| Time |
Nick |
Message |
| 00:35 |
|
sandbergja joined #evergreen |
| 00:48 |
|
cmalm joined #evergreen |
| 07:13 |
|
rjackson_isl joined #evergreen |
| 07:30 |
|
rfrasur joined #evergreen |
| 07:58 |
|
tlittle joined #evergreen |
| 08:45 |
|
mmorgan joined #evergreen |
| 08:49 |
|
mantis1 joined #evergreen |
| 08:49 |
|
mantis1 left #evergreen |
| 08:49 |
|
Dyrcona joined #evergreen |
| 09:04 |
|
stephengwills joined #evergreen |
| 09:20 |
|
Bmagic joined #evergreen |
| 09:36 |
|
mllewellyn joined #evergreen |
| 09:46 |
|
mllewellyn left #evergreen |
| 09:47 |
Bmagic |
Dyrcona: did you say that you were running 16.04 on production bricks? Any reason to hold postpone 18.04? |
| 09:49 |
alynn26 |
gmcharlt++ |
| 09:50 |
|
tlittle joined #evergreen |
| 09:53 |
pinesol |
[evergreen|Andrea Buntz Neiman] Docs: corrected image file reference in carousels.adoc - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=2d6e15f> |
| 09:53 |
Dyrcona |
Bmagic: We are. |
| 09:57 |
Dyrcona |
Bmagic: I'm not entirely sure if bug 1801163 applies. |
| 09:57 |
pinesol |
Launchpad bug 1801163 in Evergreen "SendEmail A/T reactor broken for recent version of Encode::MIME::Header" [High,Confirmed] https://launchpad.net/bugs/1801163 - Assigned to Jeff Godin (jgodin) |
| 09:58 |
Dyrcona |
I'll check since I was about to update a test vm that's on 18.04 anyway. |
| 09:58 |
Bmagic |
at this point, 18.04 is still not "fully" functional an issue with Evergreen? |
| 09:59 |
Dyrcona |
Well, can we say we support a distro if Evergreen is not fully functional? |
| 09:59 |
Bmagic |
It's debatable I'm sure |
| 10:00 |
Bmagic |
if the Evergreen brick can't send email... then I am of the mindset that the install instructions would say that |
| 10:00 |
Dyrcona |
Well, the install instructions don't say that. |
| 10:01 |
berick |
Bmagic: note bug 1855199 (likely not an issue for most prod sites) |
| 10:01 |
pinesol |
Launchpad bug 1855199 in Evergreen "Vandelay record queuing can fail if spool directory is /tmp" [Medium,Confirmed] https://launchpad.net/bugs/1855199 |
| 10:01 |
Bmagic |
If they did say that, I probably wouldn't have posed the question here |
| 10:02 |
berick |
that's the only 18.04 issue i've encountered on dev machines. of course, I'm not sending email |
| 10:02 |
pinesol |
News from qatests: Failed Building the AsciiDoc output formats <http://testing.evergreen-ils.org/~live/test.81.html#2019-12-06T09:38:30,140377482-0500 -0> |
| 10:02 |
Dyrcona |
Bmagic: I would say that Ubuntu 18.04 is affected by 1801163 give that Encode is version 2.88 and Encode::MIME::Header is 2.24. |
| 10:03 |
Dyrcona |
It also probably affects Debian Stretch and Buster. |
| 10:03 |
Bmagic |
what a bummer |
| 10:04 |
Dyrcona |
I think we need to revamp how we send email, particularly from A/T, given that there are bugs to be able to send HTML email as well as emails in the patron's native language. |
| 10:05 |
Dyrcona |
We've also been relying too much on bugs/undocumented features of various modules, particularly Encode. Seems like every 2nd time a bug is fixed in Encode, we have make changes to Evergreen. |
| 10:06 |
csharp |
your fix for bug 1801163 looks pretty straightforward |
| 10:06 |
pinesol |
Launchpad bug 1801163 in Evergreen "SendEmail A/T reactor broken for recent version of Encode::MIME::Header" [High,Confirmed] https://launchpad.net/bugs/1801163 - Assigned to Jeff Godin (jgodin) |
| 10:07 |
csharp |
we're on the precipice of upgrading to 18.04, but probably not with next month's EG upgrade |
| 10:07 |
Dyrcona |
