| Time |
Nick |
Message |
| 02:39 |
|
dbs joined #evergreen |
| 06:59 |
|
agoben joined #evergreen |
| 07:13 |
|
rjackson_isl joined #evergreen |
| 07:26 |
|
rfrasur joined #evergreen |
| 07:26 |
|
Dyrcona joined #evergreen |
| 08:46 |
|
mmorgan joined #evergreen |
| 08:48 |
|
jvwoolf joined #evergreen |
| 08:53 |
|
sandbergja joined #evergreen |
| 09:18 |
|
tlittle joined #evergreen |
| 09:19 |
csharp |
any major sites running PG 10+? I'm upgrading our next-release test servers to Ubuntu 18.04 which comes with 10 and I'm trying out PG11 |
| 09:31 |
|
yboston joined #evergreen |
| 09:49 |
|
Stompro joined #evergreen |
| 10:14 |
* Dyrcona |
shakes his fist at slow websites and networks everywhere. |
| 10:14 |
|
collum joined #evergreen |
| 10:15 |
alynn26 |
Slow websites -- |
| 10:34 |
bshum |
csharp: Be curious what you find, we keep working on it, but not enough data to know :) |
| 10:35 |
Dyrcona |
csharp: We've not looked at PG11, yet. I've meant to start, but other issues have come up. |
| 10:35 |
Dyrcona |
As for Pg10, I don't think anyone has tried it in any serious way. |
| 10:35 |
bshum |
That's what I think too |
| 10:36 |
bshum |
Getting the community to standardize on PG9.6 was the latest official step |
| 10:36 |
bshum |
For the next release |
| 10:36 |
Dyrcona |
bshum and I have rune the tests, and JBoyer found a bug with Vandelay that tests don't cover. |
| 10:36 |
bshum |
And we fixed that |
| 10:36 |
bshum |
At least I think it went into master |
| 10:36 |
Dyrcona |
Yeah, we did fix it. |
| 10:36 |
Dyrcona |
It was committed. |
| 10:37 |
Dyrcona |
csharp: God speed and good luck! :) |
| 10:37 |
bshum |
Huzzah! |
| 10:39 |
Bmagic |
A quick double check: when AT templates result in nothing in the "To:" address, will it attempt to pass it to the mailer? And result in an error at the Sendmail level? |
| 10:42 |
Bmagic |
I don't see anything in my mail logs with to=<> |
| 10:43 |
Dyrcona |
Bmagic: Yes. See bug 1837454 |
| 10:43 |
pinesol |
Launchpad bug 1837454 in Evergreen "SendEmail Reactor Will Try to Send Email with no valid To, Cc, or Bcc Headers" [Undecided,New] https://launchpad.net/bugs/1837454 |
| 10:44 |
Bmagic |
Dyrcona++ |
| 10:45 |
Bmagic |
I wonder if there is a repercussion when using a relay service like sendgrid? Would it debit the allowed "bounces" |
| 10:46 |
Dyrcona |
Dunno. We have exim set up to relay via Google. That error in the bug description is from exim, so the mail never goes anywhere in our case. |
| 10:47 |
Bmagic |
I guess the question is: would sendmail even attempt to deliver the envelope to the relay (if sendmail is used on the AT server) |
| 10:49 |
|
Christineb joined #evergreen |
| 11:00 |
Dyrcona |
Bmagic: I don't use sendmail, so I don't know, but I doubt it. |
| 11:01 |
Dyrcona |
Bmagic: You could search your logs for something like "No recipient addresses found in header" |
| 11:01 |
Bmagic |
looking |
| 11:02 |
pinesol |
News from qatests: Testing Success <http://testing.evergreen-ils.org/~live> |
| 11:04 |
Bmagic |
Nothing like that in the logs |
| 11:05 |
Dyrcona |
You searched all of the logs, not just Evergreen logs? |
| 11:05 |
Bmagic |
/var/log/mail.log* |
| 11:06 |
Dyrcona |
There may be different logs for sendmail, but I haven't used it in over 17 years. |
| 11:06 |
Bmagic |
:), no worries, thanks for thinking about it with me |
| 11:06 |
Dyrcona |
That also varies by O/S and distro. |
| 11:06 |
Dyrcona |
I really doubt that sendmail is relaying mail with no recipient, though. |
| 11:07 |
Bmagic |
