Search in #evergreen

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

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.

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

00:58	phasefx	@later tell Dyrcona the live tests happen on a near pristine snapshot of wheezy and runs all the tests
00:58	pinesol_green	phasefx: The operation succeeded.
02:53		sarabee joined #evergreen
03:27		gsams joined #evergreen
09:05	kmlussier	And I see rain in the forecast for Tuesday and Wednesday. Figures.
09:13	Stompro	mmorgan, have you started on the release notes for bug 1124498, I also started on that but forgot to assign it.
09:14	pinesol_green	Launchpad bug 1124498 in Evergreen "Wishlist: Patron notification via email when card is about to expire" (affected: 6, heat: 40) [Wishlist,Confirmed] https://launchpad.net/bugs/1124498 - Assigned to Michele Morgan (mmorgan)
09:15	mmorgan	Stompro: I was actually working on final testing to sign off. You are welcome to the release notes :)
09:18	Stompro	mmorgan, Wonderful, I hope your testing goes well.
09:18	kmlussier	mmorgan++ Stompro++
09:19	kmlussier	Stompro: I think MVLC might do some of the things you described in your email with staff clients for test servers. tsbere might be able to answer your questions. I don't think he's around this week.
09:21	* mmorgan	is also interested in that info. I often find a need to run 2 clients simultaneously on the same workstation.
09:21	kmlussier	I use MVLC's servers quite a bit for testing. Icons for the clients that access the test servers are a different color and they don't share settings with the other clients on my local machine.
09:21	* kmlussier	has many, many clients on her local machine.
09:22	Stompro	kmlussier, great, any guidance would be appreciated. I'll send him an email if I don't hear from him.
09:22	kmlussier	mmorgan: Yes, that's what I like about testing on MVLC's servers. I can have two or more clients at the same time. tsbere also configured the clients for the MassLNC VMs to work the same way.
09:23	kmlussier	tsbere++ #for general wizardry
09:24	mmorgan	Yes, I noticed that. I can run a MassLNC client concurrently with a NOBLE client, but not a NOBLE Production along with a NOBLE training client.
09:25	mmorgan	Stompro: We will just change the hostname when the different servers are running the same version of Evergreen, but often our servers are on different versions.
09:27	Stompro	mmorgan, I think you can run multiple instances at the same time by enabling multi profile mode and multi-instance mode - http://wiki.evergreen-ils.org/doku.php?id=mozilla-devel:building_the_staff_client#multiple_workstations_on_one_install
09:28	Stompro	Add -P and -no-remote to your client shortcut.
09:30	mmorgan	Stompro: we use -profilemanager all the time. I'll have a look at -no-remote, though.
09:34	Stompro	I just wanted to make it super easy for staff to connect to the training/testing server, without needing to change the hostname, to take out a step where mistakes can be made. I'm just worried about staff accidently thinking they are are production when they are on testing and vice versa. Maybe adding a prefix to all usernames on the test system would help.
09:35	mmorgan	Stompro++ You made my day!
09:42	Stompro	Did the -no-remote work?
09:44	mmorgan	Yes! I now have 2 clients open at the same time :) Not sure how I missed that detail, but am happy to know it now!
10:41	Bmagic	Ok, that is whay I thought, just changing the parent OU should do it with autogen. Sounds easy enough
10:42	Bmagic	/whay/what
10:42	Dyrcona	You might have some memcached issues, so I'd wait and do it in the middle of the night when you can restart memcached without wrecking too much.
10:45	Dyrcona	I'd also suggest trying it on a test/training system first, just to make sure.
10:50	* kmlussier	is spending more time re-learning AsciiDoc conversion than writing documentation.
10:50	csharp	Bmagic: yes, we've moved a branch from one parent to another - it's not a big deal - just move it and run autogen (though you might schedule it after business hours)
10:51	csharp	all holdings and patrons, etc. remain intact
11:03	Dyrcona	My guess is that might be a mistake.
11:04	kmlussier	Yes, I don't think that was the intent at all.
11:04	Dyrcona	But, I'd need to play with with some samples and don't have time at the moment.
11:04	Bmagic	kmlussier: yep! That's what it looks like. I created a new 700 field on a test record and only created a subfield "a" and violla it was indexted
11:05	Bmagic	I am assuming that the contents of config.xml_transform are lifted from some global standard?
11:11	Dyrcona	Bmagic: http://www.loc.gov/standards/mods/
11:11	Bmagic	Dyrcona: right, and the xsl structure was downloaded somewhere or did the EG community create it from their guideline?
13:21	kmlussier	Dyrcona: No.
13:22	kmlussier	Bmagic: I added a comment to https://bugs.launchpad.net/evergreen/+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

