Evergreen ILS Website

IRC log for #evergreen, 2018-09-12

| 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
00:26 rlefaive_ joined #evergreen
00:27 rlefaive__ joined #evergreen
01:29 rlefaive_ joined #evergreen
06:11 agoben joined #evergreen
07:06 rjackson_isl joined #evergreen
07:21 Dyrcona joined #evergreen
07:28 bdljohn joined #evergreen
08:47 pinesol [evergreen|Cesar Velez] LP#1739087 - add [x] to volcopy editor volume rows - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=ad1d99c>
08:47 pinesol [evergreen|Cesar Velez] LP#1739087 - add [x] to volcopy editor volume rows-FollowUp - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=7417b39>
08:47 pinesol [evergreen|Cesar Velez] LP#1739087 - enable vol to be zeroed - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=ef3fa14>
09:05 lsach joined #evergreen
09:36 Dyrcona joined #evergreen
09:37 yboston joined #evergreen
09:54 Dyrcona berick: Will you be around or are you running from the hurricane?
09:58 aabbee joined #evergreen
09:59 berick Dyrcona: i'll be around
10:00 Dyrcona berick: OK. Thought I had a question about the OpenSRF README for websocketd, but think I have an answer.
10:05 * berick nods
10:08 stephengwills joined #evergreen
10:18 bshum bug 1735816
10:18 pinesol Launchpad bug 1735816 in Evergreen "Cannot Delete Existing Copy Notes" [Medium,Confirmed] https://launchpad.net/bugs/1735816
10:21 JBoyer It just needs looooovve!
10:23 Dyrcona Yeah, we're going to look at it today, IIANM.
10:26 JBoyer Dyrcona++
10:26 aabbee what's the status of evergreen3.2 on ubuntu18.04? https://evergreen-ils.org/docume​ntation/install/README_3_2.html only lists 16.04 and 14.04. is this likely to work, or should i stick with xenial?
10:27 bshum aabbee: Stick with xenial
10:27 bshum 18.04 is not supported yet
10:27 bshum OpenSRF is not compatible with it
10:27 bshum And we're still working on that before we get to the Evergreen parts
10:27 aabbee bshum++ # ok, thank you!
10:29 JBoyer berick, I do have a question about customizing eg2 templates. I know there was talk of a way to load custom overrides (a templates_local dir), did that make it in or is the generally accepted way to customize the new templates to make changes in the Evergreen source pre-build?
10:30 berick JBoyer: for now, changing the source pre-build is the only option
10:30 JBoyer berick++
10:30 JBoyer I was kind of hoping that was the case, otherwise I would begin to doubt my reading ability. ;)
10:33 Dyrcona Fun with OpenSRF 3.0 and Evergren 3.0: gateway 2018-09-12 10:30:12 [WARN:9710:transport_session.c:869:] Text of error message received from Jabber: XML stanza is too big
10:44 bshum Dyrcona: Possibly -->  https://bugs.launchpad.net/evergreen/+bug/1710293
10:44 pinesol Launchpad bug 1710293 in Evergreen 3.1 "Remaining chunk/bundle work" [Medium,New]
10:44 Dyrcona Yes.
10:46 jvwoolf joined #evergreen
10:47 khuckins_ joined #evergreen
10:53 Christineb joined #evergreen
10:56 mmorgan joined #evergreen
10:57 stephengwills Two questions: Apropos JBoyer’s question: Are there any best practice workflow docs for pulling a pre-build branch, applying org specific mods and maintaining it until it’s time to merge a new release onto it?  The dev docs i’ve found speak to contributing to a working branch but that is not the place one maintains their custom production code correct?
10:57 JBoyer stephengwills, No, you wouldn't want to store anything local in a working branch on git.evergreen-ils.org.
10:58 berick stephengwills: it varies, we keep local changes in a private repo and regularly rebase it over the stock code.
10:58 JBoyer What we do here is run a local git server that pulls from git.evergreen-ils.org every morning, then we can pull from it and push branches that only exist locally and on our local git server.
10:59 JBoyer Or in fewer words, what berick said.
10:59 miker stephengwills: not documented, but a common way is to use a repo (or, just a convention-based namespace in a repo) to hold the main branch you're tracking, a set of topic branches, and then create a combined branch as needed by rebasing the main branch, forking it, and cherry picking the topic branch commits onto the new combined branch
11:02 stephengwills k, tx.  I’m trying to get those workflows documented for Balsam and figured it was something like that.
11:02 JBoyer So long as you don't put anything sensitive in there you could use a github repo to do this rather than managing it yourself.