I am thinking that's the case based on the log "tea leaves" |
| 11:09 |
berick |
Bmagic: i would expect Email::Send would crap out on the message before relaying. |
| 11:09 |
Bmagic |
berick: yep, that's exactly what I was looking for |
| 11:09 |
berick |
which would result in evergreen logs like "SendEmail Reactor: unable to send ..." |
| 11:10 |
Bmagic |
SendEmail Reactor: unable to send email: $VAR1 = bless( {'prop' => {},'string' => 'Can\'t send data','errno' => 1,'type' => 'failure'}, 'Return::Value' ); |
| 11:10 |
Bmagic |
SendEmail Reactor: failed email template: To: From: |
| 11:10 |
berick |
yeah |
| 11:10 |
Bmagic |
yep, there's the smoking gun |
| 11:12 |
Bmagic |
I was dreaming of altering the template to intentionally blank the To: address under some IF condition and wondering if that would be a bad thing. I think it would be "ok" |
| 11:13 |
Dyrcona |
I find a few of those messages from today. |
| 11:14 |
Dyrcona |
Well... The template shouldn't be generated if there's no recipient email address. |
| 11:14 |
Dyrcona |
Hence, the chained validators bit. |
| 11:15 |
Bmagic |
there should be a way to short circuit the whole template to make a completely blank result (no headers, no nothing) at the template generation level. Of course, a better way would be to add validators into Evergreen as you suggest |
| 11:15 |
mmorgan |
Bmagic: Under what circumstances would you want to blank the To: address? What are you trying to accomplish? |
| 11:16 |
Bmagic |
the library would like certain patron groups to be removed from reactor SendEmail, SendSMS |
| 11:17 |
Dyrcona |
Our email setup is pretty crufty, and I don't think it even works on Perl versions higher than 5.22. There are a number of email bugs assigned to me that are so far going nowhere. |
| 11:18 |
Dyrcona |
Bmagic: You can do that with a filter on the events that run those reactors, I think. I'm not sure if filters work on events fired by a previous event, i.e AutorenewNotify. |
| 11:18 |
mmorgan |
Bmagic: What about a custom a_t_filter to filter out those patron groups.... |
| 11:18 |
Bmagic |
opt-in comes to mind as well |
| 11:18 |
mmorgan |
What Dyrcona said:) |
| 11:19 |
Dyrcona |
Thing is, you might want the main event to trigger but not the notification event, and I've not looked at how filters interact with those. |
| 11:20 |
Bmagic |
the filter sounds attractive |
| 11:21 |
Dyrcona |
Well, all right, there is 1 email bug assigned to me, that I think I'm going to drop for now. I also need to take a look at the 9 other bugs and get some testing scheduled, but other things have been a higher priority.... |
| 11:29 |
jeff |
I may pick that up, if it's the one I think it is. |
| 11:31 |
Dyrcona |
jeff: https://bugs.launchpad.net/evergreen/+bug/1466502 |
| 11:31 |
pinesol |
Launchpad bug 1466502 in Evergreen "Replace Email::Send with Email::Sender" [Wishlist,Confirmed] |
| 11:32 |
Dyrcona |
I was looking at this one, but dropped it some time ago: https://bugs.launchpad.net/evergreen/+bug/1801163 |
| 11:32 |
pinesol |
Launchpad bug 1801163 in Evergreen "SendEmail A/T reactor broken for recent version of Encode::MIME::Header" [High,Confirmed] - Assigned to Jeff Godin (jgodin) |
| 11:32 |
Dyrcona |
jeff: I see you grabbed the latter. |
| 11:33 |
* Dyrcona |
runs out to get some lunch. |
| 11:49 |
Stompro |
Bmagic, we pre-process action_trigger pending runs with a sql script to do things like that, like removing certain permission_groups, and just mark those events as invalid. |
| 11:49 |
Bmagic |
Stompro: interesting |
| 11:50 |
