| Time |
Nick |
Message |
| 04:51 |
|
zxiiro joined #evergreen |
| 04:56 |
|
zxiiro joined #evergreen |
| 07:12 |
|
timlaptop joined #evergreen |
| 07:53 |
|
collum joined #evergreen |
| 07:54 |
|
rjackson-isl joined #evergreen |
| 08:23 |
|
misilot joined #evergreen |
| 08:33 |
|
Shae joined #evergreen |
| 08:34 |
|
kbeswick joined #evergreen |
| 08:49 |
|
RoganH joined #evergreen |
| 08:54 |
|
mmorgan joined #evergreen |
| 09:52 |
|
yboston joined #evergreen |
| 11:03 |
|
krvmga joined #evergreen |
| 11:03 |
krvmga |
i thought, just for the record, i should post what the solution to my problem the other day was |
| 11:04 |
krvmga |
problem: in advanced search, a numeric search under Call Number (Shelf Browse) for 910.4 brought the results up in the middle of the 910.4 range rather than at the beginning. |
| 11:05 |
bshum |
Yep |
| 11:05 |
pinesol_green |
[evergreen|Bill Erickson] LP#1238015 hide patron credit summary when disabled - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=dcfa1d4> |
| 11:06 |
krvmga |
solution: the call number "J 910.4" had an extra space between the "J" and the rest of the call number. This caused the system to think this was the first record. |
| 11:06 |
bshum |
Ha |
| 11:06 |
bshum |
That's special |
| 11:06 |
krvmga |
as soon as i removed the extra space, the results displayed as expected. |
| 11:06 |
|
misilot left #evergreen |
| 11:07 |
bshum |
krvmga++ |
| 11:07 |
krvmga |
i discovered this by accident. i pulled all the 910.4 data from that library and looked at it in a spread sheet. |
| 11:07 |
csharp |
@marc 092 |
| 11:07 |
pinesol_green |
csharp: unknown tag 092 |
| 11:08 |
krvmga |
when i sorted on call number, the double-spaced call number was at the top of the list. |
| 11:08 |
* berick |
suddenly remembers he's supposed to be re-cutting a release today |
| 11:08 |
dbwells |
csharp: that's a local call number, usually Dewey I think |
| 11:08 |
csharp |
dbwells: isn't that the only way numeric search works? it doesn't search asset.call_number.label, right? |
| 11:09 |
dbwells |
csharp: There are two call number searches which search different data sets. |
| 11:09 |
csharp |
ah - gotcha |
| 11:09 |
csharp |
I never remember which one is which |
| 11:09 |
dbwells |
The shelf browse searches asset.call_number, the "Bib" Call Number search searches what's in the bib record. |
| 11:10 |
csharp |
makes sense - thanks |
| 11:10 |
krvmga |
the double spacing in the call number wasn't visible in the public OPAC or the staff client. in those places, only a single space showed up. |
| 11:11 |
krvmga |
an HTML artifact, i'm guessing |
| 11:11 |
mmorgan |
Does shelf browse search/sort by asset.call_number.label, or asset.call_number.label_sortkey? |
| 11:12 |
senator |
csharp: to see what bib call number searches exactly on your system, (it's not the 092 by default) see: select format, xpath from config.metabib_field where field_class = 'identifier' and name = 'bibcn'; |
| 11:12 |
csharp |
aha |
| 11:12 |
eeevil |
dbwells: thougts on my final comment on our funtime QP bug? (also, did you see Liam's note about sitka testing?) |
| 11:13 |
dbwells |
krvmga: mmorgan: the shelf browse can be tricky. It uses the non-normalized value to find the "pivot" (that is, the closest match to what the user typed in), then uses the sortkey to sort the surrounding results. |
| 11:14 |
csharp |
ours is set to 099 - however, we disable bib cn searching because all of our call numbers are free text and locally set, so we could have many variants on a single bib record |
