 mmorgan left #evergreen
17:14 pinesol_green Incoming from qatests: Test Success - http://testing.evergreen-ils.org/~live/test.html <http://testing.evergreen-ils.org/~live/test.html>
17:53 vlewis joined #evergreen
18:39 mrpeters left #evergreen
18:40 bmills joined #evergreen

Results for 2015-08-13

02:16 jonadab_znc joined #evergreen
03:22 gsams joined #evergreen
03:24 bshum @later tell dbwells Curious to know if you had any time to poke at https://bugs.launchpad.net/evergreen/+bug/1379815.  We've been thinking about adding copy stat cats more permanently into some of our workflows, but can't if Vandelay doesn't support something like it.
03:24 pinesol_green bshum: The operation succeeded.
03:24 pinesol_green Launchpad bug 1379815 in Evergreen "Assign stat cats during Vandelay import/overlay of items" (affected: 1, heat: 8) [Wishlist,New] - Assigned to Remington Steed (rjs7)
03:24 jcamins joined #evergreen
03:43 gsams joined #evergreen
03:43 Mark__T joined #evergreen
03:47 dbs joined #evergreen
03:53 mnsri_away joined #evergreen
04:50 Mark__T joined #evergreen
05:16 pinesol_green Incoming from qatests: Test Success - http://testing.evergreen-ils.org/~live/test.html <http://testing.evergreen-ils.org/~live/test.html>
06:14 rlefaive joined #evergreen
07:15 artunit_away joined #evergreen
08:02 gsams joined #evergreen
11:56 Bmagic once I removed that config and restarted rsyslog on the app servers, the log file stopped adding those error messages
12:00 jeff logical. your app servers were trying to log to files that they didn't need to log to, and their errors about not being able to do so were being logged on the target/remote log host. :-)
12:01 Bmagic jeff: I think that was it, time will tell of course
12:02 Bmagic jeff: We have been having weird issues and these logs might be a clue. Every week for the last 4 weeks, ALL of our app servers get taken out of LVS rotation because apache stops responding to the test from lvs
12:02 Bmagic Do you think that an rsyslog misconfiguration could possible pile up high enough to bring apache down to a slow enough crawl?
12:03 bshum mrpeters: What do you mean when you say xpath issues?
12:04 mrpeters ive done various ways of doing pg dump and it seems like i had some kind of issue with xpath or search path or something like that
13:31 Christineb I can't stress how happy this made me :D
13:31 jihpringle it's slated for 2.9 but we've backported to 2.8 and have been using it since July
13:31 mcrawley don't know what that one is?
13:33 lstjohn we are also about to sign off on #1402018 workstation scoping - we've tested it on 2.7.2
13:33 jihpringle the acquisitions admin menu was moved to be under the Admin menu rather than under Admin -> Server Administration
13:33 Christineb The Acquisitions Admin menu has been moved directly under Admin instead of having to to into Admin --> Server Admin
13:33 Christineb lstjohn - that is great news!!
13:40 lstjohn I have a question. When you delete a line item from a purchase order that has not been activated, it deletes the line item from the selection list but not from the OPAC - is this your experience also?
13:40 Christineb lstjohn - I have seen this happen
13:41 Christineb when a MARC order file is loaded in "Load MARC Order Records" and Load Items for Imported Records is selected
13:42 Christineb Is on my to-do list to do some more testing to figure out if this is a bug or desired behaviou
13:42 Christineb does this scenario match what you are seeing?
13:44 lstjohn Yes, order records loaded, copies created  - there seems to be a related bug https://bugs.launchpad.net/evergreen/+bug/1269574
13:44 pinesol_green Launchpad bug 1269574 in Evergreen "ACQ lineitems canceled via EDI not deleting linked bibs/items" (affected: 5, heat: 26) [Medium,Confirmed]
15:29 Dyrcona jeff csharp bshum: lp 1484655
15:29 pinesol_green Launchpad bug 1484655 in Evergreen "ftp://ftp.mozilla.org has moved to http://archive.mozilla.org" (affected: 1, heat: 6) [High,New] https://launchpad.net/bugs/1484655 - Assigned to Jason Stephenson (jstephenson)
15:29 Dyrcona Oops. Should unassign myself. ;)
15:30 jeff say, why didn't live tests break? because it doesn't build a staff client?
15:31 Dyrcona I successfully built a client last week, so don't know when this actually changed.
15:31 Dyrcona Also, maybe it doesn't build a client.
15:32 jeff timestamp on the README in question is "8/7/15, 10:22:00 PM"
15:33 Dyrcona Missed it by a couple of days.
16:25 ericar joined #evergreen
16:35 jihpringle joined #evergreen
16:59 pinesol_green Incoming from qatests: Test Success - http://testing.evergreen-ils.org/~live/test.html <http://testing.evergreen-ils.org/~live/test.html>
17:01 Dyrcona jeff: Looking through the test logs above, it doesn't appear to make a client. It just does the server-side bits.
17:01 * Dyrcona disappears.
17:08 mmorgan left #evergreen
17:32 jlundgren left #evergreen
18:00 dbwells rfrasur: what's up?
18:02 rfrasur I think Rogan is taking care of it.  Maybe?  Thank you :-)
18:02 dbwells okay :)
18:04 berick Bmagic: yes, i think requesting a static file from your apache server via http is a common way to confirm it's alive
18:04 berick Bmagic: many load balancers use that as a way to take server in/out of rotation.  request a file (e.g. /openils/var/web/ping.tx) and if it's not there, the server is removed from rotation
18:05 berick well, ping.txt (but i imagine it can be anything)
18:06 berick that style of test is great because you can take a server out w/o having to touch the load balancer
18:37 rfrasur Okay, I'm thinking through some stuff for our library.  In a consortium...or library system...can you hide all the collections of a branch from the rest of the library system?
18:38 rfrasur I mean, using Evergreen...not a book truck and a closet.
18:41 buzzy joined #evergreen