11:02 JBoyer (less sure how to pull from elsewhere in that case though)
11:03 Dyrcona We also use a private repo for customization and configuration.
11:04 Dyrcona berick: I updated https://bugs.launchpad.net/opensrf/+bug/1777180 You'll probably see the email, but I thought I'd let you know.
11:04 pinesol Launchpad bug 1777180 in OpenSRF "Wishlist: Websocketd Gateway Support" [Wishlist,Confirmed]
11:05 berick thanks Dy
11:05 terran joined #evergreen
11:05 berick hm, tab-complete, you're my only hope
11:05 berick thanks Dyrcona
11:05 csharp Dyrcona: JBoyer: we're about to dig into looking at bug 1735816 too, FYI
11:05 pinesol Launchpad bug 1735816 in Evergreen "Cannot Delete Existing Copy Notes" [Medium,Confirmed] https://launchpad.net/bugs/1735816
11:06 JBoyer Now there's an autocompletion I haven't heard in a long time...
11:07 berick :)
11:16 jvwoolf1 joined #evergreen
11:17 csharp Hello there!
11:17 sandbergja joined #evergreen
11:20 pinesol [evergreen|Jane Sandberg] LP1665052: Fix checkin workstation display on recent circs view - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=6de97f9>
11:28 bshum berick: FYI, I'm going to push the websocketd to OpenSRF master.  Worked for me and I like Dyrcona's doc updates too.
11:28 berick bshum++
11:29 bshum berick++ Dyrcona++ jeffdavis++ # trailblazers
11:29 berick then i can merge the ansible branch
11:30 pinesol [opensrf|Bill Erickson] LP#1777180 Websocketd gateway and test scripts - <http://git.evergreen-ils.org/?p​=OpenSRF.git;a=commit;h=21c9c76>
11:30 pinesol [opensrf|Jason Stephenson] LP#1777180 Update README for websocketd - <http://git.evergreen-ils.org/?p​=OpenSRF.git;a=commit;h=d1c33b4>
11:31 csharp okay for bug 1735816, after applying the patch, we're not seeing already-entered copy notes when we click copy notes - I can confirm they are created in the DB
11:31 pinesol Launchpad bug 1735816 in Evergreen "Cannot Delete Existing Copy Notes" [Medium,Confirmed] https://launchpad.net/bugs/1735816
11:31 csharp gathering console output now
11:33 csharp https://pastebin.com/MpTDEThb - no actual errors showing
11:35 JBoyer :/ Have you done all of the clear cache/hard reload voodoo that Chrome wants you too doo?
11:37 csharp yeppers
11:38 blongwell joined #evergreen
11:40 bshum sandbergja++ # your testing notes are great!
11:42 sandbergja Thanks!
11:43 JBoyer Well that's fabulous, it looks like that patch has fallen off of production here... and all of our other machines. woo.
11:43 JBoyer One moment.
11:46 pinesol [evergreen|Jason Stephenson] Lp 1746251: Fix Circulation Modifier typos. - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=b549d59>
11:48 bshum sandbergja: Well I like them mainly because I clearly don't remember how to actually use Evergreen anymore :(
11:48 bshum So knowing which menus to click through or interfaces to look at saves me time, haha
11:50 khuckins joined #evergreen
11:58 sandbergja joined #evergreen
11:59 yboston joined #evergreen
12:00 JBoyer csharp, annoyingly, they are showing here one the one server that apparently has that patch installed.
12:03 jihpringle joined #evergreen
12:03 * JBoyer will investigate post-lunch
12:24 beanjammin joined #evergreen
12:34 pinesol [evergreen|Jane Sandberg] LP1775640: Adds deposit columns to Item Status and Holdings View grids - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=f3e7a33>
12:37 yboston joined #evergreen
12:48 pinesol [evergreen|Steven Callender] LP1792014: KPAC fix for opac.holds.org_unit_not_pickup_lib in org_selector. - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=7567914>
12:53 montgoc1 joined #evergreen
12:53 bshum berick: I'm curious, if using the websocketd ansible branch, do we have to install unzip too?  (the unarchive command documentation seems to indicate that we need it installed to do the work)
12:53 beanjammin joined #evergreen
12:53 bshum And I don't see it defined in the installer
12:56 berick bshum: yes, you probably do
12:56 berick should probably add that to the installer
12:56 bshum berick: Do you think we should just toss it into the prereqs for OpenSRF in general?  So that folks don't trip on that either if doing it manually?
12:57 bshum I mean most people would notice when they went to run the unzip command and it fails
12:57 bshum And know to get the unzip package
12:57 berick bshum: i would say so, I install zip/unzip by default everywhere.
