Evergreen ILS Website

IRC log for #evergreen, 2017-09-12

| 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
03:50 HTTP_____GK1wmSU joined #evergreen
03:53 HTTP_____GK1wmSU left #evergreen
05:01 pinesol_green News from qatests: Test Success <http://testing.evergreen-ils.org/~live>
06:02 kmlussier joined #evergreen
07:17 rjackson_isl joined #evergreen
07:43 Dyrcona joined #evergreen
08:19 kmlussier Dyrcona++ #Fixing the MassLNC VMs!
08:19 collum joined #evergreen
08:19 kmlussier @praise Dyrcona
08:19 * pinesol_green Dyrcona can run a report without assistance
08:20 Dyrcona Aw, shucks....
08:54 bos20k joined #evergreen
09:07 mdriscoll joined #evergreen
09:12 yboston joined #evergreen
09:12 jvwoolf joined #evergreen
09:18 rlefaive joined #evergreen
09:31 jvwoolf1 joined #evergreen
09:35 rlefaive joined #evergreen
09:58 mmorgan joined #evergreen
09:59 mmorgan Dyrcona++
09:59 mmorgan mmmpudha
09:59 mmorgan oops.
10:07 kmlussier bank password?
10:08 jonadab Nah, everyone knows bank passwords have to be four-digit numbers.
10:08 * jonadab ducks.
10:08 kmlussier heh
10:10 mmorgan not a password anymore ;-)
10:17 dbs joined #evergreen
10:27 dbs gmcharlt++ # quick responses
10:32 collum_ joined #evergreen
10:47 rlefaive joined #evergreen
10:57 DPearl joined #evergreen
11:28 rlefaive_ joined #evergreen
11:43 _adb joined #evergreen
11:55 _bott_1 joined #evergreen
12:02 dwgreen joined #evergreen
12:10 jihpringle joined #evergreen
12:20 khuckins joined #evergreen
12:22 Christineb joined #evergreen
13:18 kmlussier joined #evergreen
13:20 LSachjen joined #evergreen
13:27 * Dyrcona has working OpenSRF on Debian 9 Stretch.
13:34 kmlussier gmcharlt: Does 3.0 require an authority reingest?
13:34 gmcharlt Dyrcona++
13:35 gmcharlt kmlussier: it probably does, although I hope to put something together that would make it as painless as possible
13:35 kmlussier gmcharlt: OK, I didn't see anything in the upgrade script, but it sounds like you're on it.
14:22 jvwoolf joined #evergreen
14:42 Dyrcona So, using Chromium 60 on Ubuntu 17.04 with OpenSRF and Evergreen master, text boxes do not appear to have focus when I click in them.
14:43 Dyrcona No blinking cursor. I saw this the other day on just the login page, but it seemed to fix itself after the first time.
14:43 Dyrcona And the blinking cursor is back....Must be just one of those things.
14:44 jeffdavis That happens to me from time to time too.
14:44 Dyrcona So, cool. I can login via srfsh and the web staff client and search returns results on a Debian 9 host. I'll add an extra step to the README and fix up my branch and signoff on bshum's commit.
14:44 jeffdavis I think the focus is there, just the cursor is missing.
14:45 Dyrcona jeffdavis: Yes, that's how it appears. If I type, the box works.
15:01 Jillianne joined #evergreen
15:03 mmorgan1 joined #evergreen
15:17 * berick has also had the missing cursor issue
15:25 Dyrcona One of the negative balance live tests fails for me on Concerto. It's test 90.
15:26 Dyrcona Also, I think we have two live tests with the same number again.
15:26 Dyrcona Yeah, two 24s.
15:30 bshum And two 20s
15:31 Dyrcona you're right. I missed the 2 20s.
15:32 Dyrcona It's not a big deal, but what's the point of numbering them.
15:32 bshum Isn't that some tests have to run sequentially?
15:33 Dyrcona Could be that.
15:33 Dyrcona I'm just saying if we duplicate numbers, then why number them? ;)
15:34 Dyrcona Anyway, I did login and do a search prior to running prove, so I've reloaded a fresh db and will try again.
15:34 Dyrcona I also ran make check, but that doesn't do the live tests.
15:34 bshum That's probably a good idea.  I think login could goof something up like if you register a workstation ahead of whatever they add during the test
15:36 Dyrcona It passed this time.