I think jeff wants to improve it or do something different. I'm also not personally thrilled with the "fix." |
| 10:08 |
Dyrcona |
We're using 18.04 on 1 utility VM that runs the fine generator and a couple of other crons that do not send email themselves, though the standard output gets emailed to a local admin list. |
| 10:08 |
jeff |
hrm. i should look at that again. |
| 10:09 |
jeff |
especially since it's assigned to me (oops). |
| 10:09 |
Dyrcona |
:) |
| 10:09 |
dbs |
I guess iteratively stamping out bugs related to Encode's seemingly continual discovery of new ways in which their grafted-on encoding support has failed is one of the downsides of working in perl5 |
| 10:09 |
jeff |
i think i was under the impression that someone had pushed a quick-and-dirty fix and closed the bug. |
| 10:10 |
jeff |
to be fair to Encode, I believe we (or one of our deps) were using it wrong in this case. |
| 10:11 |
Dyrcona |
jeff: The deprecated Email::Send module is using it wrong, in the sense that it got away with working via a bug in Encode, yes. |
| 10:11 |
Dyrcona |
cf. bug 1466502 |
| 10:11 |
pinesol |
Launchpad bug 1466502 in Evergreen "Replace Email::Send with Email::Sender" [Wishlist,Confirmed] https://launchpad.net/bugs/1466502 |
| 10:12 |
jeff |
I may have gotten to the point of "and now we'll need to change every deployed A/T SendEmail template" and decided to sleep on it. |
| 10:13 |
jeff |
I'll take a fresh look. |
| 10:13 |
Dyrcona |
Yeahp. I think you're right. |
| 10:13 |
Dyrcona |
About changing the templates. |
| 10:14 |
* dbs |
found that using the Cookie Autodelete extension in Firefox to greylist the test server hostname and then hitting the trash icon for the hostname (keeping localstorage and cookies) fixed the "Possibly unhandled rejection: {"isTrusted":true}" webby white screen problem |
| 10:14 |
Dyrcona |
There are a number of related wish list items on Launchpad that would require similarly radical changes. |
| 10:15 |
Dyrcona |
dbs: Good to know. I haven't encountered that issue switching from production to test/training servers. |
| 10:18 |
csharp |
we're happy to reconfigure our (many) templates if it keeps us from chasing our tail in the future (fwiw) |
| 10:19 |
* jeff |
nods |
| 10:21 |
Dyrcona |
csharp | jeff: same here. |
| 10:23 |
pinesol |
Showing latest 5 of 6 commits to Evergreen... |
| 10:23 |
pinesol |
[evergreen|Bill Erickson] LP1830923 Marc edit in-place additions - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=b2de21e> |
| 10:23 |
pinesol |
[evergreen|Bill Erickson] LP1830923 Vandelay queued record editing - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=1623058> |
| 10:23 |
pinesol |
[evergreen|Galen Charlton] LP#1830923: add support for editing queued authority records - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=d5c058d> |
| 10:23 |
pinesol |
[evergreen|Galen Charlton] LP#1830923: do not show Enhanced MARC editor tab when editing in-place - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=b20d457> |
| 10:23 |
pinesol |
[evergreen|Bill Erickson] LP1830923 Replace some pesky tabs - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=f1f34b2> |
| 10:23 |
Dyrcona |
And, I rebased my local branches yesterday..... :) |
| 10:26 |
* Dyrcona |
looks at his branch for 1801163, again, just to jog the memory. |
| 10:29 |
Dyrcona |
It needs to be updated for buster, and doesn't look to me like it requires template changes. It could be a useful stop gap. |
| 10:31 |
Dyrcona |
