Evergreen ILS Website

IRC log for #evergreen, 2017-04-14

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

All times shown according to the server's local time.

Time Nick Message
00:46 dbwells_ joined #evergreen
00:50 remingtron joined #evergreen
00:50 dbwells_ joined #evergreen
04:30 pinesol_green News from qatests: Test Success <http://testing.evergreen-ils.org/~live>
06:40 rlefaive joined #evergreen
07:31 Callender joined #evergreen
08:13 kmlussier joined #evergreen
08:34 mmorgan joined #evergreen
08:38 _adb joined #evergreen
08:39 Dyrcona joined #evergreen
08:40 kmlussier gmcharlt++ Duck trivia
08:41 kmlussier Well, the rest of the blog post was good too, but the duck trivia is what earns the karma.
08:42 Dyrcona :)
08:49 kmlussier OK, before I dive into work today, I might try to figure out why JetPack has stopped publishing to the Google+ Page.
08:50 Dyrcona @blame 9 G+
08:51 pinesol_green Dyrcona: G+ is why we can never have nice things!
08:52 Dyrcona @blame 27 JetPack
08:52 pinesol_green Dyrcona: JetPack is probably integrated with systemd
08:53 Dyrcona :D
08:54 * Dyrcona waits for a create database to finish.
09:02 gmcharlt And for those strange people who are not subscribed to the RSS feed: https://evergreen-ils.org/everg​reen-3-0-development-update-1/
09:10 pinesol_green [evergreen|Bill Erickson] LP#1675899 Targeter --retarget-interval override fix - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=ea4d8b9>
09:11 Dyrcona gmcharlt++
09:14 kmlussier joined #evergreen
09:14 kmlussier RSS feed? I found it in Facebook.
09:15 kmlussier OK, after jumping through a bunch of hopes and constantly reminding Google which account to use, I'm hoping I fixed the Google+ issue. But I won't know for sure until somebody publishes a new blog post.
09:17 * kmlussier also discovered in the process that the 'About Me' section of her wordpress.com account is very much outdated.
09:18 kmlussier And includes a link to open-ils.org
09:24 Dyrcona I suppose that I could upgrade to zesty zapus while I wait.
09:25 Dyrcona nah.
09:25 kmlussier What will they do now that they're at z?
09:25 gmcharlt Aactive Aardvark
09:25 Dyrcona Yeah, they didn't start with A in the first place.
09:26 Dyrcona They started with Warthog, then went to E with Edgy Eft.
09:26 Dyrcona Or, was it D?
09:27 kmlussier Breezy Badger
09:27 Dyrcona Hoary Hedgehog -> Breezy Badger -> Dapper Drake.
09:28 kmlussier No C
09:28 Dyrcona After D, they went alphabetically.
09:30 * kmlussier writes email to dev list proposing Durable Duck for the Evergreen 3.0 release.
09:30 Dyrcona I didn't really start using Ubuntu until Lucid Lynx in 2010.
09:30 Dyrcona heh.
09:30 Dyrcona The Quacken!
09:32 kmlussier We haven't had a good version naming debate in a few years now. We're overdue.
09:33 jonadab The only Ubuntu releases I ever used were warty and hoary.  Because the Debian that was available at the time was ridiculously out of date.
09:33 Dyrcona :)
09:33 jonadab Then sarge came out and I haven't looked at Ubuntu since.
09:34 jonadab kmlussier: As long as each new version number A) is strictly larger numerically and also B) sorts strictly later asciibetically than all previous version numbers, I'm good.
09:34 * Dyrcona eyeballs Devuan.
09:34 jonadab Dyrcona: I'm using devuan and hoping it catches on and eventually more or less entirely replaces Debian.
09:35 jonadab Seems good so far but hasn't got a lot of momentum yet.
09:35 Dyrcona Koha seems to have switched to the Ubuntu YY.MM scheme.
09:35 Dyrcona yeah. I've been meaning to set it up on a vm and give it a spin.