Results for 2015-08-12

01:16 Mark__T joined #evergreen
05:01 pinesol_green Incoming from qatests: Test Success - http://testing.evergreen-ils.org/~live/test.html <http://testing.evergreen-ils.org/~live/test.html>
07:18 mrpeters joined #evergreen
07:36 sarabee joined #evergreen
07:37 rjackson_isl joined #evergreen
11:45 jeff berick: ah, good. logical, even! (sorting the params)
11:46 Stompro Bmagic, how are your page loads for other types of results.  How long does it take to load a record detail page?
11:46 berick at debug level, you can see in the logs if a search is loaded from cache, fyi
11:46 Bmagic this test still has CALL: open-ils.cstore open-ils.cstore.direct.config.​coded_value_map.search.atomic {"id":{"!=":null}}
11:46 Bmagic and it was for sure the exact same search
11:47 berick Bmagic: you may have to run multiples searches before the coded_value_map call goes away
11:47 berick depends on the number of apache backends you have running

Results for 2015-08-11

02:23 Mark__T joined #evergreen
04:48 Mark__T joined #evergreen
05:12 artunit_ joined #evergreen
06:43 csharp ldw++ # test wrangling
07:37 Callender joined #evergreen
07:45 TaraC joined #evergreen
07:46 mrpeters joined #evergreen
08:50 krvmga in eg 2.5 it was easy to change the order of search format filters in the opac dropdown list. since 2.6, they have been listed alphabetically. is it possible to change the order of the filters somehow other than by alphabetical manipulation?
08:56 mmorgan joined #evergreen
09:14 jwoodard joined #evergreen
09:25 Stompro mmorgan++ thanks for testing the patron expiration notice.. now to get the release notes done.
09:25 yboston joined #evergreen
09:27 * Dyrcona wonders if hopkinsju had any success with backstage yesterday evening.
09:27 mmorgan Stompro++ Thanks for polishing the code, and (in advance) for the release notes!
09:29 ericar_ joined #evergreen
09:35 collum joined #evergreen
10:07 pmurray_away joined #evergreen
10:10 Stompro ldw++, thanks for all the test writing bugs, I assigned myself to the simplest one I could find :-)
10:23 krvmga is there any way to change the order of search format filters in the opac other than by playing with the alphabetical listing?
10:29 dbwells krvmga: I don't think there is anything built in.  We ordered ours as you are thinking by manipulating the strings.  In our case, we prepended a certain number of spaces to the "Search Label"s.  The spaces do not show up.
10:29 krvmga dbwells: can i look at your catalog url?
14:30 Bmagic I was having trouble finding a report about the in-house barcode box not clearing the box after the enter button was pressed (ready to accept the next barcode)
14:39 Bmagic https://bugs.launchpad.net/evergreen/+bug/1483857
14:39 pinesol_green Launchpad bug 1483857 in Evergreen "Web based staff client in house use barcode box clear" (affected: 1, heat: 6) [Undecided,New]
14:59 ldw Stompro: thank you for taking a bug.  Have you worked with the pgTap testing suite before?
15:00 Stompro ldw, Never, All I know about it is from reading the docs that you linked to previously.
15:00 rlefaive joined #evergreen
15:03 ldw Stompro: If you have any questions about it do not hesitate to ask.
15:04 akilsdonk joined #evergreen
15:17 jboyer-isl I don't think I've asked this before, but what are the thoughts re: tests and simply adding new seed data? Don't bother, check to see if a few values are present, check every value, etc?