It may do some unnecessary things, like guaranteeing only 1 of each header. I don't recall if that was necessary for Email::MIME or not. |
| 10:49 |
* csharp |
still thinks of this Buster when Debian buster is mentioned: https://bit.ly/2RskeF8 |
| 10:49 |
dbs |
me too, csharp |
| 10:51 |
csharp |
@band add Unlimited Juice |
| 10:51 |
pinesol |
csharp: Band 'Unlimited Juice' added to list |
| 10:52 |
csharp |
@who is going to see [band] who is opening for [band] this weekend? |
| 10:52 |
pinesol |
rfrasur is going to see Evil Placebo who is opening for Osrf Websocket Studio this weekend. |
| 10:53 |
mmorgan |
:) |
| 10:53 |
rfrasur |
I mean....I'm cool with that. |
| 10:53 |
csharp |
@praise rfrasur for going to see [band] |
| 10:53 |
* pinesol |
rfrasur LOVES the RESISTANCE! for going to see Slosh so helped |
| 10:54 |
rfrasur |
I do also love the Resistance. |
| 10:55 |
berick |
that's a fun weekend |
| 10:56 |
berick |
though I'd expect Evil Placebo to headline |
| 10:58 |
csharp |
berick++ |
| 10:59 |
|
sandbergja joined #evergreen |
| 11:00 |
rfrasur |
I mean...OSRF Websocket Studio is a cultural touchstone band that'll probably always headline. |
| 11:00 |
dbs |
Dyrcona: that white screen resoution was mostly for the logs |
| 11:00 |
Dyrcona |
Yeahp. |
| 11:01 |
berick |
rfrasur: I don't even know who you are anymore |
| 11:01 |
rfrasur |
berick: me either. |
| 11:01 |
Dyrcona |
:) |
| 11:02 |
* rfrasur |
looks at wishlist tickets while humming "don't go chasing waterfalls." |
| 11:02 |
dbs |
these days I grep my own logs, as neither search option on http://irc.evergreen-ils.org/evergreen/ seems to return much; for "white screen", native search stops in 2014? and google search turns up almost nothing |
| 11:05 |
Dyrcona |
dbs: I start with my own logs, too. |
| 11:05 |
Dyrcona |
Older IRC logs are here: http://evergreen-ils.org/irc_logs/evergreen/ . Good luck searching them. |
| 11:09 |
jeff |
likewise with "own logs". there might be a process for indexing the ilbot search. i'll make a note to look into it. |
| 11:09 |
jeff |
(it might be dead/hung/forgotten/etc) |
| 11:11 |
pinesol |
[evergreen|Dan Briem] LP#1803406 Due date box in check out has display issues at wider resolutions - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=1711f0f> |
| 11:47 |
csharp |
@band add Forgotten Logs |
| 11:47 |
pinesol |
csharp: Band 'Forgotten Logs' added to list |
| 11:50 |
dbs |
hrm, maybe Cookie AutoDelete didn't resolve the problem; ran into another whitescreen, this time "Error loading shared worker" (message: "DataCloneError: The object could not be cloned." filename: "https://hostname/js/ui/default/staff/offline-db-worker.js") |
| 11:58 |
Dyrcona |
dbs: Latest Firefox? |
| 11:59 |
Dyrcona |
dbs: Also, I'm assuming this is 3.4, not 3.1. Correct me if that assumption is wrong. |
| 12:00 |
dbs |
Dyrcona: Firefox 71 (Linux), Evergreen 3.4 |
| 12:01 |
dbs |
Chrome in Incognito Mode is my fallback, and I haven't run into that problem there. Yet. |
| 12:01 |
Dyrcona |
I have Firefox 70. I'm going to check something after lunch. You might want to take a look at the little shield to the left of the location bar. Tracking protection may be kicking in. |
| 12:02 |
dbs |
Dyrcona: no shield (alas) |
| 12:02 |
dbs |
Err, there's a shield, but it's grey: no trackers were detected for this site |
| 12:04 |
dbs |
Also disabled all of my plugins and hit Shift-Reload |
| 12:08 |
dbs |
Manually deleted everything under the Storage Inspector, reloaded, logged in, and then hit "Error loading shared worker: NotFoundError: The operation failed because the requested database object could not be found. For example, an object store did not exist but was being opened" |