09:35 jonadab Dyrcona: That meets both of my criteria, so it'd be ok with me.
09:36 jonadab Going to a Java-style version numbering scheme where 1.6 comes after 2 would NOT be ok with me.
09:36 Dyrcona jonadab: Y'know that Java 2, was actually just 1.2, right?
09:36 jonadab Dyrcona: So I've been told.
09:37 jonadab But it was still confusing.
09:37 Dyrcona Java 2 was just marketing hype.
09:37 Dyrcona They (Oracle) still call it JDK 9, JDK 8.
09:37 Dyrcona So they have split versioning going on.
09:39 Dyrcona And my created database finished a few minutes ago. Time to test my eg_staged_bib_overlay script.
09:40 Dyrcona OK. So I messed up an option, obviously. :)
09:42 Dyrcona All right. I omitted --dbhost, and this gives me an opportunity to make the script more generic.
09:43 krvmga joined #evergreen
09:43 Dyrcona It's a shell driver for eg_staged_bib_overlay from the EOLI migration-tools repo.
09:53 yboston joined #evergreen
09:57 jeff kmlussier: are you actively working on bug 1378829?
09:57 pinesol_green Launchpad bug 1378829 in Evergreen "Typo COPY_STATUS_LONGOVERDUE.override permission" [Low,Confirmed] https://launchpad.net/bugs/1378829 - Assigned to Kathy Lussier (klussier)
09:58 jeff kmlussier: i was thinking i'd get a branch in place to fix the typo and clean up after any installations that had added a new perm instead of editing the existing perm's code.
09:58 jeff kmlussier: but if you're going to do that, i don't need to. :-)
09:59 kmlussier jeff: No, I'm not actively working on that one. When I saw Galen's comment, I put it on my potential to-do list for today, but, to be honest, there are other things ahead of it.
09:59 kmlussier Trying to finish up some MassLNC business this morning before I dive into bugs.
09:59 jeff kmlussier: mind if i snag?
09:59 kmlussier jeff: Go for it!
10:00 kmlussier jeff++
10:03 kmlussier heh, I just noticed that the long overdue bug was assigned to me, hence the questions from jeff. :)
10:03 mmorgan1 joined #evergreen
10:03 jeff kmlussier: hah!
10:03 jeff kmlussier: yes, i thought it had been unassigned also, but i think you grabbed it during the conference. :-)
10:04 * kmlussier wonders what else she grabbed.
10:04 jeff now i'm tempted to fix it in a different way...
10:04 jeff hrm.
10:05 jeff we use LONGOVERDUE far more than LONG_OVERDUE
10:09 jeff ban, but LONG_OVERDUE has snuck into string IDs.
10:09 jeff s/ban/bah/
10:59 jeff we have two XUL interfaces for "replace barcode".
11:00 Dyrcona jeff: Sounds 'bout right.
11:00 Dyrcona And, one works better than the other, IIRC.
11:03 jeff one suffers from bug 1498141, the other suffers from a bug where you can edit the call number but it fails if you don't blur that field before continuing, and i suspect they may both suffer from bug 1362743
11:03 pinesol_green Launchpad bug 1498141 in Evergreen "Replace Barcode Cancel Button Produces Error Message" [Undecided,New] https://launchpad.net/bugs/1498141
11:03 pinesol_green Launchpad bug 1362743 in Evergreen "No error message when using duplicate barcodes with Replace Barcode in Holdings Maitenance interface" [Undecided,Confirmed] https://launchpad.net/bugs/1362743
11:06 kmlussier Are any of these problems fixed in webby?
11:06 Dyrcona Well, I imagine webby has a completely different interface for both of those, no?
11:06 kmlussier @eightball will webby solve all of our problems?
11:06 pinesol_green kmlussier: You're kidding, right?
11:07 Dyrcona Ha! eightball gets it right, again!