| 11:14 |
csharp |
and none of those variants gets entered into the 099 |
| 11:14 |
eeevil |
krvmga: html sqaushes all whitespace to one inter-character space marker, regardless of the length of the run of whitespace. (excepting of course) ... that's actually an XML thing, too, but you can control it to a greater degree in xml processing |
| 11:15 |
dbwells |
krvmga: mmorgan: There is a knowledge gap when searching, as we don't know which normalization to apply to the input, since multiple normalizations may be in use. |
| 11:16 |
krvmga |
eevil: that's what i was guessing. |
| 11:16 |
dbwells |
As long as the user enters a call number which actually exists in the system, it works 100%. When the call number doesn't exist, results may vary. |
| 11:17 |
mmorgan |
dbwells: Interesting. That explains some results we sometime see. |
| 11:20 |
|
_bott_ joined #evergreen |
| 11:22 |
dbwells |
mmorgan: We put in a possible fix a few months ago where we basically lookup the call number twice. We find the closest match non-normalized, grab the normalizer from the match, then search a second time with the normalized value. This of course trades performance for convenience and accuracy. There are certainly other solutions, but I should probably open a bug for that, since it works for us. |
| 11:24 |
dbwells |
eeevil: yes, I'll reply on the bug |
| 11:27 |
krvmga |
i just did a 2.5 install on my dev machine and, despite everything being all right in eg_vhost.conf, eg.conf, opensrf.xml, opensrf_core.xml, apache still thinks the DocumentRoot is /etc/apache2/htdocs |
| 11:29 |
eeevil |
dbwells: cool, thanks |
| 11:31 |
|
hopkinsju joined #evergreen |
| 11:32 |
krvmga |
j'en ai ras le bol |
| 11:33 |
bshum |
krvmga: Did you look and see what sites are enabled in sites-enabled? Maybe something is awry there. |
| 11:33 |
krvmga |
bshum: yep |
| 11:33 |
krvmga |
bshum: but thanks for the thought |
| 11:34 |
|
fparks joined #evergreen |
| 11:37 |
krvmga |
well, i'm really not that fed up |
| 11:37 |
* dbs |
trying a fresh master install with no /openils prefix, etc again |
| 11:37 |
krvmga |
just quel frustrate |
| 11:39 |
krvmga |
pardon my frenglish |
| 11:42 |
bshum |
krvmga: What's the dev machine running? Maybe it's something in the way the distro is setup |
| 11:42 |
krvmga |
bshum: hmmm, could be. it's debian-squeeze |
| 11:43 |
|
dMiller joined #evergreen |
| 11:43 |
krvmga |
i'll find out what's up and be back later. |
| 11:46 |
bshum |
@later tell krvmga See http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=569026 which seems to indicate that it could be a vhost naming issue and apache crying home to the defaults. |
| 11:46 |
pinesol_green |
bshum: The operation succeeded. |
| 11:46 |
pinesol_green |
Debian bug 569026 in apache2-mpm-worker "apache2-mpm-worker: Configured DocumentRoot for vhost ignored by compiled variable" [Normal,Open] |
| 11:51 |
* dbs |
still running into a few hard-coded "/openils/" prefixes (action_trigger_runner.pl and offline-blocked-list.pl I'm looking at you) but still need to fine the one that is killing TPAC |
| 11:52 |
dbs |
mmm. OpenILS/Utils/Cronscript.pm - really |
| 11:53 |
dbs |
oh, just hardcoded in pod. hmm |
| 11:55 |
dbs |
"apache2[25696]: [error] Unable to Bootstrap OpenSRF Client with config /openils/conf/opensrf_core.xml.." is pretty damning |
| 12:10 |
|
tater joined #evergreen |
| 12:21 |
dbs |
ohai osrf_http_translator.c and your '#define DEFAULT_TRANSLATOR_CONFIG_FILE "/openils/conf/opensrf_core.xml"' |
| 12:22 |
|
JennB joined #evergreen |
