Time |
Nick |
Message |
06:00 |
pinesol |
News from qatests: Testing Success <http://testing.evergreen-ils.org/~live> |
07:21 |
|
collum joined #evergreen |
08:31 |
|
mantis joined #evergreen |
08:31 |
|
stephengwills joined #evergreen |
08:52 |
|
mantis joined #evergreen |
09:07 |
|
Dyrcona joined #evergreen |
10:21 |
jeff |
well that's annoying. I'm only receiving a subset of launchpad emails. |
10:25 |
Dyrcona |
jeff: How can you tell? Also what groups are you in? |
10:37 |
pinesol |
[evergreen|Garry Collum] LP1908612 Bootstrap Opac - Limit to Available - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=6d4c6b6> |
10:44 |
|
tlittle joined #evergreen |
11:03 |
jeff |
I received the original report on bug 1936986 and the "Thank you! It works now!", but no email related to Terran's comment on that bug. |
11:03 |
pinesol |
Launchpad bug 1936986 in Evergreen "Hatch chrome extension affects evergreen website" [Undecided,Invalid] https://launchpad.net/bugs/1936986 |
11:05 |
jeff |
also looked like I had some missing emails related to bug 1937290 but 1) i think they might have shown up later and 2) I don't remember exactly what I thought was missing, and think I chalked it up to it being weird because of it being marked duplicate. |
11:05 |
pinesol |
Launchpad bug 1937299 in Evergreen "duplicate for #1937290 Multiple issues with the Collections: Exempt user setting" [Medium,Confirmed] https://launchpad.net/bugs/1937299 |
11:07 |
Dyrcona |
OK. That's more info than I bargained for. AFAICT, I don't appear to be missing any emails. |
11:08 |
|
nfBurton joined #evergreen |
11:10 |
Dyrcona |
Email can be flaky. |
11:46 |
tlittle |
I have a Perl question that I'm finally frustrated enough to ask about. I'm trying to edit the create_volume sub in Order.pm, to go out and check before creating volumes if the bib is deleted. If so, undelete the bib. I'm able to get down and connect to the undelete sub, but it's not picking up the authtoken. It tries to retrieve the authtoken |
11:46 |
tlittle |
using the bib id value instead. |
11:46 |
tlittle |
Can anyone see what I'm doing wrong? https://www.diffchecker.com/Kg3ytc7y |
12:03 |
berick |
tlittle: you'll want to call undelete_biblio_record_entry() in a slightly different fashion |
12:06 |
berick |
it's an api call running in a different service, so better to call it like an API call instead of just running the function |
12:08 |
berick |
tlittle: something along the lines of https://gist.github.com/berick/dfb38340bb935b40f31df3a22dd078a7 |
12:08 |
berick |
$resp will be '1' on success, Event hash on error |
12:10 |
berick |
the call to OpenILS::Application::Cat::AssetCommon->find_or_create_volume() can be made directly because OpenILS::Application::Cat::AssetCommon is explicitly a set of shared functions and not API calls |
12:10 |
tlittle |
Oh, that helps! You beat me to it, I was just about to ask why that one was different |
12:12 |
tlittle |
Great, thank you so much! And that's good info about why the AssetCommon is different, I'll know to look at that next time. berick++ |
12:16 |
Dyrcona |
berick++ |
12:21 |
Dyrcona |
A better patch might be to move the functional code for undelete_biblio_record_entry to BibCommon.pm and call that from Cat.pm and AssetCommon.pm. Just throwin' that out there. |
12:25 |
Dyrcona |
So, my upgrade to master-d9bb7c9205 from 3.5.3 took 25 hours, and that's not counting the "did you mean" setup. |
12:25 |
pinesol |
Dyrcona: [evergreen|Jane Sandberg] LP1718782: follow up to fix failing test - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=d9bb7c9> |
12:34 |
|
collum joined #evergreen |
12:54 |
|
collum joined #evergreen |
13:20 |
|
abowling joined #evergreen |
13:20 |
abowling |
Currently trying to build from source and getting this error: |
13:21 |
abowling |
npm ERR! command git --no-replace-objects ls-remote ssh://gitgithub.com/rxfork/ngOrderObjectBy.git |
13:21 |
abowling |
npm ERR! Warning: Permanently added the RSA host key for IP address '140.82.113.3' to the list of known hosts. |
13:21 |
abowling |
npm ERR! gitgithub.com: Permission denied (publickey). |
13:21 |
abowling |
npm ERR! fatal: Could not read from remote repository. |
13:21 |
abowling |
anyone run into this and have a fix or workaround? |
13:21 |
Dyrcona |
npm-- |
13:33 |
jeffdavis |
Depending on your circumstances, as a temporary workaround you could skip building the web client and copy over the client code from another server. |
13:33 |
abowling |
Dyrcona, agree |
13:39 |
abowling |
jeffdavis, that's likely the next step |
13:41 |
Dyrcona |
abowling: I think you can fix that by adding a ssh key on github. If you already have one, make sure that you're using ssh-agent and agent forwarding or add your key on the VM where you're working. |
13:42 |
berick |
abowling: what branch? |
13:42 |
Dyrcona |