11:07 Dyrcona eightball++
11:07 kmlussier @karma
11:07 pinesol_green kmlussier: Highest karma: "gmcharlt" (11), "berick" (10), "dbs" (9), "bshum" (7), and "Bmagic" (5).  Lowest karma: "oracle" (-6), "ie" (-5), "comcast" (-3), "bzr" (-1), and "Launchpad" (-1).  You (kmlussier) are ranked 6 out of 23.
11:07 Dyrcona @bartender eightball
11:07 * pinesol_green fills a pint glass with Dos Equis Amber, and sends it sliding down the bar to eightball (http://beeradvocate.com/beer/profile/57/171/)
11:08 jeff webby avoids some of the issues by nature of it not re-using the unified volume/copy editor for a replace barcode interface, and does not suffer from the cancel button error referenced above.
11:08 berick ok.. evergreen 2.7 z39.50 encoding issues.. this should be a joy ride
11:09 jeff i haven't tried to create a duplicate barcode across copies/serials
11:09 kmlussier jeff: Yeah, I imagine that bug probably still exists.
11:10 jeff webby suffers from a few other issues, like a silent failure if the barcode to replace doesn't exist.
11:11 jeff also a silent error if the desired new barcode already exists
11:11 Dyrcona Those should be on Lp if they're not already there.
11:12 Dyrcona Hint....Hint... ;)
11:12 jeff just found. i'll try to get them there.
11:12 Dyrcona Cool.
11:13 jeff technically webby has two interfaces also, at least one that behaves differently depending on how it is accessed.
11:15 jeff berick: that sounds thrilling.
11:19 sandbergja joined #evergreen
11:20 khuckins_ joined #evergreen
11:21 jeff berick: local news site posted the list of upcoming artists that will be performing this summer at a local venue. there were at least a half dozen encoding errors in the list.
11:21 berick hah, i feel slightly better
11:24 berick Dearest Eliza, I don't know which is worse, the typhoid, the bad coffee, or the encoding errors *cue sad banjo*
11:24 jeff since then, they apparently edited the article body, but in several cases they ended up replacing something like ō with o.
11:24 berick you don't hear a lot of sad banjo, only in the civil war
11:25 jeff Macron: Not to be confused with Macaron, Macaroon, or Makron.
11:28 Dyrcona Or umlaut. :)
11:41 eady joined #evergreen
11:42 jeff Dyrcona: Not to be confused with omelette. See also, https://en.wikipedia.org/wiki/Metal_umlaut
11:42 berick lol @ rock dots
11:55 berick ah, changing outputcharset="marc-8" to outputcharset="utf-8" in xml2marc-yaz.cfg fixes the encoding issue.  now to see who's day that's going to mess up.
12:38 sard joined #evergreen
12:39 Bmagic marc--
12:39 Bmagic I just realized that it's possible someone would use that name for their nick.....
12:49 kmlussier heh
13:01 marc-- Sure, why not?
13:02 mmorgan joined #evergreen
13:29 tsbere Now I almost want to use that as my nick just to get the side effect on marc's karma when people try and get my attention. ;)
13:42 bshum Ha
13:46 berick bshum++ # ansible installer additions
13:46 berick huh, never occurred to me to use variables in the name: fields
13:48 kmlussier bshum++
13:48 bshum berick: I saw that while I was reading the documentation guide
13:49 bshum Seemed like it would benefit the user to know what we were changing where we changed it
13:49 jeff kmlussier++ gmcharlt++ bug approach feedback
13:49 berick bshum: indeed
13:49 gmcharlt jeff: oh, it's bug /approach/ feedback you wanted!
13:49 gmcharlt I suggest you approach bugs from behind, on tiptoes
13:49 jeff bugfix approach feedback?
13:49 jeff heh
13:49 gmcharlt otherwise they tend to fly away
13:50 jeff gmcharlt: i was just happy that when i returned to that window (having done most of the work already) that the new comment from you was "sounds sensible." instead of "this is a terrible idea, and here's why:"
13:51 gmcharlt heh
13:53 * kmlussier grumbles at her web host and wonders if she'll ever have time to review patches before taking off for a week.