12:58 berick or add it to the websocketd docs, at least
12:58 Dyrcona We should add zip to the prereqs for OpenSRF. Installing the zip package typically installs unzip, too.
12:59 Dyrcona Or we could just install unzip.
13:00 aabbee left #evergreen
13:00 aabbee joined #evergreen
13:02 khuckins joined #evergreen
13:05 bshum berick: I'll mock up a branch to add those to the prereq list
13:05 bshum I consider it a continuation of the websocketd stuff, so I'll tag it with that LP#
13:06 Dyrcona bshum +1
13:09 bshum http://git.evergreen-ils.org/?p=work​ing/OpenSRF.git;a=shortlog;h=refs/he​ads/user/bshum/lp1777180-continued
13:14 Dyrcona bshum: Works for me, I'll signoff and push it.
13:14 bshum Dyrcona++
13:14 beanjammin joined #evergreen
13:18 pinesol [opensrf|Ben Shum] LP#1777180: Add zip/unzip to prerequisites - <http://git.evergreen-ils.org/?p​=OpenSRF.git;a=commit;h=da7e927>
13:21 dkyle1 joined #evergreen
13:25 bshum phasefx: FYI, I'm thinking to rebase the branches to eliminate wheezy install targets
13:25 bshum And push those through, since Wheezy is definitely EOL
13:25 bshum I know you're in the process of building a new live test server, but I just wanted to give you a head's up before I go ahead
13:26 miker bshum: IMO, that's kinda a version-y thing... we know that up to 3.1 works on wheezy, so I'd say remove it for master only. thoughts?
13:26 bshum miker: Yes I was only going to push that change for master only
13:26 bshum (for 3.2)
13:26 phasefx sounds good to me
13:26 miker cool, thanks
13:26 bshum I didn't think to backport that removal
13:26 bshum Seemed far too impactful :D
13:27 bshum But I'm glad you brought it up
13:27 miker yeah .. "hey, why won't my 3.1 upgrade anymore?"
13:27 bshum miker: Yeah, lol
13:28 Dyrcona "Tempted by the fruit of another..." Wait, wrong Squeeze.
13:33 terran joined #evergreen
13:35 aabbee any chance i could get a second opinion on bug 1731272? on the one hand, i don't know if the stupid fix actually works (works on *my* machine...). on the other hand, there's gotta be a better way. on the THIRD hand, how worthwhile is it to try to fix this when ang6 is seemingly just around the corner?
13:35 pinesol Launchpad bug 1731272 in Evergreen "web client: "Set default view" breaks record page loading" [High,Confirmed] https://launchpad.net/bugs/1731272 - Assigned to Galen Charlton (gmc)
13:39 bshum Does anyone have any strong opinions about the naming of the example folder for apache_24 vs. just plain apache?  with the removal of wheezy, we can just retain one set of configs and keep the folder named apache_24 in case there's further future versions?  Or just fall back to having just everything in an "apache" folder?
13:39 miker aabbee: I suspect the approach is correct, but the ng way of spelling setTimeout is $timeout()
13:42 aabbee miker: ok, that's easy enough to change. as far as i can tell, the only reason we need to muck around with the event queue anyway is because we're using angularjs promises instead of ecma6 promises. but if the whole thing is being rewritten for ang6 anyway, why bother, y'kno?
13:43 miker aabbee: well, there's no telling when the ang6 version will drop for production use, so it's good to fix, IMO
13:44 berick yeah, we def. want to keep fixing the bugs
13:45 aabbee *if* it actually fixes the problem and doesn't introduce any new problems (going back to: "there has got to be a better way"). hoping to get confirmation from Jane whether or not the fix even works, but thanks for the insight.
13:48 miker fwiw, there are lots of spots where we have to sprinkle a $timeout() due to apply/digest race conditions
13:49 miker I'm glad we don't support IE ... think of all the phantom click sounds!
13:49 berick that means it's working!
13:50 berick well, not functioning, of course, but putting forth effort
13:54 jihpringle aabbee: that fix is on my list to test in a sandbox later today
13:55 aabbee jihpringle: great news, thank you! i hope it just misplaced the line when it was merged and is an easy fix. looking forward to knowing either way.
