| Time |
Nick |
Message |
| 00:43 |
|
zerick joined #evergreen |
| 01:52 |
gmcharlt |
eeevil: dbwells: the patch for bug 1311467 should probably get eyes on it soon |
| 01:52 |
pinesol_green |
Launchpad bug 1311467 in Evergreen "validate headings broken in 2.6.0" (affected: 1, heat: 6) [High,New] https://launchpad.net/bugs/1311467 |
| 01:58 |
gmcharlt |
@later tell yboston I've increased the upload limit again |
| 01:58 |
pinesol_green |
gmcharlt: The operation succeeded. |
| 06:54 |
|
timf joined #evergreen |
| 07:44 |
|
jboyer_isl joined #evergreen |
| 07:51 |
eeevil |
gmcharlt: looking now |
| 07:53 |
|
collum joined #evergreen |
| 08:02 |
|
jboyer-isl joined #evergreen |
| 08:06 |
|
akilsdonk joined #evergreen |
| 08:09 |
gmcharlt |
eeevil: thanks |
| 08:12 |
|
mdriscoll joined #evergreen |
| 08:14 |
|
George_Frideric joined #evergreen |
| 08:21 |
|
ericar joined #evergreen |
| 08:27 |
jboyer-isl |
Floating Point + Math = sucks. |
| 08:35 |
|
Dyrcona joined #evergreen |
| 08:44 |
|
mmorgan joined #evergreen |
| 08:48 |
jl- |
morning |
| 08:50 |
dbwells |
jboyer-isl: I think you mean " = sucks.0000000000000004" |
| 08:50 |
jboyer-isl |
dbwells++ |
| 08:51 |
jboyer-isl |
Or suckr, because I'm usually off by one. (too much int($float * 100)...) |
| 08:51 |
|
RoganH joined #evergreen |
| 08:55 |
jl- |
I want to enable the opac.located_uri.act_as_copy global flag |
| 08:55 |
jl- |
would this be sufficient for a located URI (see Marc Record at bottom) |
| 08:56 |
jl- |
856.‡9SHIP |
| 08:57 |
RoganH |
I'm assuming SHIP is an org unit name for you? You'll want the URI and maybe a display subfield too. $u and $z |
| 08:57 |
jl- |
RoganH: it is an org unit |
| 08:58 |
jl- |
$u seems to be a link which I dont want |
| 08:58 |
jl- |
I want just the record |
| 08:58 |
jl- |
and $y would be link text |
| 08:58 |
jl- |
which I also don't want / need since I'll be batch adding 856 9 SHIP |
| 08:58 |
jl- |
is this feasible? |
| 08:58 |
eeevil |
jl-: it doesn't work like that |
| 08:59 |
jl- |
hmm |
| 08:59 |
eeevil |
jl-: located URIs are for URIs. if you don't want a link, you need to attach a copy to the record |
| 08:59 |
RoganH |
jl-: are you trying to restrict what orgs the bib shows up to? |
| 08:59 |
jl- |
eeevil: the dilemma is, I really want scope search for demonstration purpose.. and migrating the copies from the previous ILS.. would be overkill for me with my resources. |
| 09:00 |
jl- |
if I could batch add generic copies that would be fine |
| 09:00 |
jl- |
RoganH: yes |
| 09:00 |
jl- |
you can see my dev install |
| 09:00 |
RoganH |
jl-: eeevil was ahead of me on my response to that :) |
| 09:00 |
RoganH |
jl-: you can batch add copies via sql inserts, especially if you don't care about call numbers being random :) |
| 09:01 |
jl- |
see "KLN > Shippensburg University" for units, next will be Lincoln University |
| 09:01 |
dbwells |
jl-: I guess what you are trying might work strictly for fake, demo purposes. |
| 09:01 |
jl- |
ah, the master of SQL is speaking ;) |
| 09:01 |
jl- |
dbwells: that would be nice |
| 09:01 |
dbwells |
jl-: You'll probably need a placeholder URI in there, though. |
| 09:02 |
jl- |
ok so I have no problem doing that for my current bibs |
| 09:02 |
jl- |
they are all ship bibs |
| 09:02 |
dbwells |
jl-: Also, I think you need certain indicators to have it picked up as a located URI. 4|0 or 4|1 (OTTOMH) |
| 09:02 |
RoganH |
[off topic] Occasionally I think about adding fake books that should exist just to see if anyone would put a hold on them like "The Love Life of the Polar Walrus". |
| 09:03 |
jl- |
rofl! |
| 09:03 |
jl- |
dbwells: I thought if I have copies, I won't need located URI? |
| 09:03 |
jl- |
because scope search will look for the copies? |
| 09:04 |
RoganH |
jl-: yeah, restrict your scope by org unit and only copies at those orgs will show up |