15:36 berick the live tests have chaotic consequences.  until we have a system that does a full data rollback (a la pgtap) with each test we have to guarantee an order.
15:37 Dyrcona I know that. I'm being rhetorical, sort of... :)
15:37 bshum So... branch needed to rename the tests in a proper order?  :D
15:37 Dyrcona Guess someone (me) should open a LP bug to renumber them.
15:38 berick Dyrcona: ah, i'm being too literal
15:38 Dyrcona Or, bshum can open it. That works for me. :)
15:38 * bshum lets Dyrcona have the honor
15:38 Dyrcona berick: S'OK.
15:38 * bshum is too busy reading up on iPhone specs anyways
15:39 Dyrcona :)
15:40 Dyrcona I should probably look at reports on my training server. Clark seems to be running more than I expected it would.
15:40 berick i don't think sharing numbers is really such a problem as long as they run in the desired order.
15:41 Dyrcona yeah, probably not, but someone will eventually ask... :)
15:41 berick true
15:41 Dyrcona Anyway, Evergreen now runs on Debian 9. bshum++
15:41 berick woot
15:42 Dyrcona I should make sure the lib naming changes don't break anywhere else, but it shouldn't.
15:42 bshum Yeah I'll test the new commit too for Debian 9 but also apply it on my next Ubuntu 16 test build just in case
15:42 bshum Dyrcona++
15:43 bshum Maybe those changes will help get it working for Fedora again, hehe :)
15:43 Dyrcona Shouldn't hurt.
15:43 Dyrcona It's always better not to fight the tools.
15:43 roycroft hello, folks
15:43 * miker deposits $0.02: not that we shouldn't make the numbers unique (and the order perfectly predictable for scripting purposes), but I see them like sysv init scripts ... the numbers are for ordering deps only, and and same number means "meh, don't care"
15:43 roycroft is this channel for evergreen developers, or are general questions ok here?
15:44 abneiman roycroft: all Evergreen questions welcome here!
15:44 roycroft awesome
15:45 roycroft so if you'll indulge me briefly, i'll describe my situation
15:45 roycroft first of all, i'm not a librarian, nor do i have any kind of formal training in library science
15:46 roycroft i'm helping out a non-profit put together a library, which consists of two different collections that may be merged at some point, but will remain separate for now
15:46 roycroft there's a total of about 10,000 items in the two collections
15:46 roycroft i've been "cataloging" and classifying the collection using readerware books on a mac right now, with an external postgresql database
15:46 roycroft the classifying is going fine
15:47 kmlussier joined #evergreen
15:47 roycroft readerware's cataloging capability is rather lacking, though, and we've found that as we add more books to the database it has become difficult to find the appropriate books by doing readerware queries
15:48 roycroft so i thought that a proper library system such as evergreen would be a good idea
15:48 roycroft again, i have no formal training, but i think i have a sense of what to do
15:49 roycroft i started installing evergreen on an existing web server, intending it to be a vhost on that server, but before i got to the server installation it became fairly clear to me that it wasn't designed to fit a vhost situation such as i have
15:49 roycroft is that correct?
15:49 roycroft i can spin off a dedicated vm just for evergreen
15:49 roycroft i'm running debian squeeze on my web servers currently, btw
15:50 roycroft so i guess my initial questions are:
15:50 roycroft 1. is running it on a dedicated vm the best way to go?
15:50 roycroft 2. should i be installing a 3.x beta for evaluation purposes? i see that 3.0 is about to be released, but i don't know what evergreen release cycles mean
15:51 roycroft i.e. whether a x.0 release is really stable and should be used in production, or whether that is wishful thinking on th part of the devlopers :)
15:51 Dyrcona squeeze is old and not supported by Debian anymore. Upgrade to Jessie at least.
15:51 roycroft er, sorry
15:51 roycroft not squeeze
15:51 roycroft whatever the latest is
15:51 roycroft i can't keep track of the names
15:51 Dyrcona Stretch?
15:52 roycroft the one that comes after jessie
15:52 roycroft yeah, that's it
15:52 Dyrcona You just missed it. We just added support for Jessie to experimental branches in the working repository.
15:52 roycroft 3. i have several thousand items in readerware already - is it worth trying to migrate those data to evergreen?