15:18 berick jboyer-isl: for example?
15:18 jboyer-isl I'm looking at LP #1371647 and have a large upgrade script ready to go, but I haven't decided what to do about a test case.
15:18 pinesol_green Launchpad bug 1371647 in Evergreen "config.marc21_ff_pos_map needs an audit" (affected: 1, heat: 6) [Undecided,New] https://launchpad.net/bugs/1371647
15:20 jboyer-isl berick: example incoming
15:21 jboyer-isl Bah, it would have, but I mistyped my nick...
15:22 berick sorry, i was confuzzled
15:23 berick i've just been checking a few values, like a basic sanity check.
15:23 berick don't think we have a policy on that
15:25 jboyer-isl Excellent. I don't mind spot checking a few things to be sure the patch was applied, but a "full check" would be a huge pain to write for things like that.
15:26 jboyer-isl I'll try to get something put together and posted for proper discussion/testing asap; this is holding up some format search improvements that we're trying to do and I don't want it to be a local only fix.
15:27 Dyrcona Well, you could use a macro in your text editor to write the tests for you... I've done that recently from other upgrade scripts.
15:30 berick you're gonna need a unit test on that macro
15:32 Dyrcona You're gonna need a unit test on your unit test, too. :)
15:33 Dyrcona But actually, I didn't use a macro just a regex search and replace. I misspoke.
15:45 jboyer-isl That's true, but then there's only full coverage for a chunk of data. If there's ever interest in fully verifying seed data, I'd rather see one big test that checks it all (one file per schema/table, not necessarily the entire seed database)
15:52 Dyrcona jboyer-isl: That sounds reasonable to me. When do you start working on it? :P
15:52 Dyrcona j/k
15:54 jboyer-isl Well, I suppose someone could find and replace 950.config.seed-data.sql, heh.
16:12 jlitrell Living in soup.
16:21 Dyrcona Hmm. Looking into something from yesterday that I thought would require a staff client change and now it looks like I can handle it all on the back end.
16:22 Dyrcona berick: Do you remember the LP bug that was created for the branch that allowed pre cat copies to be created for ILL and other purposes?
16:22 Dyrcona kmlussier and I tested it, but I can't find the bug now, even search fix committed bugs.
16:23 mmorgan1 joined #evergreen
16:23 mmorgan1 left #evergreen
16:24 mmorgan2 joined #evergreen
16:24 pinesol_green Launchpad bug 1308239 in Evergreen "Support targeting and fulfillment of precat copy holds (for ILL)" (affected: 2, heat: 10) [Wishlist,Fix released] https://launchpad.net/bugs/1308239
16:24 mmorgan2 left #evergreen
16:25 Dyrcona berick++ # that's a handy feature.
16:26 Dyrcona Looks like bshum tested it, too.
16:27 hopkinsju Dyrcona: Just made a breakthrough.
16:27 Dyrcona hopkinsju: What was the problem?
16:28 hopkinsju Initial issue was the default last_run_date and the fact that we didn't use the script to do the initial export
17:00 Dyrcona Well, time to go home.
17:12 mmorgan1 joined #evergreen
17:12 mmorgan1 left #evergreen
17:18 pinesol_green Incoming from qatests: Test Success - http://testing.evergreen-ils.org/~live/test.html <http://testing.evergreen-ils.org/~live/test.html>
18:20 rjackson_isl_ joined #evergreen
18:20 mmorgan joined #evergreen
18:23 TaraC joined #evergreen