Stompro |
Bmagic, it is also how we roll failed bill emails over to print bills, and limit sms & phone calls to one per day per patron. |
| 11:51 |
Bmagic |
very cool |
| 11:51 |
Bmagic |
got a branch on that? |
| 11:52 |
Stompro |
No, but I can put it in a gist. |
| 11:52 |
mmorgan |
Stompro: That is cool! |
| 11:52 |
Bmagic |
That sounds like a worth-while feature for everyone! |
| 11:53 |
Stompro |
It isn't as maintainable as the built in methods, and requires editing scripts on the server. |
| 11:54 |
|
rfrasur joined #evergreen |
| 11:54 |
Bmagic |
editing scripts? oh, well, in that case nevermind.... haha |
| 11:54 |
mmorgan |
I'd still favor filters wherever possible to prevent certain events from being created in the first place. |
| 11:56 |
|
jvwoolf joined #evergreen |
| 11:57 |
Stompro |
mmorgan, filters cannot specify joins though, so no way to add in criteria like patron email exists as far as I could tell. |
| 12:01 |
|
aabbee joined #evergreen |
| 12:02 |
* mmorgan |
thought someone had shared a way to do that in a filter on irc at one point. Can't locate it though, and never got a chance to try it. |
| 12:04 |
Dyrcona |
mmorgan Stompro: You can specify subqueries. |
| 12:06 |
Dyrcona |
Pretty much all of our custom filters have an '-and' : [ {"-exists" : { "select" : .... |
| 12:06 |
Stompro |
Dyrcona, could you share your custom filters please, I would love to see examples. |
| 12:07 |
berick |
Bmagic: FYI, https://git.evergreen-ils.org/?p=working/Evergreen.git;a=shortlog;h=refs/heads/user/berick/el-integrate-circa-3.4 -- wasn't really ready to start talking about it, but I'm hoping to start testing locally before too long |
| 12:07 |
|
jihpringle joined #evergreen |
| 12:08 |
Bmagic |
berick++ # is there more on this topic that I am not aware of? |
| 12:09 |
berick |
Bmagic: no, so far it's just been a local project. |
| 12:09 |
berick |
borrowing an idea or 2 from jeff's mobile/elastic catalog |
| 12:09 |
Stompro |
Bmagic, here is an example of how we pre-process AT events. https://gist.github.com/stompro/45bc96af05c015769a95d7b5b8e99316 |
| 12:11 |
jeff |
berick++ bookmarking for later |
| 12:11 |
Stompro |
Dyrcona, I would love to update the AT filter docs with info like this, the json sql syntax really makes my brain hurt though. |
| 12:11 |
Bmagic |
berick: nice! Glad I didn't get too far into it myself! |
| 12:11 |
Bmagic |
Stompro++ |
| 12:13 |
Dyrcona |
Stompro: Here's a simple example for our member who sends lost notices after 14 days: https://pastebin.com/CECvqLqx |
| 12:14 |
mmorgan |
Dyrcona++ |
| 12:14 |
Dyrcona |
Limits to certain circ modifiers via the subquery. |
| 12:14 |
Bmagic |
Stompro: and here is mine https://pastebin.com/ZmRJ7mHp |
| 12:16 |
Dyrcona |
You could use that to limit to patrons with an email, or a SMS setting. |
| 12:16 |
|
jvwoolf1 joined #evergreen |
| 12:17 |
Stompro |
Dyrcona++ Bmagic++ thanks, I think I need to go back and amend some emails to the listserve now. "New information (to me) has come to light" |
| 12:19 |
Bmagic |
berick: wow man, you've been busy! This is great! Mind if I link it to the LP? |
| 12:21 |
berick |
Bmagic: no, i don't mind |
| 12:22 |
Dyrcona |
berick++ |
| 12:23 |
Bmagic |
berick++ # ferreal |
| 12:23 |
sandbergja |
berick++ |
| 12:24 |
|
khaun joined #evergreen |
| 12:24 |
Bmagic |
I'd say it was something close to telepathy that I posted that LP bug today only 7 days after this code started appearing from berick |
| 12:24 |
berick |
Bmagic: oh, no, it was rebased 7 days ago |
| 12:24 |
berick |