| 12:23 |
* dbs |
adds "OSRFTranslatorConfig /usr/local/etc/opensrf_core.xml" to eg_vhost.conf and expects that problem to go away |
| 12:23 |
JennB |
hello Evergreeners. I have a question about Spell Check in the Staff Client using 2.3.7 & 2.4.2. Does it exists and if so, where is it? Dropdown, automatic? Thanks! |
| 12:28 |
jeff |
I'm not aware of any spell check functionality in the staff client, but what are you looking to spellcheck? |
| 12:29 |
JennB |
Hi Jeff, thats what I am thinking because i have never seen it in S Client either. One of my libraries is asking where it is and thought she used it before. Only the OPAC had a spell checker suggestion. |
| 12:29 |
jeff |
That opac spelling suggestion on search is the only thing I'm aware of. |
| 12:30 |
* berick |
uploads updated 2.3.12 tarball to previews |
| 12:30 |
JennB |
ok! thats for reconfirming what I knew/was aware of |
| 12:30 |
jeff |
Now, there is support for verifying authority controlled fields in the MARC editor -- which isn't strictly "spell check", but can bring to light typos and such. |
| 12:30 |
jeff |
berick++ |
| 12:30 |
jeff |
and dbs++ for testing hardcoded paths |
| 12:31 |
JennB |
OK, thanks! |
| 12:32 |
dbs |
from what I can see, the install instructions will get simpler if we can avoid the /openils prefix |
| 12:33 |
jeff |
ouch: SIPServer.pm: OILS: SIP Checkin request took 6.386 seconds |
| 12:34 |
dbs |
Double-ow. wow. |
| 12:34 |
* berick |
wonders how many hold target attempts were made |
| 12:36 |
jeff |
zero ahcm entries for this copy, so i'm guessing none. |
| 12:39 |
jeff |
SIPServer.pm: OILS: SIP Checkin request took 7.813 seconds |
| 12:39 |
jeff |
bummer. |
| 12:43 |
|
Bmagic joined #evergreen |
| 12:44 |
jeff |
two yesterday, two today. |
| 12:50 |
|
mjingle joined #evergreen |
| 12:55 |
* dbs |
opened bug 1255184 accordingly |
| 12:55 |
pinesol_green |
Launchpad bug 1255184 in Evergreen "Reduce hardcoded paths in Evergreen code" (affected: 1, heat: 6) [Undecided,New] https://launchpad.net/bugs/1255184 |
| 12:59 |
csharp |
dbs++ # easing the path to debs |
| 12:59 |
csharp |
(no pun intended) |
| 13:18 |
jeff |
logs++ |
| 13:25 |
jeff |
hrm. last monday had some up to 11 seconds, but i've no corresponding complaint from a sip client reconnecting. |
| 13:25 |
jeff |
but logs (logs++) should be able to tell me if it did need to reconnect. |
| 13:29 |
jeffdavis |
dbs++ # "grammatically-challenged branch" |
| 13:46 |
pinesol_green |
[evergreen|Mike Rylander] Fix boolean lists; Better atom regex; Caching - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=61302b7> |
| 14:06 |
|
misilot joined #evergreen |
| 14:12 |
|
misilot left #evergreen |
| 14:38 |
berick |
well, whaddya know, this actually works in Chrome/FF: Date.parse('2016-11-26T14:08:49-0500') |
| 14:43 |
dbs |
All hail the evolution of ecmascript! |
| 14:50 |
|
stevenyvr joined #evergreen |
| 15:21 |
jeff |
interesting. circ block doesn't appear to block on a pre-cat. |
| 15:28 |
* eeevil |
taps fingers waiting for i18n build to complete for 2.4.4 wrapping |
| 15:30 |
jeff |
eeevil: in testing, a self checkout terminal from vendor 3 doesn't seem to care about the "renewal privileges denied" flag. if charge privs are denied, the patron can't use the terminal. |
| 15:32 |
eeevil |
jeff: so even if we did the "right" thing with the renew flag and block check, their thing would not act correctly... :( |
| 15:32 |
jeff |
eeevil: still worth adjusting our response there to be correct (now that we can), but i suspect i'm also going to add a config option to the sip server to skip over certain standing penalties. |
