| Time |
Nick |
Message |
| 06:01 |
pinesol |
News from qatests: Testing Success <http://testing.evergreen-ils.org/~live> |
| 07:25 |
|
rjackson_isl_hom joined #evergreen |
| 07:52 |
JBoyer |
@band add CONNECTED TO THE NETWORK!!! |
| 07:52 |
pinesol |
JBoyer: Band 'CONNECTED TO THE NETWORK!!!' added to list |
| 08:30 |
|
mmorgan joined #evergreen |
| 08:33 |
|
mantis joined #evergreen |
| 08:34 |
csharp_ |
JBoyer++ |
| 08:34 |
* csharp_ |
imagines raging house music beats |
| 08:50 |
|
Dyrcona joined #evergreen |
| 09:18 |
|
rfrasur joined #evergreen |
| 09:26 |
|
jvwoolf joined #evergreen |
| 10:02 |
|
rfrasur joined #evergreen |
| 10:03 |
jeff |
we have a group of libraries where -- thanks to duplicate item barcodes and "legacy" barcodes that require transformation -- barcode completion expands to 12 possible values |
| 10:04 |
jeff |
hooray for ANY() |
| 10:21 |
* Dyrcona |
has been using ANY(?) a good bit with Perl DBI programs, you can jam a Perl arrayref into that parameter slot and it just works. |
| 10:32 |
jeff |
evergreen.get_barcodes already uses ANY(), and I don't recall offhand if it's a huge performance boost here (text arrays, not int arrays), but *shrug* |
| 10:33 |
jeff |
I now have a table where I can put regex-based barcode transformations. |
| 10:33 |
jeff |
(which are only used in get_barcodes style autocompletion contexts... doesn't do any kind of general input normalization, etc) |
| 10:33 |
jeff |
that would be a nice next step. |
| 10:35 |
Dyrcona |
I have a TODO in my latest script to normalize a date input parameter, but meh.... I'm the only one who will ever use the option, right? :) |
| 10:36 |
Dyrcona |
Just added it this morning, in fact. |
| 10:37 |
Dyrcona |
Also, speaking of regexes, gotta love code where: if (this_option) $query = ~ s/some comparison/some other comparison/; |
| 10:52 |
Dyrcona |
Looks like the modified queries work so long as the date time is properly formatted. |
| 11:24 |
|
jihpringle joined #evergreen |
| 12:30 |
csharp_ |
*sigh* - while the ejabberd shaper settings seemed to help a bit, still seeing open-ils.actor die off without warning and without clues as to why |
| 12:45 |
Dyrcona |
csharp_: Happens to me with storage. Do you have log messages about it trying to restart? |
| 12:46 |
csharp_ |
Dyrcona: what do the logs say so I can grep? |
| 12:47 |
Dyrcona |
I see you mentioned it yesterday morning. Mine looks different: http://irc.evergreen-ils.org/evergreen/2021-11-02#i_494286 |
| 12:48 |
Dyrcona |
I see this for the restart attempt: "server: restarting after fatal crash..." |
| 12:49 |
Dyrcona |
After that, the listener is running, but it can't start drones. |
| 13:08 |
Dyrcona |
csharp_ ^^ |
| 13:13 |
Dyrcona |
Meh. Searching Lp for a bug that I'm "sure" exists, and can't find it, not even with fix released checked. |
| 13:14 |
Dyrcona |
Does anyone remember a bug about the pre-cat item call number being editable? |
| 13:14 |
jeff |
yes. |
| 13:14 |
Dyrcona |
jeff: You wouldn't happen to have the Lp number handy, would you? |
| 13:15 |
jeffdavis |
bug 1849334 |
| 13:15 |
pinesol |
Launchpad bug 1849334 in Evergreen "Precat owning library should not be editable" [Medium,Confirmed] https://launchpad.net/bugs/1849334 |
| 13:15 |
csharp_ |
pinesol: ahem |
| 13:15 |
pinesol |
csharp_: have you tried local mean solar time for the named city as the reference point? |
| 13:16 |
jeff |
in addition to that, there have been recent bugs regarding precat bib merges: |
| 13:16 |
jeff |
bug 792595 |
| 13:16 |
pinesol |
Launchpad bug 792595 in Evergreen "it's possible to merge the precat dummy bib" [Medium,Fix released] https://launchpad.net/bugs/792595 |
| 13:16 |
jeff |
bug 827356 |
| 13:16 |
pinesol |
Launchpad bug 827356 in Evergreen "DB Doesn't protect against precat merges" [Medium,Fix released] https://launchpad.net/bugs/827356 |
| 13:16 |
jeff |
(one of those is more recent than the other) :-) |
| 13:17 |