| 12:08 |
dbs |
(Makes some sense because I deleted all of the storage I guess) |
| 12:10 |
* dbs |
scrolls through 146 service workers in about:debugging#workers |
| 12:13 |
dbs |
Unregistered the service worker, then deleted all cookies and local storage, reloaded the page, logged in, registered and used a workstation, and whitescreened again :( |
| 12:13 |
dbs |
Back to "DataCloneError: The object could not be cloned." |
| 12:15 |
dbs |
Time to reboot into Windows and see if I can get this Dymo LabelWriter Turbo 330 set up with Hatch... |
| 12:16 |
|
gsams joined #evergreen |
| 12:44 |
|
dbwells joined #evergreen |
| 13:14 |
berick |
tried to recreate the DataCloneError in FF 70 on master, no dice. pleased to see FF shows shared worker logs along with the main page logs. |
| 13:15 |
Dyrcona |
berick: I was just getting ready to test it. |
| 13:16 |
Dyrcona |
Except I got an Internal Server Error, so I probably have something broken on my vm. |
| 13:18 |
Dyrcona |
I think my problem is ejabberd. |
| 13:20 |
Dyrcona |
Cannot connect to server public.localhost: Connection refused |
| 13:33 |
Dyrcona |
Hrm.. This is harder to fix than usual... |
| 13:34 |
Dyrcona |
Now, I've totally busted it. |
| 13:36 |
Dyrcona |
This isn't good. It can't load mod_legacy_auth. |
| 13:54 |
Dyrcona |
OK. Fixed it. Turns out mod_legacy_auth was not my issue. I missed one of the registrations of users, router wasn't registered with private.localhost. |
| 13:56 |
Dyrcona |
Well, no, that turns out not to be it, either.... So, I'm not sure what fixed it, now, but after trying to register the users again, it worked. |
| 14:03 |
|
khuckins joined #evergreen |
| 14:03 |
|
sandbergja joined #evergreen |
| 14:07 |
csharp |
I do know that upgrading 16.04 to 18.04 in-place requires an ejabberd hostname change dance, but that's alerted when you run the release upgrade |
| 14:08 |
Dyrcona |
csharp: Yeah, I'm not sure what happened here. I think I may have not completely configured this vm, or used a wrong password or something. |
| 14:08 |
Dyrcona |
The vm wasn't upgraded. |
| 14:09 |
Dyrcona |
I'm not getting the DataCloneError on FF 70, either. |
| 14:10 |
Dyrcona |
Also, I just noticed it is snowing here again. |
| 14:18 |
|
dbs_windows joined #evergreen |
| 14:19 |
* dbs_windows |
spends an hour installing JRE 1.8 + Hatch and trying to figure out why printing to a Dymo 330 is failing with margin errors when he finds https://bugs.launchpad.net/evergreen/+bug/1817932 |
| 14:19 |
pinesol |
Launchpad bug 1830391 in Evergreen "duplicate for #1817932 Hatch omnibus circa 3.3 (Java updates and more)" [Undecided,New] |
| 14:20 |
rfrasur |
Dyrcona++ #snow_noticing |
| 14:25 |
dbs_windows |
goodbye, java 8 jre and hello hatch-0.3.0 (and annoying Windows SmartScreen) |
| 14:35 |
dbs_windows |
Yay! Dymo 330 works with hatch-0.3.0. Would be nice to update evergreen-ils.org/downloads accordingly |
| 14:35 |
rfrasur |
dbs++ |
| 14:36 |
rfrasur |
dbs_windows++ |
| 14:37 |
dbs_windows |
berick++ |
| 14:37 |
rfrasur |
berick++ |
| 14:39 |
jeff |
also potentially worth noting: at present, there is no "modern" Hatch support for 32-bit Windows, mostly because there is no reasonably available (at last I checked) 32-bit OpenJDK for Windows |
| 15:04 |
dbs_windows |
hmm, font family/size/weight don't seem to get pulled in from exported templates. That will be a bug to chase down for the future I guess... |