14:18 pinesol [opensrf|Jason Stephenson] Lp#1718459: Remove Debian 7 Wheezy installation support. - <http://git.evergreen-ils.org/?p​=OpenSRF.git;a=commit;h=265aa9f>
14:20 rlefaive joined #evergreen
14:20 montgoc01 joined #evergreen
14:20 pinesol [evergreen|Jason Stephenson] Lp#1718459: Remove Debian 7 Wheezy Installation Support - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=c6437c2>
14:23 bshum Reading that stuff, it looks like our server_upgrade.adoc hasn't been updated for master since 2.12
14:23 bshum I guess we haven't been forward porting those from the release branches
14:27 jihpringle joined #evergreen
14:28 pinesol [evergreen|Ben Shum] LP#1718459: Remove apache 2.2 language from README - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=bacd986>
14:36 sandbergja joined #evergreen
14:36 pinesol [evergreen|Jane Sandberg] LP1743654: Fleshing out the Created By column in Item Status - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=5eaa171>
14:36 pinesol [evergreen|Jane Sandberg] LP 1743654: Removing a duplicate flesh_fields array entry - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=be9f2cc>
14:40 pinesol [evergreen|Cesar Velez] LP#1776917 - fix patron bill grid lacking bill ID - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=cf2298b>
14:51 pinesol [evergreen|Chris Sharp] LP#1269574 - Exclude backordered copies from deletion. - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=8689047>
14:57 rlefaive joined #evergreen
14:58 khuckins_ joined #evergreen
15:02 yboston joined #evergreen
15:10 bshum berick: I'm going to try using the latest ansible from the PPA and see if it helps us get around that npm problem.  The github bug you linked to seems to indicate that it's "fixed" in 2.3+ and the ansible version that ships with Ubuntu 16.04 is 2.0.ish
15:10 bshum So... fingers crossed that nothing bigger blows up?  :D
15:11 bshum Ansible 2.6, whee
15:12 JBoyer bshum, I can confirm that it should work with 2.6. ;)
15:12 bshum JBoyer++ # cool!  I'll find out in a few moments too, but glad to know that it'll be good
15:14 bshum If it's good, then I think we can just modify the readme to include the PPA step prior to installing ansible and get back to a happy place then.
15:15 berick bshum++
15:15 JBoyer Ubuntu is really irritating in that regard. If I wanted to have to constantly add vendor repositories to get access to modern-ish versions of things I could always run RedHat. :/
15:15 berick oh, need to merge websocketd branch..
15:16 Dyrcona JBoyer: LFS.
15:17 berick bshum: just merged websocketd commits
15:17 bshum berick++
15:21 * bshum waits for the magic to happen now
15:22 mmorgan left #evergreen
15:27 bshum Aww, it still was unhappy with npm :(
15:30 JBoyer Oh. :/
15:30 JBoyer No more than I've looked at that branch I'm not sure what it's doing exactly, what's the big ugly red blob of an error say?
15:33 berick JBoyer: the problem is it won't run 'npm install'
15:33 berick if you've already done that in the past, it will contineu working
15:33 bshum Yeah, npm install is my problem
15:33 berick rather, you won't notice the issue
15:33 bshum I just ran them manually and continuing my merry way I guess
15:33 JBoyer Ah. Which I have. I see.
15:34 JBoyer Is it actually an ansible thing or is that just where it's most apparent?
15:35 jeff for context, where is the github issue in question?
15:35 bshum Running the steps manually, I haven't had a problem yet.
15:35 bshum So I think it's only with ansible on clean VMs without the stuff installed
15:37 bshum http://paste.evergreen-ils.org/14​100?wr=on&amp;submit=Format+it%21
15:37 bshum JBoyer: This is the error that's shown on my screen
15:38 bshum I'm going to check for that debug log next
15:39 jeff yes, i'd be interested in seeing the contents of the referenced /home/benjamin/.npm/_logs/2018​-09-12T19_26_40_044Z-debug.log file
15:39 bshum Hmm
15:40 pastebot "bshum" at 64.57.241.14 pasted "debug npm log" (1428 lines) at http://paste.evergreen-ils.org/14101
15:41 jeff oh, it's:
15:41 jeff "npm ERR! Command failed: /usr/bin/git clone --depth=1 -q -b npm git://github.com/rxfork/ngOrderObjectBy.git /home/benjamin/.npm/_cacach​e/tmp/git-clone-83a9574e",
15:41 jeff "npm ERR! /home/benjamin/.npm/_cacache/​tmp/git-clone-83a9574e/.git: Permission denied",
15:41 bshum Yeah that
15:41 JBoyer I wonder what user it's run as. I should really look at this playbook directly rather than just guessing what it's doing. Is this in berick's github or somewhere else?
15:41 jeff jq can be handy for formatting the JSON that ansible spits out.
15:42 berick JBoyer: yeah, mine
15:42 * berick reads up
15:42 jeff though it's still a lot of blobs of text with escaped newlines, at least it breaks things into nicely formatted key/value (blob) pairs. :-)
15:42 JBoyer Or just a s/\\n/\n/g :)
15:43 berick bshum: is /home/benjamin/.npm/_cacache​/tmp/git-clone-83a9574e/.git owned by root?  (or parent directory)?