13:54 kmlussier gmcharlt: A question about next week's point release. I know 2.10 is officially in security-fix-only mode, but we missed the March maintenance release, and there are a lot of bug fixes in the 2.10 branch that didn't make it into a packaged release.
13:55 kmlussier gmcharlt: Did you have any thoughts about doing one last 2.10 release before it really is security only?
13:55 gmcharlt kmlussier: indeed; FWIW, I've no objection to stretching it one more release
13:56 kmlussier I'm +1 to one more release too.
13:57 khuckins__ joined #evergreen
14:08 khuckins_ joined #evergreen
14:13 Jillianne joined #evergreen
14:18 jonadab tsbere: Heh.
14:28 khuckins__ joined #evergreen
15:03 khuckins_ joined #evergreen
15:31 kmlussier miker / phasefx / gmcharlt: Would you say everything in collab/gmcharlt/webstaff-sprint-4-5 are in backportable, bug fix territory?
15:33 kmlussier Actually, it appears as if the Lovefield commit is more new feature than bug fix. hmmm
15:33 jeff bug 1378829 now has a pullrequest tag
15:33 pinesol_green Launchpad bug 1378829 in Evergreen "Typo COPY_STATUS_LONGOVERDUE.override permission" [Low,Confirmed] https://launchpad.net/bugs/1378829
15:34 kmlussier jeff++
15:35 gmcharlt kmlussier: Lovefield is definitely for future as part of offline
15:36 gmcharlt as part of prepping for serials, we're thinking of merging sprint-4-5 into master in next day or two, but if you want to go ahead and pick stuff, that would be great
15:36 kmlussier Yeah, I was just thinking of picking the bug fix stuff.
15:37 kmlussier gmcharlt: After the merge, are you all planning to put bug fix commits in branches separate from new feature?
15:40 gmcharlt kmlussier: yeah, I think we're now in a mode where we will be mostly handling bugfixes normally via LP
15:40 kmlussier gmcharlt: Excellent! Thanks!
15:53 kmlussier My first time installing the web client since the recent changes went in. The only change I need to make when installing is that I no longer need to run the 'bower install' command. Is that right?
15:54 kmlussier I'm using the ubuntu-trusty-developer Makefile.install target, which should cover the changes to node.js
15:55 berick kmlussier: taht's right
15:56 Bmagic jeff++ # LOD
15:58 kmlussier OK, I'm going to try again. It wasn't too happy on my first try.
16:04 jeff Bmagic: lots of lines of upgrade script for a single character typo. there's probably some law about the longer a small bug sits unfixed, the more code and effort it takes to actually fix it. ;-)
16:05 Bmagic yep! I agree
16:05 Bmagic The unmerged branch I have for the LOD -> Lost will need revised
16:06 Bmagic Probably just remove the perm renaming bits and let your branch do it
16:06 jeff That, and I really had to resist the urge to just settle on one spelling of "long overdue" in constants and codes and variable names...
16:06 Bmagic yeah, I remember having that argument with myself as well
16:09 mmorgan jeff++
16:09 mmorgan me resists the urge to say that the fix was long overdue...
16:09 mmorgan oops.
16:10 berick *cue sad banjo*
16:11 jeff mmorgan: I said it everywhere -- it's all in how you interpret the words!
16:11 jeff * faba81f - LP#1378829 Fix long overdue override permission
16:11 jeff Open-ILS/src/sql/Pg/upgrade/XXXX​.data.fix_long_overdue_perm.sql
16:11 jeff etc :-)
16:12 mmorgan :)
16:13 jeff Bmagic: yeah, only minor changes required, and one would probably be handled automatically by git and the other would be a no-op, but I'm commenting on bug 1331174 right now.
