| Time |
Nick |
Message |
| 06:54 |
|
collum joined #evergreen |
| 08:31 |
|
mmorgan joined #evergreen |
| 09:29 |
|
Dyrcona joined #evergreen |
| 11:07 |
|
Christineb joined #evergreen |
| 11:09 |
Bmagic |
has anyone noticed that Evergreen allows patrons to go beyond their hold limit when placing holds at night when the library is closed? |
| 11:20 |
csharp_ |
:-/ |
| 11:21 |
csharp_ |
Bmagic: we have not seen that before |
| 11:22 |
csharp_ |
might mess around with the action.hold_request_permit_test function to see which policies are getting picked at which times |
| 11:22 |
Bmagic |
There must be a magic set of YAOUS that makes it happen. I could see it where a title level hold is placed at night, and another library has a copy of it and the hold rules allow the patron to put a hold on another library's copy, therefore the hold rule that is used wouldn't be the one with the limit? |
| 11:23 |
csharp_ |
maybe something with hours of operation? |
| 11:23 |
Bmagic |
it's hours of operation for sure |
| 11:23 |
csharp_ |
are all the libraries set to be closed at the off hours? |
| 11:23 |
Dyrcona |
I don't think hours of operation would matter. |
| 11:24 |
Dyrcona |
If they do, that's a bug. |
| 11:25 |
Dyrcona |
Refresh my memory. How are hold limits determined? |
| 11:25 |
Bmagic |
what if all of the local copies are unavailable but another library's copy is available, the selected hold rule could be not a local rule? And instead it's using the remote library hold rule (which doesn't have a limit) |
| 11:25 |
csharp_ |
Dyrcona: hold matchpoints |
| 11:25 |
Dyrcona |
Well, there you go, if a matchpoint with a different limit applied, then the hold is allowed. |
| 11:26 |
Bmagic |
I'm thinking the only way it's possible is when the hold rule that get's figured out isn't the one with the limit |
| 11:26 |
csharp_ |
pickup_ou integer, request_ou integer, match_item bigint, match_user integer, match_requestor integer, retargetting boolean - those are the args for the function that matches the matchpoints |
| 11:27 |
Dyrcona |
Do hold matrix matchpoints cascade the way that circ rules do? |
| 11:27 |
csharp_ |
Dyrcona: alas, no |
| 11:27 |
Dyrcona |
csharp_++ I was about to say that I think they don't. |
| 11:28 |
csharp_ |
we're looking at a situation where we might have to add 200+ more policies to address a quirk - if they had fallthrough like circ, it would probably be one rule |
| 11:28 |
Dyrcona |
So, it library 1 allows 10 holds and library b allows 20, and library b's matchpoint is chosen the patron from library A gets 20 holds. |
| 11:28 |
csharp_ |
also if the circ policies could say things like ou1 <> ou2 |
| 11:29 |
csharp_ |
(and hold policies) |
| 11:29 |
Dyrcona |
csharp_: You could do "ou1 <> ou2" back when the rules were handled in JavaScript, but that was slightly before my time. :) |
| 11:32 |
csharp_ |
Dyrcona: I was around for the JS circ days - they were unwieldy but were quite flexible! |
| 12:00 |
jeff |
random? sure! moon phase? you got it! last two characters of the md5sum of the item barcode drives daily fines? sure! |
| 12:41 |
|
mantis1 joined #evergreen |
| 12:42 |
mantis1 |
just for clarification - when updating someone's permission group within the patron edit screen, the expiration date also updates, too, correct? |
| 12:42 |
mantis1 |
wasn't sure if this is a library setting or just baked into the system as a whole |
| 12:51 |
jeffdavis |
oh wow, js circ policies, there's a blast from the past |
| 12:55 |
jeffdavis |
mantis1: yes, I'm seeing that behavior on my system; not sure if it's governed by a setting but I would guess not (I don't immediately see a relevant library setting) |
| 12:55 |
mantis1 |
I was looking around library settings and couldn't catch any |
| 12:57 |
jeffdavis |
yeah looks like no settings checked in regctl.js either, it just applies whatever the perm_interval is from the selected profile |
| 13:01 |
mantis1 |
jeffdavis++ thanks for checking |
| 13:21 |
|
mantis2 joined #evergreen |
| 13:38 |
jeff |
yep, changing the group sets the expiration date to now + the group's perm_interval. has for a while now, as long as I can remember. |
| 13:38 |
jeff |
(not that you needed additional confirmation) |
| 13:59 |
|
mantis2 left #evergreen |
| 14:36 |
csharp_ |
goood++ # bug 2133878 |
| 14:36 |
pinesol |
Launchpad bug 2133878 in Evergreen "Store user session data in the database rather than in memcached" [Wishlist,In progress] https://launchpad.net/bugs/2133878 - Assigned to Mike Rylander (mrylander) |
| 14:40 |
goood |
csharp_: if you are just bored over the break and want to test that out ... that'd be super awesome |
| 14:40 |
goood |
oh, I need to remove me from the bug assignment! |
| 14:41 |
goood |
done |
| 14:49 |
goood |
csharp_ / Dyrcona: re rule explosion, I have a proposal (have had for ... 12 years) to address that in the form of field comparison rather than value comparison. "this rule is a match if the copy circ_lib and user home_ou are the same (range adjusted)" instead of "if circ_lib = 3 and home_ou = 3" |
| 14:50 |
csharp_ |
goood: where do I +1? :-) |
| 14:50 |
goood |
but ... no funding has come through. just considered a nice-to-have, I guess, once the static rules are written |
| 14:50 |
goood |
csharp_: well, you just sign this contract right here, and ... ;) |
| 14:51 |
csharp_ |
if only... |
| 14:54 |
goood |
jeffdavis: the /really/ fun thing about the js policies is that I implemented a whole DOM model (for working with MARCXML) /and/ a set of EG integration functions (written in perl) that were injected into the standalone spidermonkey engine we used. that was actually really cool. I miss it, TBH. |
| 14:58 |
goood |
nowadays I'd probably just teach Node to speak OpenSRF. it has sockets, right? ... oh! hello, node-redis! ;) |
| 15:01 |
* jeffdavis |
makes apotropaic gestures to ward off demons |
| 15:02 |
csharp_ |
jeffdavis++ # teaching us new words :-) |
| 15:02 |
jeffdavis |
the kids would just say that node-redis is "cursed" |
| 15:03 |
csharp_ |
@blame node-redis |
| 15:03 |
pinesol |
csharp_: It really IS node-redis's fault! |
| 15:27 |
csharp_ |
goood: I'm going to be off for the next two weeks, but I installed the in-db-session branch on csharp-main.gapines.org (admin/demo123) |
| 15:28 |
csharp_ |
works so far - able to register workstation, move around, and I see the session in the new table |
| 15:28 |
goood |
woot. it's been working on my dev machine for a month, so ... I'm not TOO surprised, but glad for the basic confirmation! |
| 15:29 |
csharp_ |
noice |
| 15:33 |
goood |
it shifts load from memcache to pg, of course, so some real stress testing is in order. (do I want to double-buffer? use pg as L2 and memcache as L1? no I do not. see: the 2 hard things in compsci) |
| 17:05 |
|
mmorgan left #evergreen |