15:52 Dyrcona s/Jessie/Stretch/
15:52 * Dyrcona can't keep the names straight with his fingers.... :)
15:52 roycroft ok
15:53 roycroft this is a kind of background project
15:53 roycroft so there's no rush
15:53 Dyrcona Stability of the .0 releases varies, some are more stable than others.
15:53 roycroft if there's an experimental branch that runs on stretch that would be fine
15:53 roycroft ok
15:53 Dyrcona roycroft: See bug 1708048
15:53 pinesol_green Launchpad bug 1708048 in OpenSRF "Add support for Debian 9 Stretch" [Undecided,New] https://launchpad.net/bugs/1708048
15:54 roycroft but if stretch support is just experimental now, and just got added, that support will probably not be in the first 3.x release, i should think
15:54 gmcharlt eh, it will likely make it into 3.0.0... although such support will almost certainly be marked experimental
15:54 roycroft ok
15:55 gmcharlt I'd anticipate that it will have stablized by 3.0.1 or 3.0.2
15:55 roycroft it sounds like it's worth pursuing then
15:55 gmcharlt maintenance releases come out monthly
15:55 Dyrcona Yeah, I was about to post gmcharlt's comment to that effect.
15:55 roycroft so by year's end it will likely be marked stable
15:55 gmcharlt yeah, barring a big surprise
15:55 roycroft sure
15:55 Dyrcona Well, we consider it stable at .0, really.
15:55 Dyrcona Officially, anyway. :)
15:56 miker roycroft: re types of installations, you /can/ run evergreen in a vhost-ish way, but the install instructions don't make that plain and it's not the common configuration. if you can set up a separate VM for evergreen, you'll be able to get a lot more help from the community in general
15:56 * roycroft has been doing this stuff for a while, and knows about surprises and expectations
15:56 roycroft thanks, miker - i'll just do that then
15:56 roycroft that was becoming apparent to me when i decided to join this channel and ask :)
15:56 Dyrcona miker++ for beating me to that comment, too.
15:56 gmcharlt yeah, we don't typically brown-bag x.0.0 releases, but we also don't deviate from the typical pattern that *.0.1 has useful bugfixes :)
15:57 roycroft i'm usually not an early adopter, and very rarely do i use bleeding edge technology
15:58 miker as for migration of existing data, if you can script something to spit out MARC data with copy (barcode) information embedded in a particular way, and more quickly than re-cataloging, it's doable. depends on the desired level of cataloging data, really
15:58 roycroft but since i'm just getting ready to start evaluating this package, and am in no rush, it makes sense to me to use the 3.x train, especially since there are major differences between that and 2.x
15:58 roycroft sadly, readerware is a consumer-grade product, and does not import marc data
15:59 roycroft it populates the database by scouring a list of websites
15:59 roycroft a typical way to add a record is to find the item on amazon and drag and drop the url into readerware
15:59 roycroft this is why i say cataloging is weak
15:59 miker indeed :)
15:59 roycroft it can go to loc, and i get call numbers from there
15:59 roycroft so classification is ok
15:59 roycroft but that's far from enough
15:59 roycroft even with a somewhat small collection
16:00 roycroft and since there are two collections in two phyical locations right now, browsing the shelves isn't particularly useful either
16:00 gmcharlt there are a variety of CSV to MARC convterers out there, but to be honest, best way to get clean MARC records is to grab them via Z39.50
16:00 roycroft i don't mind starting that work over
16:00 gmcharlt which can be eased by making sure that you keep track of ISBNs and LCCNs
16:01 khuckins_ joined #evergreen
16:01 roycroft yes, i figure i can at least export a list of things to recatalog
16:01 gmcharlt would I be correct in guessing that most of the collections consist of commercially published books?
16:01 roycroft hopefully i'm not wearing out my welcome already
16:02 roycroft this has been all good information so far - these few minutes have saved me a lot of time
16:02 roycroft the majority of the collection, yes
16:02 roycroft however
16:02 roycroft there are a lot of old (pre-isbn) books
16:02 roycroft and a lot of foreign (no lccn) books
16:02 roycroft so there will be some manual cataloging to do no matter what
16:02 gmcharlt OK
16:03 gmcharlt pre-ISBN books published in the US are pretty likely to have records in the Library of Congress' Z39.50 database
16:03 roycroft is there a place to get marc records for free, without having to get a loc subscription?