| 09:04 |
dbwells |
jl-: Yes, you only need one or the other. |
| 09:04 |
dbwells |
I thought we were talking about adding fake URIs vs adding fake copies. |
| 09:04 |
RoganH |
jl-: dbwells is right, you probably can get the URI way to work but without copies things will still be off and not give an accurate impression even for a demo |
| 09:05 |
RoganH |
jl-: just my 0.02 dinars |
| 09:05 |
jl- |
I think if I can use SQL for fake copies that might be fine |
| 09:06 |
jl- |
I'm ok with random barcodes |
| 09:06 |
jl- |
(instead of located uri's) |
| 09:07 |
dbwells |
I'd say that's the better route. Certainly a less fake fake. |
| 09:07 |
jl- |
cool |
| 09:08 |
eeevil |
dbwells / gmcharlt: 2d65580268 looks good to me. signing off. dbwells, I can push if you like |
| 09:09 |
dbwells |
eeevil: please do, thanks |
| 09:09 |
|
kbeswick joined #evergreen |
| 09:10 |
jl- |
so RoganH, what would the SQL look like? :) would this go into asset.copy? |
| 09:11 |
dbwells |
jl-: for quick creation of fake assets, I'd check out the stuff in the "Concerto" test set. |
| 09:11 |
RoganH |
jl-: I'm not sure what data you're working off of. the concerto stuff would be a good jumping off point. |
| 09:11 |
dbwells |
jl-: specifically, the functions in Open-ILS/tests/datasets/sql/env_create.sql and their use in Open-ILS/tests/datasets/sql/assets_concerto.sql |
| 09:11 |
RoganH |
jl-: basically you need entires in asset.call_number that reference bib records |
| 09:12 |
RoganH |
jl-: then entries in asset.copy that reference the call numbers |
| 09:12 |
eeevil |
dbwells: done, sir |
| 09:12 |
|
dluch joined #evergreen |
| 09:14 |
pinesol_green |
[evergreen|Galen Charlton] LP#1311467: fix heading validation in bib editor - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=3665b7c> |
| 09:15 |
|
RoganH left #evergreen |
| 09:15 |
|
RoganH joined #evergreen |
| 09:27 |
jl- |
RoganH: I did use the concerto_loader but I attached my bibs to it and copied out assets part |
| 09:27 |
jl- |
load_concerto |
| 09:27 |
jl- |
copied out = commented out |
| 09:29 |
bshum |
General question, now that 2.6.0 is on the downloads page, shouldn't we remove 2.3 series from there since it's past it's 3 months security fix time period and eventually when we add 2.4.7, change 2.4's status to security updates only? |
| 09:31 |
csharp |
bshum: I think that's the right approach |
| 09:34 |
RoganH |
jl-: That's a good start. So you have your bibs but no call numbers or assets yet, right? |
| 09:35 |
jl- |
RoganH: correct |
| 09:36 |
|
yboston joined #evergreen |
| 09:37 |
RoganH |
jl- : :) |
| 09:38 |
RoganH |
jl-: you have a few options, one would be to modify the concerto asset sql, downside being that it's really oriented towards that data set and might take a lot of modification since you want your records there |
| 09:39 |
RoganH |
jl-: I would lean towards grabbing random bibs by ID and creating call numbers for each org to whatever limit you want using something (author's name?) as the call number label |
| 09:39 |
RoganH |
jl-: then auto creating copies for each of those with random barcodes |
| 09:39 |
* bshum |
changed the status of 2.4 to security updates only anyways. Since basically that's where we are already. |
| 09:40 |
RoganH |
jl-: you could get fancier with distinguishing fiction and non-fiction, doing dewey ranges if you want to spend that time as well |
| 09:42 |
RoganH |
jl- : assets_concerto.sql will give some great code for working off of so you don't have to reinvent the wheel |
| 09:42 |
jl- |
RoganH: I don't need anything fancy, at all |
| 09:42 |
jl- |
just functional |
| 09:43 |
RoganH |
jl-: it'll get you close enough that it'll look real, things like call numbers will be wonky but that's easy to explain off |
| 09:44 |
* bshum |
is also updating all the other places, Windows Staff Clients (for 2.5 and 2.4) and the Code Museum since he's poking anyways. |