Results for 2015-08-06

02:12		jonadab_znc joined #evergreen
05:09	pinesol_green	Incoming from qatests: Test Success - http://testing.evergreen-ils.org/~live/test.html <http://testing.evergreen-ils.org/~live/test.html>
06:40		rlefaive joined #evergreen
07:03		rlefaive joined #evergreen
07:25		rlefaive joined #evergreen
09:41	Dyrcona	Dunno, bshum. You tell me. :)
09:42	bshum	Dyrcona: Oh I was just musing on dbs' musings
09:42	Dyrcona	Yep.
09:43	* Dyrcona	goes back to writing a pgtap test for jeffdavis_afk
09:43	Dyrcona	Well, one of his branches, anyway.
09:43	bshum	:P
09:44	Dyrcona	But, bshum's latest comment on the bug he mentioned might prove helpful to dbs if that is the problem.
10:10	bshum	But not to master
10:10	bshum	And since translations in LP track against master
10:10	bshum	Those strings never showed up for needing translation in LP.
10:10	Dyrcona	miker: If you want sprint2 in master before the alpha/beta release, then you'll need to clean up the upgrade scripts and provide pgtap tests.
10:11	Dyrcona	miker: I'm also concerned about the FIXME comment in XXXX.schema.marc-tag-tables.sql
10:11	Dyrcona	I get errors on line 9 and will try reordering the drop statements, but also wonder if I should even be running those drops given the comment.
10:12	phasefx	bshum: fun
10:13	bshum	phasefx: I would imagine we could still cherry-pick berick's changeset from 2.8 to master now. And then when Dyrcona does i18n dances for 2.9, we just make sure to bring in all the translations
10:13	bshum	But I'm not sure how we can fix 2.8. Yet.
10:13	bshum	Still rethinking it over.
10:13	Dyrcona	I should do a test release build whether there is an alpha or not.
10:14	bshum	Dyrcona: It's always good practice.
10:15	Dyrcona	Yep. Trouble is: not much time for practice.
10:16	phasefx	bshum++ I'm glad you're thinking about it; it's over my head at the moment. So we do something to extract strings from source files into language files, commit those somewhere, and Launchpad can then change them or let us export a new updated set of files based on its global translation effort?
10:27	bshum	That should go on a t-shirt
10:28	* csharp	copyrights it before bshum gets a chance
10:30	Dyrcona	Hmm... Wonder if IRC chat is copyrightable.... That would be an interesting case.
10:31	Dyrcona	And, I still haven't written the pgtap test, yet.
10:32	dbs	Your chat in IRC is an expression of an idea, thus very likely copyrightable.
10:32	* dbs	was just answering a question around copyright and fair dealing for our online learning folks
10:32	RoganH	Yep. If you create the speech you can copyright it. You're just exercising your own right to perform it in public.
10:38	Dyrcona	(C) is not valid, so in Brazil, what you just said is in the public domain, csharp. :)
10:39	bshum	I'm comparing our outputs now, for giggles
10:39	Dyrcona	In Brazil it must have the c in the circle symbol, the word copyright is not good enough by itself. (That I remember from research from a number of years ago.)
10:40	Dyrcona	But, anyway.... that pgtap test...
10:40	* bshum	makes an adjustment to the query
10:40	bshum	Since he realizes current_shelf_lib = 22 doesn't mean the same thing in his system
10:41	Dyrcona	bshum: Try my database and use current_shelf_lib = 17, pretty much guaranteed to blow up at our busiest library. :)
10:42	bshum	I still got a return about 2533 ms
10:42	bshum	So 10 times faster
10:42	bshum	Chose our largest library
10:42	csharp	bshum: my current experiment is deleting (aka, "aging") most of the hold_request table on a test server - gonna see if that helps.
10:43	bshum	One of our busiest ILL libs, got 2977 ms
10:43	bshum	So yeah
10:43	bshum	The shape of the query is different
10:44	Dyrcona	Speaking of slow queries....
10:45	Dyrcona	bshum and berick do you think that lp 1479953 should be backported to 2.8 and 2.7 as a bug fix?
10:45	pinesol_green	Launchpad bug 1479953 in Evergreen "Deleting queues containing many records is slow, can time out" (affected: 2, heat: 10) [Undecided,New] https://launchpad.net/bugs/1479953
10:46	Dyrcona	That's what I'm trying to write the pgtap test for.
10:46	bshum	Dyrcona: Remind me what happens if you try to create an index where one already exists?
10:46	Dyrcona	The create index statement dies.