it's been in progress for years now |
| 12:24 |
Bmagic |
well I guess I take that back then |
| 12:25 |
berick |
but over last month i've been putting what I think are the finishing touches |
| 12:25 |
Bmagic |
I have to admit: It's been on my mind on and off for years now. |
| 12:25 |
berick |
at least enough to support the staff work flow we need |
| 12:25 |
Bmagic |
and just recently got serious about it |
| 12:28 |
Bmagic |
Now I have to re-plan my freed 6-18 month future. Thanks a lot berick! |
| 12:30 |
berick |
heh, well, let me know if you have questions, etc. eyes and testing obviously appreciated |
| 12:31 |
* berick |
should probably post a tech doc to the repo |
| 12:39 |
|
jvwoolf joined #evergreen |
| 13:18 |
|
nfBurton joined #evergreen |
| 13:21 |
|
khuckins joined #evergreen |
| 13:28 |
* jeff |
eyes a 3.1 db where most of the vr_format values in config.coded_value_map are duplicated once |
| 13:28 |
|
sandbergja joined #evergreen |
| 13:29 |
jeff |
(all but " " Unspecified and "s" Blu-ray disc) |
| 13:36 |
|
rfrasur joined #evergreen |
| 14:04 |
|
bos20k joined #evergreen |
| 14:06 |
|
jvwoolf joined #evergreen |
| 14:29 |
|
khuckins joined #evergreen |
| 14:54 |
jeffdavis |
How do you all handle Postgres upgrades? Do you just do pg_dump and restore to a new cluster, or use pg_upgrade, or...? |
| 14:58 |
|
mmorgan1 joined #evergreen |
| 15:00 |
Dyrcona |
jeffdavis: When I upgraded from Pg 9.2(?) to Pg 9.5 I also got new servers, so I did a dump and restore. I have done some test upgrades using pg_upgrade and that seems to work, too. I'm planning to use pg_upgrade to go from 9.5 to 9.6 (or maybe 10) before we upgrade to Evergreen 3.4 next spring. |
| 15:01 |
Dyrcona |
Having replication makes it more interesting, though you can disable replication, upgrade the software on the replicant, to a pg init db, and start replication over from scratch if anything goes wrong. |
| 15:03 |
gmcharlt |
jeffdavis: we use pg_upgradecluster most of the time |
| 15:04 |
|
jihpringle joined #evergreen |
| 15:16 |
|
bwillis joined #evergreen |
| 15:20 |
bwillis |
I am trying to add some new Z39.50 targets in Evergreen. I am using the The Z39.50 Target Directory compiled by IndexData (http://irspy.indexdata.com/) I have had success connecting to only one of several targets that I have tested. I don't know if the directory information is incorrect or if the targets I selected require authorization. Can |
| 15:20 |
bwillis |
anyone recommend one or more targets that your system is currently using? |
| 15:22 |
jeffdavis |
Dyrcona, gmcharlt: thanks! |
| 15:29 |
|
bwillis53 joined #evergreen |
| 15:33 |
|
bwillis3856 joined #evergreen |
| 15:36 |
dbs |
@later tell bwillis it can be tricky to set up new Z39.50 targets, things like truncation and mapping to fields correctly make a big difference. As does the need for authentication (in some cases) |
| 15:36 |
pinesol |
dbs: The operation succeeded. |
| 15:37 |
|
bjwillis joined #evergreen |
| 15:38 |
dbs |
bjwillis it can be tricky to set up new Z39.50 targets, things like truncation and mapping to fields correctly make a big difference. As does the need for authentication (in some cases) |
| 15:38 |
Stompro |
@later tell bwillis bwillis++ that is a great project. I have had so little luck adding Z39.50 targets that I cannot be any help. We wanted to add a library across the river from us, they were willing, we were willing, but no one could tell us the correct settings and I had no luck figuring it out. |
| 15:38 |
pinesol |