| 09:44 |
jl- |
max(id) 628807 |
| 09:44 |
jl- |
so all of those need copies |
| 09:44 |
jl- |
some of those are authorities tho |
| 09:45 |
jl- |
we don't want those authorities to get copies right? |
| 09:45 |
RoganH |
jl- : no. but you can make another table to work off of, do a create table as select non authories non-deleted (though none of yours are probably deleted yet) |
| 09:45 |
jl- |
nvm |
| 09:45 |
jl- |
authority.record_entry |
| 09:46 |
RoganH |
jl- : yeah, I'd just use biblio.record_entry |
| 09:47 |
|
ericar joined #evergreen |
| 09:50 |
jl- |
RoganH: right off the bat, which lines can I take out of assets_concerto |
| 09:50 |
jl- |
I need to strip it down first |
| 09:52 |
RoganH |
jl- : If you're going for a big bulk insert of semi random items you'll probably just want to cherry pick out of it the code you want to use |
| 09:52 |
RoganH |
jl- : it's more of a crafted realistic (but small) sample set |
| 09:53 |
RoganH |
[off topic] I still think an alternative sample data set for the Miskatonic Consortium would be fun. |
| 09:54 |
jl- |
RoganH: http://collabedit.com/vsc4w |
| 09:54 |
jl- |
do I need to create call numbers? |
| 09:56 |
RoganH |
jl- : yep, think of the chain as bibs -> call numbers -> copies |
| 09:58 |
|
mrpeters joined #evergreen |
| 10:00 |
* bshum |
reminds jl- of http://ur1.ca/h4zj2 and how copies are organized |
| 10:00 |
jl- |
thanks |
| 10:00 |
bshum |
RoganH++ |
| 10:04 |
yboston |
bshum: who created that google doc explaining copies / call nums? |
| 10:04 |
bshum |
yboston: I did |
| 10:05 |
bshum |
Earlier this week |
| 10:05 |
yboston |
bshum: I like it, we should added to the offical docs |
| 10:05 |
bshum |
I've been thinking about where we could stick something like it in the core |
| 10:05 |
yboston |
s/added/add it/ |
| 10:05 |
yboston |
I'll think about ti too if it helps |
| 10:05 |
bshum |
Could probably enhance it further, but as a starting point, I was just trying to get the point across about how information hangs together. |
| 10:05 |
yboston |
very well done |
| 10:06 |
yboston |
learned about asset.uri |
| 10:06 |
bshum |
eeevil: gmcharlt: Does the most recent changeset 3665b7c need to be backported to the rel_2_6 branch too? |
| 10:06 |
pinesol_green |
[evergreen|Galen Charlton] LP#1311467: fix heading validation in bib editor - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=3665b7c> |
| 10:06 |
mmorgan |
bshum++ |
| 10:06 |
|
montgoc1 joined #evergreen |
| 10:07 |
eeevil |
bshum: it was |
| 10:07 |
bshum |
Oh duh |
| 10:07 |
bshum |
I see what I did wrong now |
| 10:07 |
bshum |
The top commit in rel_2_6 wasn't the same because it was a docs change |
| 10:07 |
bshum |
Nevermind I see it now |
| 10:07 |
eeevil |
cool |
| 10:08 |
|
denishpatel joined #evergreen |
| 10:21 |
|
ldwhalen joined #evergreen |
| 10:22 |
pinesol_green |
[evergreen|Thomas Berezansky] LP#1310283: Propogate .staff through Z3950 searches - <http://git.evergreen-ils.org/?p=Evergreen.git;a=commit;h=f6bb791> |
| 10:29 |
jl- |
does each copy have to have a unique call nr? |
| 10:31 |
Dyrcona |
jl-: No, but call number labels have to be unique per org. unit. |
| 10:31 |
Dyrcona |
jl-: More than 1 copy can use the same call number. |
| 10:31 |
Dyrcona |
In database terms it is a one to many relationship from call number to copy. |
| 10:31 |
jl- |
with org unit you mean owning_lib ? |
| 10:31 |
Dyrcona |
Yes. |
| 10:32 |
jl- |
what table stores owning_lib ? |
| 10:32 |
Dyrcona |
owning_lib refers to actor.org_unit.id |
| 10:33 |
Dyrcona |
If you do \d asset.call_number in psql, you can get all the information about the table and its foreign key relationships. |
| 10:34 |
jl- |
got it |
| 10:35 |
|
kmlussier joined #evergreen |
| 10:37 |
jl- |