| 15:33 |
|
dMiller_ joined #evergreen |
| 15:35 |
jeff |
eeevil: right. i forced OpenILS::SIP::Patron to always return ok for renewal privs, and the terminal still rejected the patron before giving them a chance to renew things. |
| 15:36 |
|
janes joined #evergreen |
| 15:37 |
jeff |
eeevil: i don't think it would be good to "always ignore this circ-blocking penalty" in a hardcoded sense, since i can only test things on certain clients, but i suspect it'll be a "ignore this block for this penalty type" config option with a suggested default. |
| 15:40 |
jeff |
"ignore these penalties completely" might be simpler, at the expense of some expression/flexibility |
| 15:41 |
eeevil |
jeff: wonder if it's time to provide a config for each flag allowing "force-Y/force-N/what-ILS-says" |
| 15:41 |
eeevil |
in the sip config file |
| 15:41 |
eeevil |
that would allow per-client-device tuning |
| 15:42 |
jeff |
yes. i'd still like to do per-client config for some things, not all related to this. |
| 15:43 |
jeff |
though maybe per-sip-user isn't needed if we can do it now with multiple institutions without too much overhead. not sure i'd thought of that yet. |
| 15:44 |
eeevil |
<sip-flags><circ>auto</circ><renew>Y</renew></sip-flags> ... default for unconfigured flags being "auto" |
| 15:45 |
jeff |
off to test how well the client reacts to "sure this patron can check stuff out" followed by not allowing anything to be checked out. |
| 15:45 |
* jeff |
grabs some rfid tags |
| 15:46 |
sandbergja |
Acquisitions question for everyone: on eg 2.4.0, is there a trick to placing holds via Acquisitions > Patron Requests? I have double checked that "Place Hold" was checked, and I have gone through the whole Add to selection list, Activate PO, Receive PO process. |
| 16:03 |
|
dMiller joined #evergreen |
| 16:11 |
|
dMiller_ joined #evergreen |
| 16:32 |
Bmagic |
would anyone out there be able to confirm the arguments required for 'open-ils.actor.user.penalties.update'. I am attempting to call it like this: $session->request('open-ils.actor.user.penalties.update', '', $authtoken, $userID ); The source code looks like imy($self, $conn, $auth, $user_id) = @_; which makes me wonder about $conn |
| 16:36 |
jeff |
$self and $conn are not something you need to worry about. |
| 16:37 |
jeff |
they're used internally -- that's a bit of an over-simplification, but accurate enough. |
| 16:38 |
Bmagic |
I was starting to get that idea as I was looking at some of the other functions - I guess that opensrf fills those in when it makes the call |
| 16:39 |
eeevil |
Bmagic: here's some old docs: http://wiki.evergreen-ils.org/doku.php?id=osrf-devel:user_s_guide |
| 16:39 |
|
hopkinsju joined #evergreen |
| 16:40 |
eeevil |
in particular http://wiki.evergreen-ils.org/doku.php?id=osrf-devel:terms and http://wiki.evergreen-ils.org/doku.php?id=osrf-devel:example-server |
| 16:41 |
eeevil |
opensrf does a ton of magic stuff between client and implementing code so that you don't have to understand or know about endpoints, scaling, etc |
| 16:42 |
Bmagic |
eeevil and Jeff - thanks for the help! I had already expieremented with sending only the auth and userid with error results, I am expecting the opensrf call to create rows in actor.usr_standing_penalty but it doesnt |
| 16:45 |
jeff |
when you call ->request on a session object, you're expected to pass the method name and any parameters. in your example, i'd remove the blank argument after the method name. |