16:03 gmcharlt ah, Z39.50 access  to LC is free
16:03 roycroft yes, pre-isbn us published books are pretty well represented at the loc
16:03 roycroft oh, cool
16:03 * roycroft repeats that he's doing this for a non-profit, and doesn't mean to come off as being cheap
16:04 gmcharlt and they recently released a dump of their entire catalog in MARC format as of 2014
16:04 gmcharlt (although that's almost certainly waaaaaay more data than you'd want to chow through)
16:04 Dyrcona :)
16:05 roycroft so it sounds like i should build a vm, install the latest beta, do a csv export from readerware with probably title, author, lccn, isbn, and start importing records from loc
16:05 roycroft does evergreen have a way of doing those imports?
16:05 gmcharlt yeah, that's a reasonable plan
16:05 roycroft i.e. can i uploade the csv to evergreen and it will go suck down the details?
16:06 Dyrcona roycroft: You'll need to convert the CSV to MARC.
16:06 roycroft there's a library system that covers most of eastern oregon (i'm in western oregon) who use evergreen, and folks seem to think it's a really nice package, which is what got me looking at it in the first place
16:06 roycroft so csv to incomplete marc, and then evergreen can get the full marc records?
16:06 Dyrcona Also, if you want to test on stretch use the branches on the bug I referenced earlier.
16:07 roycroft ok, i'll do that
16:07 gmcharlt yeah, Evergreen doesn't have a direct script to batch-import records from a Z39.50 source, but here's another tool that can do it: http://marcedit.reeset.net/batch-m​arc-record-retrieval-using-z39-50
16:07 Dyrcona They're basically the beta for 3.0.
16:07 gmcharlt also, for the foreign records, http://irspy.indexdata.com/ is a directory of Z39.50 servers across the world
16:08 roycroft brilliant
16:08 roycroft thank you~!
16:08 gmcharlt (and to expand on what I mean with the MARCEdit example: Evergreen /can/ batch import files of MARC records)
16:09 roycroft this seems pretty doable
16:09 roycroft so one other thing for now
16:10 roycroft i'm currently printing spine labels on some avery label sheets
16:10 roycroft i export a csv of 60 call numbers from readerware, run through an awk script to format them for printing, and then load them into an avery app that prints the labels
16:11 roycroft is there some similar mechanism from within evergreen to do that?
16:11 roycroft i can the mark those records as labeled in readerware, so i'm not repeating myself
16:12 roycroft i ask this because if evergreen can't do it (or can't do it easily), it might be a good idea to keep building the collection in readerware until everything is classified/labeled
16:12 roycroft and then move to evergreen
16:12 gmcharlt yeah - you can do batches of labels directly in Evergreen, and if you prefer to stick with your current process, it's easy to grab reports of call numbers to feed into your current workflow
16:12 roycroft but if evergreen can do it we might want to stop adding to readeware and start adding directly to evergreen
16:12 roycroft ok
16:13 roycroft do actual libraries use dedicated spine label printers?
16:13 Dyrcona Yes, they do.
16:13 roycroft we tried one of those dymo things, but it's thermal, and the labels fade quickly
16:13 Dyrcona Well, some do, some don't.
16:13 roycroft which is why we went to laser labels
16:13 Dyrcona Some use laser labels.
16:13 roycroft but i don't want to have to print 60 labels every time i add a book
16:14 roycroft discussion of spine label printers is probably outside the scope of this channel, so i'll leave it at that for now
16:14 roycroft and do some research on my own
16:14 Dyrcona If you search the general mailing list archives you can find some information about spine label printers that work well with Evergreen.
16:14 roycroft i'll do that
16:14 roycroft thank you again
16:14 Dyrcona Don't worry about being off topic here, no one will ban you for it.
16:15 Dyrcona And spine label printers is closer than some of the things I get up to. :)
16:19 Dyrcona Well, since that Debian 9 thing took less time than I thought, I may look at some more bugs this week.
16:20 Dyrcona I've got a list of "showstoppers" from C/W MARS.
16:20 roycroft i try to be respectful
16:20 roycroft i've been on irc for over 25 years, and i've seen a lot of stuff :)
16:20 Dyrcona roycroft: Same here, though Usenet was generally worse.