Dyrcona: I'm trying to create fake copies for all existing bibs with SQL for library scope search demonstration since I dont want to migrate real copies for now -- what do you mean with call number labels have to be unique per org.? |
| 10:38 |
jl- |
look here http://collabedit.com/vsc4w is the Call Number Label 'Import Concerto'? |
| 10:38 |
Bmagic |
At the Evergreen conference there was a mention of redesigning the UI for reporting. Im having trouble finding documentation on this via open-ils.org |
| 10:38 |
Dyrcona |
asset.call_number.label has a unique index per asset.call_number.owning_lib. |
| 10:39 |
Dyrcona |
Bmagic: PINES has contracted Emerald Data to work on that AFAIK. |
| 10:39 |
Bmagic |
Ah |
| 10:40 |
Dyrcona |
jl-: That means if you have two copies with a call number of 324.56 JUL at Branch 1, then you can only have 1 asset.call_number entry, and both copies refer to the same call_number. |
| 10:43 |
jl- |
right |
| 10:44 |
Dyrcona |
jl-: In your example the label is the second parameter of the evergreen.populate_call_number functions. |
| 10:44 |
Dyrcona |
asset.call_number.label is the call number that gets displayed, the prefix and suffix, if any will be added to it. |
| 10:45 |
jl- |
what are the other parameters in there? |
| 10:45 |
Dyrcona |
Not sure. I've never used the function. |
| 10:46 |
|
RoganH left #evergreen |
| 10:46 |
jl- |
I wish there was a generate_fake_copies.pl |
| 10:46 |
kmlussier |
Bmagic: There were some wireframes for that project floating around at some point. But I think they may be a couple of years old. |
| 10:47 |
Dyrcona |
jl-: I've heard of such scripts floating around. It would not be so difficult to write one. |
| 10:47 |
Dyrcona |
jl-: You should try looking in some of the lesser known repos out there. You might find something. |
| 10:51 |
kmlussier |
Bmagic: http://evergreen-ils.org/dokuwiki/doku.php?id=evergreen-reports:reports_wireframes |
| 10:52 |
Bmagic |
kmlussier++ #thanks for digging out that link |
| 10:52 |
jl- |
Dyrcona: let me know if you can find something |
| 10:52 |
jl- |
I haven't been able to |
| 10:53 |
Dyrcona |
jl-: I'm not doing your homework. :) |
| 10:53 |
jl- |
right |
| 10:53 |
jl- |
not asking for that either |
| 10:53 |
jl- |
:) |
| 10:54 |
|
Dyrcona left #evergreen |
| 10:54 |
jl- |
would be useful for demo purposes tho |
| 10:54 |
|
Dyrcona joined #evergreen |
| 10:55 |
csharp |
Bmagic: the alternative reports creator UI is under development now - I don't have a lot of details, nor do I have an ETA, but I"m happy to answer/redirect any questions |
| 10:55 |
* Dyrcona |
grumbles something about "tap to click...." |
| 10:55 |
Dyrcona |
jl-: It turns out, we have a copy of the old ILS-Contrib repo. |
| 10:55 |
remingtron |
akilsdonk: ericar: I'm getting 404 errors on old ESI docs pages, where should I be looking? |
| 10:55 |
* csharp |
likes tap to click (except when he doesn't) |
| 10:55 |
remingtron |
Example: http://esilibrary.com/docs/asciidoc/2.5_Default_Values_Vandelay_Item_Import_doc/2.5_Default_Values_Vandelay_Item_Import.txt |
| 10:56 |
Dyrcona |
jl-: Have a look at this: http://git.mvlcstaff.org/?p=Evergreen/ILS-Contrib.git;a=tree;f=import_demo/trunk;h=32ba80abf9363c86de5533083a2e0e9f38714a66;hb=refs/heads/master |
| 10:56 |
* bshum |
mumbles something about publishing development plans for community review early in case it'll matter to the next RM whether something will go in or not to the next release.... |
| 10:56 |
Bmagic |
csharp: I was just seeking general info. Some people in our consortium were curious |
| 10:57 |
csharp |
sure thing - it will be a way to run pre-designed templates in a more polished interface (plus some new features) |
| 10:57 |
jl- |
Dyrcona: looks interesting, sec |
| 10:58 |
Bmagic |
csharp: but creation of the templates will be the same? |
| 10:58 |
akilsdonk |