| 16:45 |
jeff |
(sorry, phone delayed me) |
| 16:46 |
jeff |
Bmagic: if you have your code available somewhere or care to put it in a pastebin or gist, we might be able to offer suggestions. no guarantees. :-) |
| 16:46 |
|
Bmagic|2 joined #evergreen |
| 16:48 |
Bmagic|2 |
jeff: Sure, I have the code available |
| 16:51 |
Bmagic|2 |
my last commit to https://github.com/mcoia/eg-scripts should do it - Take a look at recalc_penalties_direct.pl line 146 |
| 16:54 |
|
kbeswick joined #evergreen |
| 17:06 |
Bmagic|2 |
jeff: would it be helpful to paste the opensrf object dump after the call? |
| 17:07 |
jeff |
yes. |
| 17:07 |
jeff |
well, to a pastebin or gist. not directly here if it's long. :-) |
| 17:07 |
jeff |
i'm curious what the error you're getting is. |
| 17:07 |
jeff |
i'd start with that. |
| 17:07 |
Bmagic|2 |
it's pretty good sized, I was about to use the pasting utility on pate.evergreen.... |
| 17:09 |
Bmagic|2 |
jeff: http://pastebin.com/wKE988YB |
| 17:09 |
|
mmorgan left #evergreen |
| 17:09 |
|
stevenyvr2 joined #evergreen |
| 17:10 |
|
tater-home joined #evergreen |
| 17:10 |
|
tater-home left #evergreen |
| 17:11 |
jeff |
Bmagic|2: have you tried passing the userid as "12345" instead of 12345? |
| 17:15 |
jeff |
might make a difference. might not. i just noticed that you weren't, and in my example i had been. that could have been self-cargo-culting on the part of myself, though. |
| 17:15 |
Bmagic|2 |
jeff: no I have not |
| 17:15 |
Bmagic|2 |
jeff: However, the authentication is successful |
| 17:15 |
jeff |
Bmagic|2: in your pasted output, i didn't see the error response, but i did see at least one password-looking string. FYI. |
| 17:16 |
Bmagic|2 |
jeff: Maybe the problem is not the opensrf call, it's my understanding of what should happen to the database |
| 17:16 |
jeff |
also, i'd look at your opensrf logs when you make this request, to see what happens there. |
| 17:17 |
jeff |
Bmagic|2: you could try making the request via srfsh to compare. |
| 17:17 |
Bmagic|2 |
jeff: The issue that we are facing - fine gererator does not create a penalty for users who have more than the maximum overdues when they are inside of the grace period. Therefore, they are able to check out a book. Interestingly enough though, the staff client stops them from checking out more than 1 item (because of the refresh mechanizm) |
| 17:17 |
* jeff |
nods |
| 17:18 |
jeff |
i remember. :-) |
| 17:18 |
Bmagic|2 |
jeff: so, from your understanding, would an appropriate row in actor.usr_standing_penalty be required to prevent the checkout? |
| 17:19 |
jeff |
this seems to be a bit of an inelegant workaround, though perhaps useful as a learning experiment. :-) |
| 17:20 |
Bmagic|2 |
jeff: I only ask, because the (Has overdues) comes up for this test patron even without the row in the table |
| 17:20 |
jeff |
Bmagic|2: yes, a row in actor.usr_standing_penalty corresponding to the user in question with the standing penalty type in question is what i'd expect to see for a user with that penalty |
| 17:20 |
jeff |
has overdues is something else. |
| 17:21 |
Bmagic|2 |
jeff: I think I need (max overdues) |
| 17:34 |
|
dMiller joined #evergreen |
| 18:06 |
|
dMiller_ joined #evergreen |
| 18:15 |
|
mjingle left #evergreen |
| 18:44 |
|
kbeswick joined #evergreen |
| 18:44 |
|
stevenyvr joined #evergreen |
| 20:38 |
|
tater-home joined #evergreen |
| 20:38 |
|
tater-home left #evergreen |
| 21:02 |
|
kbeswick joined #evergreen |
| 21:34 |
|
kbeswick joined #evergreen |