Stompro: The operation succeeded. |
| 15:40 |
dbs |
bjwillis: https://pastebin.com/xa6hA4tJ is an example of one that I think we have working... |
| 15:40 |
dbs |
(sorry for the database format but I'm pressed for time) |
| 15:41 |
bjwillis |
dbs: thank you for that example! |
| 15:42 |
dbs |
I've found it confusing myself, most recently: http://irc.evergreen-ils.org/evergreen/2019-07-12#i_412193 |
| 15:43 |
bjwillis |
I should mention that I have had success adding a few targets, but had hoped to add a few more that might be more relevant to our libraries |
| 15:44 |
dbs |
Ah, sounds like a good wiki-based community project. Or SQL maintained in a "contrib" git repository or something like that! |
| 15:44 |
dbs |
It can be really frustrating when Z39.50 servers advertise UTF8 but actually give you MARC8 and the like. |
| 15:46 |
Dyrcona |
There is this: https://wiki.evergreen-ils.org/doku.php?id=evergreen-admin:sru_and_z39.50 |
| 15:47 |
Dyrcona |
But that's mostly about setting up your Z39.50 server. |
| 15:48 |
bjwillis |
dbs: that IS confusing! |
| 15:49 |
Dyrcona |
There's a dokuwiki page from 2009, but I hesitate to share the link because it looks out of date. |
| 15:49 |
Dyrcona |
bshum and I were working on some instructions for setting up another Evergreen instance some years ago, but I can't find them, now. |
| 15:51 |
bjwillis |
Can someone tell me if the "Format" attribute in the Z39.50 target configuration refers to the Bib-1 Structure Attributes? |
| 15:54 |
jihpringle |
bjwillis: I couldn't figure out what Format is. I think we have it set to 1 for most or all of our sources |
| 15:54 |
phasefx |
I think it's for the fullness of the record |
| 15:56 |
phasefx |
bjwillis: it is the Bib-1 Structure Attributes |
| 15:56 |
phasefx |
raw, it gets sent with @attr 4= |
| 15:59 |
jihpringle |
just as an fyi while we're talking about Z39.50 attributes - https://bugs.launchpad.net/evergreen/+bug/1834687 |
| 15:59 |
pinesol |
Launchpad bug 1834687 in Evergreen 3.3 "Z39.50 Servers Attributes page missing" [Medium,Confirmed] |
| 15:59 |
* phasefx |
may be jumping the gun, 3 different format fields with z39.50 in the source code :) |
| 16:01 |
phasefx |
the one associated with attributes is the structure thingy |
| 16:01 |
jihpringle |
In the search attributes I've seen from Z39.50 sources they have Position, Structure, and Completeness fields that we don't have (or don't have labeled as such in Evergreen) |
| 16:01 |
jeffdavis |
jihpringle: berick pointed us at part of the solution for that bug last week - it is already possible to filter the attributes grid by Z39.50 source via URL param. I have been trying to figure out how to add a link to the Z39.50 Sources UI that links to the appropriately-filtered page. |
| 16:01 |
jeffdavis |
I should update the bug with that information. |
| 16:02 |
jihpringle |
great, thanks jeffdavis |
| 16:11 |
|
yboston joined #evergreen |
| 16:34 |
|
jvwoolf joined #evergreen |
| 16:42 |
|
cmalm joined #evergreen |
| 16:48 |
|
CMalmgren joined #evergreen |
| 16:49 |
|
CMalmgren left #evergreen |
| 16:49 |
|
CMalmgren joined #evergreen |
| 17:03 |
|
mmorgan1 left #evergreen |
| 17:14 |
|
jvwoolf joined #evergreen |
| 17:27 |
|
sandbergja_ joined #evergreen |
| 18:04 |
|
cmalm joined #evergreen |
| 18:18 |
|
khuckins joined #evergreen |
| 21:36 |
|
sandbergja joined #evergreen |
| 22:56 |
|
sandbergja joined #evergreen |
| 23:03 |
pinesol |
News from qatests: Testing Success <http://testing.evergreen-ils.org/~live> |
| 23:28 |
|
jamesrf joined #evergreen |