remingtron: we're working on getting our documentation up on the new ESI website. All of the documentation that was posted on the old site is in the official docs and we've been adding new documentation to the official docs as well. As soon as we get the new ESI docs page up, we will be adding all new documentation there and directly to the official docs repository. |
| 10:59 |
csharp |
it will leverage the existing reports structure (i.e., clark kent, template/report/output) |
| 10:59 |
remingtron |
akilsdonk: cool, thanks for the info |
| 10:59 |
csharp |
templates will be pre-set (as far as the end user goes), but admin staff will be able to add templates |
| 10:59 |
csharp |
(like EG system admin staff, not local library staff) |
| 11:00 |
Bmagic |
csharp: Good info! Thanks again |
| 11:00 |
csharp |
happy to help! |
| 11:00 |
|
tspindler joined #evergreen |
| 11:00 |
* csharp |
looks around for what bshum is mumbling about ;-) |
| 11:21 |
bshum |
Random question |
| 11:21 |
bshum |
Right now, we pull in Business::ISBN via CPAN |
| 11:21 |
bshum |
is there any reason we can't use the packaged version like libbusiness-isbn-perl |
| 11:22 |
bshum |
(since the version in Trusty is 2.0.7, same as CPAN, I think?) |
| 11:22 |
bshum |
Or is this the sort of thing we're teasing out |
| 11:22 |
bshum |
With testing and breaking |
| 11:23 |
bshum |
Oh, nevermind I can see we already covered this with Wheezy |
| 11:24 |
* bshum |
ponders some edits to the section of the README about standalone DB requirements |
| 11:26 |
jeff |
bshum: you neverminded, but to the best of my knowledge, the only reason for bringing that in via CPAN was where it wasn't packaged by the distro in question. |
| 11:27 |
bshum |
jeff: That makes good sense. I guess that means we really ought to update the standalone DB portion of the README since it still uses CPAN and not packages that may exist. |
| 11:27 |
bshum |
I'm setting up a new standalone DB and it suddenly occurred to me to check on what packages may or may not exist |
| 11:28 |
* bshum |
will do some poking and whip up a branch for this later on. |
| 11:41 |
* jeff |
had fun killing a test vm repeatedly yesterday |
| 11:41 |
csharp |
@blame the test vm |
| 11:41 |
pinesol_green |
csharp: the test vm stole bradl's tux doll! |
| 11:42 |
jeff |
(intentional triggering of OOM conditions and adjusting Apache settings to avoid said conditions) |
| 11:42 |
* Dyrcona |
had "fun" upgrading his laptop to Ubuntu 14.04. |
| 11:42 |
jeff |
creating endless loops in php designed to quickly use as much memory as the php memory limit would allow, etc :-) |
| 11:43 |
|
Erick316 joined #evergreen |
| 11:44 |
bshum |
So pretty much everything that used to be CPAN can be done with packages now for the existing standalone DB list. But I wonder what else might be missing off that list that might be required. |
| 11:45 |
Erick316 |
hello |
| 11:46 |
Dyrcona |
Erick316: hello |
| 11:47 |
Erick316 |
I need some help, I'm triyng to import marc files into Evergreen, but I couldn't do a record match set that works |
| 11:49 |
Erick316 |
I already used the expression of the official documentation, but it didn't work, here is an example : Your Expression: (020 ‡a OR 022 ‡a OR 024 ‡a OR 028 ‡a) |
| 11:52 |
kmlussier |
Erick316: When you say it didn't work, what happened? |
| 11:52 |
kmlussier |
And what version of Evergreen are you on? |
| 11:53 |
Erick316 |
I´m usign Evergreen staff client 2.5.3, when import the records without the record match set all the records are in the queue, but when I use the record match set none of the records are brought to the queue |
| 11:55 |
kmlussier |
Erick316: How large is the batch of records you're trying to import? |
| 11:55 |
Erick316 |
I've already tried with other expressions and have the same result, I don't know if is the record match set or maybe I have another problem. |
| 11:56 |
jl- |
Erick316: are you looking to import *any* records or your own? |