16:13 pinesol_green Launchpad bug 1331174 in Evergreen "Long Overdue processing needs org unit settings separate from Lost Processing" [Wishlist,Confirmed] https://launchpad.net/bugs/1331174
16:30 pinesol_green News from qatests: Test Success <http://testing.evergreen-ils.org/~live>
16:37 khuckins__ joined #evergreen
16:39 kmlussier I tried installing the web client again. I'm getting an error when I run npm install. https://pastebin.com/qevBnNc0 is what I see in the logs.
16:52 jeff kmlussier: is that at the point where you've installed the LTS version of node and you're running "sudo npm install -g grunt-cli"?
16:55 kmlussier jeff: No. I run ubuntu-trusty-developer Makefile.install target as the root user in an earlier step, which should install the dependencies for me.
16:56 kmlussier This error occurred when I'm at 4.2 in the instructions and running npm install as the opensrf user.
16:58 bshum Hmmm
16:58 _adb ubuntu repos have a rather old  version of node and npm, i had to dl from nodejs.org and update npm to get angular to do anything.
16:59 _adb node 0.10.25 vs 6.10.2; npm 1.3.10 vs 4.5.0.
17:00 bshum kmlussier: Maybe you've got some bad node install files
17:00 bshum Is this a fresh install?
17:00 jeff looks like the ubuntu-trusty-developer target will end up installing nodejs-legacy (0.10.25~dfsg2-2ubuntu1)
17:00 kmlussier bshum: Yes, I just installed it this afternoon.
17:00 bshum kmlussier: You could try removing the node_modules directory
17:00 bshum And then try npm install again
17:01 jeff as of commit 33e365b we're doing "install a modern Node.js"
17:01 pinesol_green jeff: [evergreen|Dan Scott] Install a modern Node.js - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=33e365b>
17:01 kmlussier But do we need to update the ubuntu-trusty-developer target?
17:01 bshum Not exactly
17:02 bshum Wheezy is I guess
17:03 kmlussier bshum: I could try removing the node_modules directly as you suggested, but it would just fix it for now. The next time I rebuild a VM following the same steps, I expect I would encounter the same problem, right?
17:04 bshum kmlussier: I only suggest that in case there was a bad issue with node itself
17:04 bshum Sometimes they push something new in that makes stuff unhappy
17:04 mmorgan left #evergreen
17:04 bshum I installed everything earlier this afternoon and all was fine
17:05 jeff bshum: what version of node and npm do you have installed, and what distro are you using?
17:05 kmlussier OK, I'm going to have to investigate later. I need to run off for the evening.
17:05 bshum jeff: I was testing the Ubuntu 16.04 ansible stuff
17:05 bshum So it used the makefile developer target to install node, etc.
17:06 bshum I'm spinning up a new fresh VM to try that out again
17:06 bshum On my other computer
17:07 jeff bshum: looks like you would end up with nodejs-legacy 4.2.6~dfsg-1ubuntu4 in that case, which might be the difference between success and failure in your case vs kmlussier's case
17:08 jeff bshum: i.e., what might have worked for you earlier today on xenial might not work for kmlussier on trusty.
17:08 bshum jeff: Oh I see trusty
17:08 bshum I didn't notice that part, was it in the paste details?
17:08 bshum Ah okay no I just didn't scroll up far enough
17:08 bshum I haven't built on trusty lately
17:08 jeff 16:55:27 < kmlussier> jeff: No. I run ubuntu-trusty-developer Makefile.install target as the root user in an earlier step, which should install the dependencies for me.
17:09 bshum Or at least, not before the conference
17:09 bshum It could be also the new stuff we changed for bower->npm... maybe that's unhappy with something on trusty :\
17:09 bshum And I broke it when we merged that
17:09 bshum I'll spin up a trusty VM to check I guess
17:34 * bshum thumbs his fingers waiting for Evergreen to finish installing pre-reqs
17:36 bshum Okay, moment of truth coming up
17:39 bshum Yup, unhappy :(
17:43 bshum So it is related to the bower/npm switchup
17:43 bshum It's choking on "angular-order-object-by": "rxfork/ngOrderObjectBy#npm"
17:43 bshum Maybe our aged version doesn't know what to make of that line
17:44 bshum I'll try installing a newer node and seeing what it does differently with that.