Results for 2015-08-10

10:07 Dyrcona csharp: We don't seem to have set one.
10:08 Dyrcona We often get requests to fix patrons after registration, but it's usually when the staff tries to use a barcode that got assigned as the usrname on a previous user.
10:08 collum joined #evergreen
10:08 csharp Dyrcona: we discovered the need for it when testing the built-in selfcheck, which says "check whether the string the patron has entered matches the opac.barcode_regex setting... if not, treat the string like a username"
10:09 csharp since our setting is null, all entered values are being treated like usernames
10:09 Dyrcona Well, our opac.barcode_regex setting is a bit strange. It doesn't check length, for one thing.
10:09 Dyrcona I thought you were asking about a different setting.
10:10 Dyrcona But, the patron's barcode and usrname not matching, but both being barcodes, seem to happen in registration when the staff start out using 1 barcode and end up issuing a different one.
10:41 Christineb joined #evergreen
10:56 jeff interruptions, possibly.
11:11 * mmorgan reads backscroll
11:12 pinesol_green Incoming from qatests: Test Failure - http://testing.evergreen-ils.org/~live/test.html <http://testing.evergreen-ils.org/~live/test.html>
11:12 mmorgan Our opac.barcode_regex specified the barcode must have at least five digits, that's all. Not sure where it actually comes into play.
11:14 Dyrcona The test failure looks like one of those things we can't control.
11:14 Dyrcona Got a 503 while trying to download the phantomjs package.
11:17 phasefx start local repos for everything?
11:27 Dyrcona Nope. It'll probably succeed when the issue on the server is discovered and resolved.
15:26 jlitrell joined #evergreen
15:27 jeff ah. title hold on a bib whose only volume is deleted.
15:33 jeff and we have a surprising number of those.
15:35 pinesol_green [evergreen|Michael Peters] LP#1394989: Make users_of_interest test for defined actor.usr.card values - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=13da6b8>
15:35 pinesol_green [evergreen|Bill Ott] LP#1394989: Do not include deleted users when retrieving for Collections - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=0f8ec14>
15:39 berowskim joined #evergreen
15:57 Bmagic I have a bib that is sorted to the bottom of the list search when sorted newest to oldest. Come to find out, there is no entry in metabib.record_sorter for pubdate for this bib. When I reingest the bib, it still doesnt get created. Where is the glue?
16:02 Bmagic I found it. sorry, that was easy, Date1 in the 008
16:11 Dyrcona It handles authority deletes from a file matching DEL
16:12 Dyrcona Anything else that doesn't begin with R followed by 2 digits is treated as an authorities file.
16:13 Dyrcona We get zips with bibs, authorities, and reports all in one and it handles them OK.
16:13 hopkinsju So if I have a zip file with ONLY a file called BIB.MRC in it, shouldn't that get overlayed?
16:14 hopkinsju I'm trying to do some smaller scale testing before i set it loose.
16:15 Dyrcona It depends. If you look in Backstage::Import->do_bibs you'll see there are some criteria, but you should get some output to the screen as each record is processed.
16:15 Dyrcona Sorry, it's doBibs not do_bibs.
16:15 hopkinsju Ok
17:00 Dyrcona It's probably because of the full path to BIB.MRC in the zip file.
17:01 Dyrcona The zip files that I get from Backstage have no directory information in the member file names.
17:02 Dyrcona Anyway, signing off for now.
17:03 pinesol_green Incoming from qatests: Test Success - http://testing.evergreen-ils.org/~live/test.html <http://testing.evergreen-ils.org/~live/test.html>
17:05 mmorgan left #evergreen
17:14 jeff deleted bibs with non-deleted call numbers. deleted call numbers with non-deleted copies.
17:14 jeff always fun to try to determine which way to "fix" this.