If that doesn't fix it, then it looks like someone has misconfigured the source for the ngOrderObjectBy package. It should probably use git: or https: for the protocol. |
13:43 |
Dyrcona |
berick: It's a npm/Node.js error. I'm not sure it is something that we can fix, but I'll defer to you if you think so. |
13:44 |
berick |
just curious about what versions it's trying to install |
13:44 |
Dyrcona |
Ah ha! I misconstrued. |
13:45 |
berick |
npm install on master working OK for me |
13:47 |
Dyrcona |
Could be one of those things or specific to an older release... |
14:27 |
collum |
Hmm. https://catalog.kentonlibrary.org - In the adult carousel click the previous button. Maybe we need a limit for title max string size in the carousel definitions. |
14:35 |
berick |
collum: :) |
14:50 |
JBoyer |
That's some quality 19th-century style naming there. |
15:00 |
pinesol |
[evergreen|gmontimantis] Update lsa-address_alert.adoc - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=c5af141> |
15:00 |
pinesol |
[evergreen|Gina Monti] Docs: LP1776977: Update lsa-address_alert.adoc and add images - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=d323b83> |
15:02 |
|
abowling joined #evergreen |
15:02 |
abowling |
berick: master |
15:05 |
berick |
abowling: just doing 'npm install' ? not 'npm update' ? |
15:19 |
abowling |
berick: correct |
15:20 |
berick |
abowling: huh, keeps happening if you try again? |
15:20 |
abowling |
Dyrcona++ ish ;) |
15:20 |
abowling |
adding the key now lets me ssh to github, but the package is still failing with the same error |
15:21 |
abowling |
berick: yeah. darndest thing. i've never run into this on other boxes |
15:21 |
abowling |
this is an AWS VM. i don't think that should matter, but... |
15:22 |
JBoyer |
abowling, anyhing interesting in your ~/.ssh/config? Seems odd that it's trying to use ssh to clone from github |
15:23 |
berick |
i think what Dyrcona said is the main issue, the package should not be using ssh: |
15:23 |
berick |
just tested on a VM with github key disabled and got the same error |
15:23 |
abowling |
JBoyer: berick: agree |
15:25 |
abowling |
JBoyer: not at all; straightforward |
15:26 |
JBoyer |
seeing berick's note about *removing* his makes my question kind of unnecessary. Agreed that Dyrcona nailed it above. Someone screwed up upstream. :/ |
15:28 |
JBoyer |
So if anyone wants to build anything that touches that package any time soon I guess make sure your ssh key is on your github account. :-/ Tons of fun for ephemeral / one-off machines or builds. |
15:28 |
abowling |
JBoyer: which is EXACTLY what this is |
15:29 |
berick |
correction.. I tried an actual npm install (instaed of just the git command) with my github keys disabled and it worked fine :\ |
15:29 |
Dyrcona |
I use ssh-agent to forward my ssh keys, 'cause I use git to distribute code, etc. |
15:29 |
berick |
the raw git command fails, but npm install completes OK |
15:30 |
JBoyer |
weird. |
15:30 |
Dyrcona |
What JBoyer said. :) |
15:31 |
Dyrcona |
On Windows, you can forward your keys with Pagent. |
15:31 |
Dyrcona |
If you use putty, that is. |
15:32 |
JBoyer |
abowling, which npm step, JS client, Angular client, or the bootstrap opac? |
15:32 |
abowling |
JBoyer: js |
15:41 |
JBoyer |
abowling, you might try rm -rf node_modules/angular-order-by - I just did a fresh build and it may be fixed already? But if you still have the busted version it may not re-download it if it fits the version requirements. |
15:42 |
JBoyer |
(There's always rm -rf package-lock.json node_modules, but that *shouldn't* be necessary.) |
15:42 |
abowling |
so, um, this works just fine |
15:42 |
abowling |
"angular-order-object-by": "1.3.0", |
15:42 |
abowling |
in package.json |
15:43 |
abowling |
any reason why we shouldn't just remove the github ref altogether? |
15:44 |
berick |
abowling: good catch. likely the github ref was needed a long time ago, but not anymore |
15:45 |
abowling |
berick++ makes sense |
15:45 |
JBoyer |
Oh, I didn't realize *we* were using github; I thought the package was doing something weird as part of its build process. That's absolutely the way to go then. |
15:46 |
JBoyer |
Also your aws instance is haunted; no idea why it keeps trying to use ssh://github... |
15:47 |
Dyrcona |
npm-- |
16:03 |
abowling |
JBoyer: Blue Origin told it to |
16:03 |
abowling |
I've gone ahead and pushed the change to working: https://git.evergreen-ils.org/?p=working/Evergreen.git;a=commit;h=c15d9b197371a8af373f2587f9e39d326eabb8ef |
16:10 |
Dyrcona |
Ah, so it was us... |
16:11 |
Dyrcona |
abowling++ |
17:19 |
|
denials joined #evergreen |
18:00 |
pinesol |
News from qatests: Testing Success <http://testing.evergreen-ils.org/~live> |
19:00 |
|
stephengwills left #evergreen |
20:43 |
|
stephengwills_ joined #evergreen |
20:44 |
|
stephengwills_ left #evergreen |