17:48 * bshum hacks his trusty makefile to use the wheezy style for installing newer node
17:54 * bshum waits, and waits, and keeps waiting...
18:02 bshum Hmmmm
18:02 bshum Yeah
18:02 bshum Okay
18:03 bshum So using a newer version of node helps with the problem
18:03 bshum But I have discovered that the Makefile for Wheezy has not been updated to work with a newer version of node
18:04 bshum It's still installing from source for version 0.10.28
18:04 bshum Cause that's what it still says in the Makefile.install source file
18:04 bshum We fixed it in the eg_wheezy_installer for random repo, for the live test building
18:04 bshum But the actual installer isn't working quite right
18:05 bshum So we should rip all that stuff out
18:05 bshum But yeah, Trusty also won't work without a newer node version
18:05 * bshum goes to eat
18:06 * bshum brings his laptop to think with while he eats
18:42 bshum @later tell kmlussier Check out my work in progress branch:  http://git.evergreen-ils.org/?p=working/E​vergreen.git;a=shortlog;h=refs/heads/user​/bshum/newer-node-for-wheezy-and-trusty
18:42 pinesol_green bshum: The operation succeeded.
18:42 bshum I ran those Makefile changes through a test on Trusty and ended up with newer nodejs and able to run all the npm package installs for Trusty
18:42 bshum Same changes should apply towards Wheezy too
18:43 bshum If it works out, we need some tweaks to the README instructions so that the Wheezy specific steps can go away and we can just stick to recommending use of the -developer make target
18:44 bshum I think it ought to work for the debian-wheezy-developer target too.
18:44 bshum But it's untested
18:44 bshum I'll try more of it later on.
18:44 * bshum goes out to enjoy his Friday :)
18:47 bshum dbs++ # for new nodejs inspiration from the auto installer scripts
18:48 bshum kmlussier++ # for helping to find the problem
18:48 * bshum disappears
18:57 kmlussier joined #evergreen
18:58 kmlussier bshum++
21:05 jeff Looking at circ policy for a library that makes a circ-duration distinction between "new" items and "replacement" items... currently they use a circ modifier on the "new" items, and while it's so tempting to use "Item Age <" in a circ policy matrix, that won't do it...
21:08 jeff Using bib or volume age won't quite do it either. I think they're going to continue to use a circ modifier, and possibly use a report to know when to change the circ modifier to what it would "normally" be.
21:09 jeff But again, that seems less than idea.
21:09 jeff But again, that seems less than ideal.
21:21 jonadab Our criterion for when a "new release" three-day movie gets moved over to normal 7-day loan is, in a nutshell, when we notice that it's sitting on the shelf instead of having a hold list.
21:22 jonadab That's an oversimplification, but it's basically mostly that.
21:22 jeff see, that's almost something that we could automate.
21:22 jonadab Yeah, we don't have it automated, but it theoretically could be.
21:22 jonadab Well, except someone does have to physically remove the three-day-loan sticker.
21:23 jeff I think that in a lot of systems, "circ modifier" doubles as a kind of item type.
21:24 jeff So unless you have a "NEW" variant of *every* circ modifier, you can't just "make it no longer NEW", because nothing stores the "circ modifier as a kind of item type" that it should be.
21:26 jeff And yes, physical stickers are sometimes the forgotten challenge in this kind of thing. Excellent point. :-)
21:26 jeff We have some things that go in temporary locations until they are no longer "new". They end up with different colored dots. They don't circulate any differently, they're just housed somewhere different.
21:27 jeff I've mused on ways to do a checkin alert once they've been in that location longer than X... "please remove dot and now this goes to a different location" kind of thing.
21:27 jeff (again, very roughly simplified description)
21:31 khuckins__ joined #evergreen

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