16:21 roycroft usenet used to be useful
16:21 roycroft back when i got my feed every night via uucp
16:21 roycroft and back when a person could read every post in every newsgroup every day and still get work done
16:21 roycroft but those days are so long gone they're almost totally forgotten
16:22 Dyrcona yeap. You can't get Usenet through most ISPs any more, and I've not looked in some years.
16:22 roycroft now usenet is nothing but spam and porn
16:22 roycroft as is the internet, generally :)
16:23 roycroft damn that al gore for enabling the general public to have access to the internet!
16:23 roycroft it was so much better when it was a private club
16:23 Dyrcona pfft. Maybe, maybe not.
16:23 * roycroft is being sarcastic
16:25 * Dyrcona is being too literal, eh, berick. :)
16:25 roycroft at any rate, you folks have given me some good answers, and i think i have a solid plan for moving forward
16:25 roycroft so thanks again
16:25 Dyrcona You're welcome and good luck!
16:25 roycroft i'll probably have another round of questions after i'm more familiar with the software
16:28 kmlussier joined #evergreen
16:28 Dyrcona Hm... I guess it is a bad idea to update reporter.schedule while Clark is running....
16:28 * Dyrcona counts the zombie processes.
16:29 Dyrcona 62!
16:30 mmorgan joined #evergreen
16:30 Dyrcona It's OK to just delete from reporter.schedule, yeah?
16:31 * Dyrcona tries it anyway... It's only the training server.
16:32 Dyrcona Mr. Kent seems much happier, now.
16:36 miker Dyrcona: confirmed, yes it's fine
16:37 Dyrcona miker: Thanks. That's what I thought and what it looks like from here. ;)
16:38 * Dyrcona actually makes a note to that when updating the data for the training and testing servers.....Typing it in will help me remember.
16:39 * Dyrcona imagines 62 zombie Supermen running around Gotham or whatever city Superman supposedly lives in.
16:41 gmcharlt clark-kent.pl ought to change its process name to 'Bizarro' just before terminating
16:55 Dyrcona I noticed that Clark still doesn't daemonize properly. It writes messages to the console, still.
16:56 * Dyrcona search for a LP bug on that.
17:02 pinesol_green News from qatests: Test Success <http://testing.evergreen-ils.org/~live>
17:03 kmlussier abneiman: The silent failure bug you just filed, is that the import marc permission or the create or update marc permission?
17:04 * kmlussier is wondering if it's the same as bug 1693580 or something different.
17:04 pinesol_green Launchpad bug 1693580 in Evergreen "web client: attempting to update a MARC record without required permissions fails without feedback to the user" [Undecided,New] https://launchpad.net/bugs/1693580
17:04 abneiman kmlussier: I thought it would be update or create, but the XUL error specifically calls out import
17:05 kmlussier huh.
17:05 abneiman lemme double check the console on webby
17:05 mmorgan left #evergreen
17:07 abneiman kmlussier: "open-ils.pcrud.create.bre failed" in console of webby
17:08 kmlussier abneiman: You're probably right. It just surprised me. I still don't really understand Evergreen. ;)
17:08 miker ha!
17:08 miker kmlussier: ;)
17:09 abneiman it suprised me too, but I am notoriously shaky on permissions
19:48 _adb1 joined #evergreen
19:48 pastebot0 joined #evergreen
19:54 berick joined #evergreen
20:37 bshum berick: I added a couple tweaks for the ansible installer here:  http://git.evergreen-ils.org/?p=working/​random.git;a=shortlog;h=refs/heads/colla​b/bshum/ansible-installer-ubuntu-16.04
20:37 bshum berick: It changes the ejabberd config back to default max_stanza_size and adds the rpaf apache module install for nginx
20:38 bshum I used it to test a clean installation using the Debian9 changes from Dyrcona for libaries.  Looks like everything works with the changes.
21:12 jvwoolf joined #evergreen
21:22 jvwoolf joined #evergreen
21:31 jvwoolf left #evergreen
22:58 pinesol_green [evergreen|Jane Sandberg] Docs reorg: Adding a new manual for system administration from the Web client - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=bfa1bfa>
23:13 pinesol_green [evergreen|Jane Sandberg] Docs reorg: Correcting contents of staff client system administration - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=9fa87e3>
23:35 pinesol_green [evergreen|Jillianne Presley] Docs reorg: Adding a cataloging manual - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=9f1ec3a>

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