15:46 berick maybe need to set "become_method: su" for the npm stuff so it doesn't put root-owned stuff into the login user's home dir.
15:47 jeff why are you running npm as root?
15:47 berick jeff: only for the install -g @angular/cli bit
15:47 berick i wonder if we can get rid of that, though
15:48 JBoyer I thought all of the stuff that needed to be done by root was in <blah>-developer?
15:48 bshum berick: Hmm, that directory doesn't exist on my test system anymore, but the folders leading up to it seem to be owned by me
15:48 JBoyer my entire build (after prereqs) is run by opensrf until install time.
15:49 berick JBoyer: it is...
15:49 bshum Would it matter that I'm using myself benjamin, rather than opensrf to run the step though?  Since the user is opensrf for the Evergreen repo
15:51 JBoyer It looks like the user that is running npm at that point needs write access to ~ and ./, so long as both of those are true it shouldn't matter who is it.
15:51 berick i think i just have an unnecessary 'become' or two in there
15:52 berick heh, yml using "-" makes looking at diff's weird
15:52 JBoyer I'll admit, I haven't looked at the npm module yet so I don't know what it's doing, but I'd remove the become for the AngularJS Node Build task and see if that makes it happy.
15:53 JBoyer --its
15:53 JBoyer ---dashes all
15:53 JBoyer -----the way down
15:53 JBoyer +- oops.
15:53 berick https://gist.github.com/berick/6​a5f9f111de78f23dbfd8f17227fa534 -- for example ;)
15:53 jeff berick: ah, yeah. looks like installing @angular/cli locally is possible, but weird.
15:54 berick anyway, the 2 "becomes" are what I'm trying to remove
15:54 * berick doesn't have a clean vm handy at the moment to test
15:54 bshum I can test that out on my busted VM with the error :)
15:55 bshum But I'll retest it on a clean VM afterwards
15:55 berick bshum: cool, may have to do some chown'ing first
15:55 bshum I was just going to blow away the directories and start over too, but yeah I getcha
15:55 berick the ultimate chown
15:56 berick you belong to the void now, little directory
15:59 JBoyer berick++
16:03 bshum berick: Still seems unhappy, I assumed I was blowing away the .npm directories
16:04 bshum I'll poke at this stuff more closely later
16:04 berick k, i will too
16:04 berick wondering now if the preceeding 'become' on make -f ...Makefile.install needs to change
16:04 berick it needs to be there, of course
16:04 berick but maybe modified
16:23 Dyrcona The wheezy removal branch broke it.
16:25 berick ah, i'll def spend some time on it in the morning then
16:26 bshum Failed to restart websocketd-osrf.service: Unit websocketd-osrf.service not found.
16:26 bshum berick: Well we're making some progress :)
16:26 bshum I got almost to the end this time
16:26 berick bshum: you enabled service files setting?
16:26 bshum For the logs, adding "become_user: opensrf" to the angularjs step helped
16:27 bshum berick: Aha, it's set to "false" by default
16:27 berick yeah, thought that might be presumptious
16:27 bshum Gotcha
16:27 berick or however that word is spelled
16:27 bshum Then perhaps we need to add something for the ansible step so that it doesn't call the websocketd by default
16:28 bshum If it's not selected
16:28 berick ah, yes, definitely
16:28 berick faulty assumption in the code
16:28 bshum Okay, time to blow away the whole VM and build a new one to retest cleanly :)
16:28 bshum Or maybe drive home first and deal with it later
16:30 bshum berick: I'll make a new branch with all the fixes I've noticed that we need
16:30 berick bshum++
16:41 bshum berick: It's presumptuous, but if a user did not setup the osrfwebsocket service, they'd have to start that manually, right?
16:42 bshum If we don't set it to true
16:42 * bshum ponders that further
16:43 berick right
16:44 bshum Maybe we should split the services install options
16:44 bshum And then install the one for osrfwebsocket by default so that we can use it and start it up
16:44 bshum I think for majority users, we'd want that option to exist and be turned on
16:44 bshum Or at least, I would :D
16:45 berick yeah, +1 to that.  it is an OS-specific installer and this just makes that whole thing easier
16:49 bshum berick: I'll put all those thoughts into a branch and test everyone on fresh VMs tomorrow
16:50 bshum Thanks!
16:50 bshum Signing off for now, have a nice day y'all
16:50 Bmagic lata
16:51 berick later bshum
16:51 berick and thanks again
17:50 khuckins joined #evergreen
19:20 Christineb joined #evergreen

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