| 15:06 |
|
dbwells joined #evergreen |
| 15:25 |
|
jvwoolf joined #evergreen |
| 15:46 |
Bmagic |
https://imgur.com/a/Ip4r2vy |
| 15:46 |
Bmagic |
Note to self: don't loop 24k files and push an operation into a sub process using & between "for f in $FILES" |
| 15:47 |
Bmagic |
Never seen top report 6k. LOL |
| 15:48 |
mmorgan |
Wow! |
| 15:49 |
Dyrcona |
Bmagic: What you did is called a "fork bomb." It's one variation of it. |
| 15:49 |
Bmagic |
yep! That's what devted was telling me a whole bunch |
| 15:52 |
Dyrcona |
Don't background the processes next time. :) |
| 15:52 |
Bmagic |
I am playing with getting it to be more than single threaded |
| 15:52 |
Bmagic |
I think I've got a winner now |
| 15:52 |
Dyrcona |
Bmagic: I have Perl program and C++ program for doing that sort of thing without threads..... |
| 15:55 |
Dyrcona |
If you want to use threads, I recommend Java or Python. |
| 15:55 |
Bmagic |
:) |
| 15:56 |
Dyrcona |
Here's the Perl program: https://gist.github.com/Dyrcona/7488180 |
| 15:57 |
Dyrcona |
You basically feed it a file with a list of command lines to run, and it executes so many at a time until it is done. |
| 15:58 |
phasefx |
not the same thing, but I'll mention it. I like GNU parallel |
| 16:01 |
Dyrcona |
Yeah, there are a number of tools out there for doing this sort of thing. parallel is more flexible than my simple Perl program. |
| 16:01 |
|
sandbergja_ joined #evergreen |
| 16:02 |
Bmagic |
Dyrcona++ |
| 16:07 |
mmorgan |
Since 3.1, we've noticed our popularity badges seem to have little effect when sorting catalog searches by popularity adjusted relevance. Has anyone else noticed the same thing? |
| 16:10 |
mmorgan |
a 'harry potter' search sorted by relevance shows us results where the actual novels are pretty far down in the list. |
| 16:10 |
|
jvwoolf1 joined #evergreen |
| 16:11 |
mmorgan |
Sorting by poprel used to pop them up to the top, but since the changes in 3.1, poprel doesn't seem to do much at all. |
| 16:13 |
mmorgan |
With all the changes that came in with bug 1744385, we noticed a factor of 1000 that wasn't there in earlier releases, and wonder if that is the cause: |
| 16:13 |
pinesol |
Launchpad bug 1744385 in Evergreen "Wishlist: Search & display infrastructure improvements" [Wishlist,Fix released] https://launchpad.net/bugs/1744385 |
| 16:13 |
mmorgan |
https://git.evergreen-ils.org/?p=Evergreen.git;a=blob;f=Open-ILS/src/perlmods/lib/OpenILS/Application/Storage/Driver/Pg/QueryParser.pm;h=950a081b63a30b9c0889dbc3ef27fff51150abe7;hb=HEAD#l1253 |
| 16:13 |
mmorgan |
Anyone know why that factor of 1000 was introduced? |
| 16:16 |
rhamby |
I have a flippant desire to say because 999 just didn't feel like quite enough but that might because it's Friday afternoon. :) |
| 16:16 |
|
stephengwills left #evergreen |
| 16:17 |
* mmorgan |
realizes that Friday afternoon might not be the best time for such a question :) |
| 16:18 |
mmorgan |
On a test system, I removed the factor of 1000, and our popular items zoomed right back up to the top of the poprel search. |
| 16:23 |
mmorgan |
Anyone else have issues with poprel, or is it just us? |
| 16:27 |
* mmorgan |
will open an lp bug, but wanted to look for insight here first. |
| 17:06 |
|
mmorgan left #evergreen |
| 18:02 |
pinesol |
News from qatests: Testing Success <http://testing.evergreen-ils.org/~live> |
| 18:03 |
|
sandbergja_ joined #evergreen |
| 18:33 |
|
dbwells_ joined #evergreen |
| 19:34 |
|
remingtron_ joined #evergreen |
| 19:39 |
|
sandbergja_ joined #evergreen |