Results for 2015-08-09

05:05 pinesol_green Incoming from qatests: Test Success - http://testing.evergreen-ils.org/~live/test.html <http://testing.evergreen-ils.org/~live/test.html>
09:45 graced joined #evergreen
11:47 sarabee joined #evergreen
18:24 sarabee joined #evergreen

Results for 2015-08-07

13:21 kmlussier Dyrcona: No.
13:22 kmlussier Bmagic: I added a comment to https://bugs.launchpad.net/ever​green/+bug/1331174/comments/15
13:22 pinesol_green Launchpad bug 1331174 in Evergreen "Long Overdue processing needs org unit settings separate from Lost Processing" (affected: 4, heat: 18) [Wishlist,Confirmed]
13:23 kmlussier Bmagic: I don't know what kind of time you have today, but if you were able to get that branch ready today, I could load it on a test server before I leave for vacation.
13:23 kmlussier mmorgan is interested in testing it.
13:25 * kmlussier is trying to get various billing bits in before she tackles billing documentation for 2.9.
13:25 kmlussier Which brings me to bug 1436797. Maybe that's one I can handle.
13:25 pinesol_green Launchpad bug 1436797 in Evergreen ""grocery" term confuses frontline staff and patrons" (affected: 11, heat: 48) [Undecided,New] https://launchpad.net/bugs/1436797
13:41 Stompro dyrcona++ Thanks for enlightening me about the STAFF_CLIENT_NAME variable, that is just what I needed.  I'll add that to the buidling the staff client wiki page.
13:42 Stompro Dyrcona++ (getting the caps right)
14:12 pinesol_green csharp: It is possible.
14:12 Dyrcona Stompro: I think it might use the stamp id as the product tag in windowssetup.nsi, but I never actually install the Windows staff client, so I don't know for sure.
14:14 Stompro I don't thinks so, my STAMP_ID is set to rel_2_8_2 and the PRODUCT_TAG is set to '2.8' which gets combined with "Evergreen Staff Client" to set the default install.
14:17 Stompro Dyrcona, I just tested it out, and setting a custom PRODUCT_TAG in windowssetup.nsi does set the default install directory.  So I think I've got all the pieces figured out.  Thanks for your help.
14:18 Stompro I love having all the various options on how this stuff can work.
14:18 Stompro iii--
14:21 Dyrcona Stompro: Cool! Glad you find my ramblings useful.
16:47 kmlussier Heading out for the week. Have a nice week everyone!
16:48 mmorgan Have a great vacation!
16:49 kmlussier left #evergreen
17:07 pinesol_green Incoming from qatests: Test Success - http://testing.evergreen-ils.org/~live/test.html <http://testing.evergreen-ils.org/~live/test.html>
17:08 mmorgan left #evergreen
17:17 mrpeters left #evergreen
18:09 rlefaive joined #evergreen

Result pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148