10:47	Dyrcona	But it might succeed if the names are different.
11:09	berick	bshum: arg, feel bad I caused a problem I still don't understand
11:15	bshum	berick: It's okay, I'm not sure I will ever fully understand it.
11:21	bshum	berick: When I can think through everything a bit more, I'll try making some changes to that release_process wiki page on how the i18n bits happen
11:22	Dyrcona	Hmm. Should a commit adding a pgtap test and simply renaming indexes from the previous commit get another sign off?
11:22		kbutler joined #evergreen
11:22	berick	bshum: k. let me know if I can help
11:24	bshum	berick: Looking back in the git history, I can see that changes for "po files" and "newpot" occurred as separate commits.
11:39	bshum	So yeah, every time we update the pot's in master, it'll setup new strings needing translation in LP
11:40	bshum	And then as people translate them in LP, it gets pushed to dbs' branch
11:40	bshum	Which we then pull back in when we run the updates_pofiles script
11:46	pinesol_green	[evergreen\|Jeff Davis] LP#1479953: Add indexes to vqbr foreign key references - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=679578d>
11:46	pinesol_green	[evergreen\|Jason Stephenson] LP#1479953: Rename indexes to *_idx and add pgTAP test. - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=57d756f>
11:46	pinesol_green	[evergreen\|Jason Stephenson] LP#1479953: Stamping Upgrade Script - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=768680b>
11:47	bshum	jeffdavis++ Dyrcona++
11:47	Dyrcona	kmlussier++ # she tested it, too.
11:48	bshum	berick: I'm going to forward port your 2.8 translation changes to master so that the Czechs have something to start working with.
11:48	berick	bshum++
11:51		bmills joined #evergreen
14:24	yboston	OK
14:25	yboston	#action kmlussier will complete the marc stream importer work and check on the status of the RDA docs
14:25	yboston	#info 3) remingtron will send out an email to the Dig list anouncing that a new file extension will be used for AsciiDoc files
14:25	remingtron	I was made aware of a possible problem with the way git tracks history when a file name changes. So I want to test that before any more public announcements.
14:25	yboston	interesting
14:26	yboston	I understood that it usually catches name changes
14:26	remingtron	don't want to mess up git history too much, especially for a non-essential change
14:30	Stompro	I'll do it in a few minutes, no need for an action item...
14:30	yboston	anything else on this issue/topic before me move on?
14:31	remingtron	Stompro: you're that much more ahead on next meeting :)
14:31	yboston	moving on
14:31	yboston	#info 4) remingtron will begin testing use of and conversion to the .adoc extension
14:32	yboston	is there more to say about this?
14:32	remingtron	not yet, I'll keep working on it
14:32	yboston	no problem at all
14:32	yboston	remingtron++
15:28	csharp	kmlussier: yboston learned that at Berklee College of Music :-)
15:28	yboston	1.25
15:28	kmlussier	csharp++
15:29	csharp	okay, so for the logs, I did a giant purge of old hold requests and that seems to have fixed the Clear Holds Shelf checkin modifier on our test server
15:30	csharp	now to figure out how to do the same on production without scheduling a several-hours-long downtime
15:33	Dyrcona	csharp: What did you do to purge the old holds?
15:33	dbs	yboston++
15:35		rlefaive joined #evergreen
16:10		Dyrcona joined #evergreen
16:34	berick	haha, george_duimovich++
16:42		bmills joined #evergreen
16:52	pinesol_green	Incoming from qatests: Test Success - http://testing.evergreen-ils.org/~live/test.html <http://testing.evergreen-ils.org/~live/test.html>
16:57	Dyrcona	Does the live test do a fresh install and run tests?
16:58	Dyrcona	Does it run pgtap as well as perl tests?
17:06		mmorgan left #evergreen
18:07		bmills joined #evergreen
18:10		jonadab_znc joined #evergreen

Results for 2015-08-05

16:19	berick	that is all
16:19	dbwells	berick++ # glad to see this moving forward
16:20	kmlussier	berick++
16:21	dbwells	berick: my C chops are basically non-existant, so my eyeballs won't help, but I'll plan to do some basic explosion testing.
16:21	berick	dbwells: well, the DB changes are my main concern
16:21	dbwells	ldw: action me up, good sir!
16:22	dbwells	berick: ah, ok