jeff |
that last one is the one I had in mind, but realized after I found it that it was about merging. |
| 13:17 |
Dyrcona |
jeff++ jeffdavis++ |
| 13:18 |
jeff |
oh, but the bug title is no longer accurate, as in the comments it looks like the scope expanded a bit. |
| 13:19 |
jeff |
one of the things the patch claims to do is: "adds rules to biblio.record_entry id -1, asset.copy_location id 1, asset.call_number id -1 so that edits are rejected for those ids on those tables" |
| 13:33 |
|
khuckins joined #evergreen |
| 14:48 |
Dyrcona |
Number of git branches always feels like a trade off with no real upside: At times, it seems I have too many so I consolidate branches and commits, and at another time I wish my branches and commits were more split up. |
| 14:48 |
Dyrcona |
I wish I had more branches at the moment. |
| 14:54 |
mmorgan |
Dyrcona: I have a few branches that fell in my yard you can have if you want ;-) |
| 15:14 |
Dyrcona |
mmorgan: Thanks, I have more than enough of those branches. :) |
| 15:15 |
* mmorgan |
figured as much :) |
| 15:56 |
|
jlundgren joined #evergreen |
| 16:08 |
* mmorgan |
is still struggling with group_penalty_thresholds |
| 16:09 |
mmorgan |
I have a grp_penalty_threshold at the consortium level for PATRON_EXCEEDS_LOST_COUNT with threshold 1 |
| 16:09 |
mmorgan |
I also have one set for a specific group at a specific library with threshold 99 |
| 16:09 |
mmorgan |
The idea is that if they go into that specific library, they shouldn't be blocked, but they should be at other libraries. |
| 16:11 |
mmorgan |
To confuse things further, these are new thresholds, so existing patrons with lost items will need the appropriate blocks. |
| 16:13 |
mmorgan |
I'm using actor.calculate_system_penalties to apply the blocks retroactively. Do I just need to be judicious about what org units I apply them for for these specific groups? |
| 16:14 |
mmorgan |
Or is in not possible to have a consortium level block that is stricter than a local block? |
| 16:15 |
mmorgan |
in/it |
| 16:17 |
JBoyer |
mmorgan, I believe I've seen in the past that the most restrictive penalty that can be applied will be. There isn't a concept of a more-specific override the way there is with OUS and the like. |
| 16:19 |
JBoyer |
So you could swap things around if that was what was wanted, but you can't do things where the whole consortium uses value X and this branch uses X+1. (Short of duplicating the penalty at every level where it could change and leaving it undefined at the cons level) |
| 16:23 |
mmorgan |
JBoyer: Ok, that's disappointing. So it sounds like the number of rows in my grp_penalty_threshold table might need to expand exponentially :-( |
| 16:27 |
Dyrcona |
Well, this isn't right, either: Exception: OpenSRF::EX::Session 2021-11-03T16:26:12 main /openils/bin/action_trigger_runner.pl:247 Session Error: Transport::handler(): No AppSession object returned from server_build() |
| 16:27 |
Dyrcona |
Looks like the password reset email trigger. |
| 16:28 |
Dyrcona |
Yay. The server is using almost all of its RAM. |
| 16:29 |
Dyrcona |
It's a report. |
| 16:33 |
Dyrcona |
Can I just kill the reporter that's running the report? Will that cancel the backend and clean up any temp space possibly used by it? (I suppose I've killed the reporter before.) |
| 16:48 |
|
jihpringle joined #evergreen |
| 17:05 |
|
mmorgan left #evergreen |
| 17:27 |
jeffdavis |
According to my notes, killing the relevant report process will usually but not always cancel the db query. I think our usual process is (1) stop reporter, (2) cancel PG backend manually if necessary, (3) set complete_time = now() with error_code and error_text for the relevant report in reporter.schedule, (4) start reporter |
| 17:56 |
|
jihpringle joined #evergreen |
| 18:00 |
pinesol |
News from qatests: Testing Success <http://testing.evergreen-ils.org/~live> |
| 19:43 |
|
jihpringle joined #evergreen |