| 11:57 |
Erick316 |
are about 35,000 records, but first I'm proving with less records |
| 11:57 |
kmlussier |
Unfortunately, I need to head out, but I'm sure somebody can help you out. |
| 11:57 |
Erick316 |
thank you anyway |
| 11:57 |
kmlussier |
FWIW, I would never triy to import that many records through the staff client. |
| 11:58 |
|
kmlussier left #evergreen |
| 11:58 |
Dyrcona |
Erick316: I don't think the importer is going to deal very well with 35,000 records with match sets all at once. |
| 11:58 |
Dyrcona |
Erick316: The staff client is likely to time out waiting on the server. |
| 11:58 |
jl- |
yes start with 1k |
| 11:59 |
jl- |
or even 1 |
| 11:59 |
jl- |
just for testing |
| 11:59 |
jboyer-isl |
Start with 5 to make sure the match set works the way you want, after that your upper limit is determined by the complexity of your matches and your hardware. |
| 11:59 |
jboyer-isl |
It will be a lower limit than you would hope. |
| 12:01 |
Erick316 |
yeah actually I have 10 records only for testing |
| 12:02 |
Erick316 |
but the expression for the record match set, I don't know if it is correct, have anyone some expression to make another test? |
| 12:02 |
jboyer-isl |
How many are already in the system? Match set timing is effected by both incoming and existing. |
| 12:03 |
Erick316 |
I' ve already used this : Your Expression: (020 ‡a OR 022 ‡a OR 024 ‡a OR 028 ‡a) and Your Expression: (item_lang AND (240 ‡a OR 245 ‡a)) |
| 12:03 |
Erick316 |
are 3 queues already in system some of 300, another of 600, and other of 3500 |
| 12:04 |
jboyer-isl |
Try a single field with a record that you know will match before trying 4 at a time. |
| 12:04 |
jboyer-isl |
That shouldn't be too many for it to work. |
| 12:06 |
jboyer-isl |
Does anyone currently in here use vandelay much? I'd be worried that the command line import methods are most common here. |
| 12:10 |
Dyrcona |
I don't think too many catalogers are in the channel. |
| 12:10 |
Dyrcona |
I imagine the ones who are are usually too busy to notice questions like these. |
| 12:12 |
jboyer-isl |
Erick316, have you tried the openils-general or openils-docs lists? Responses may not be immediate, but they may be better than here. |
| 12:12 |
|
jihpringle joined #evergreen |
| 12:13 |
|
Christineb joined #evergreen |
| 12:13 |
|
remingtron joined #evergreen |
| 12:18 |
Erick316 |
No, I already haven't |
| 12:20 |
csharp |
~lists | Erick316 |
| 12:20 |
pinesol_green |
Erick316: information on the Evergreen project's mailing lists is available here: http://www.evergreen-ils.org/listserv.php |
| 12:25 |
Dyrcona |
There is also a catalogers list. |
| 12:26 |
jl- |
Dyrcona: I'm going through http://git.mvlcstaff.org/?p=Evergreen/ILS-Contrib.git;a=blob;f=import_demo/trunk/generate_copies.sql;h=4724219c1c50270291403ed64951bfde14ae0c97;hb=refs/heads/master so does this generate copies from existing bibs biblio.record_entry? it seemsto be referring to staging_items, which is what? |
| 12:27 |
Dyrcona |
jl-: That script depends on one of the other ways of importing stuff, I guess. |
| 12:28 |
bshum |
jl-: So, going by the README file for that contrib |
| 12:29 |
bshum |
There's a link on line 74 to the Evergreen wiki |
| 12:29 |
bshum |
jl-: http://evergreen-ils.org/dokuwiki/doku.php?id=importing:holdings:import_via_staging_table |
| 12:29 |
Dyrcona |
jl-: There's lots of other stuff in that repo. |
| 12:29 |
bshum |
Which is how a staging table might be constructed I would imagine, based on how that generate_copies script operates. |
| 12:29 |
bshum |
I wasn't involved with it and it's for Evergreen 1.4, so it might be flaky / needs serious updating to work |
| 12:30 |
bshum |
But I would read the many things involved with the repo to see how things come together. |
| 12:32 |
bshum |
Ugh, this page is horrible. |
| 12:32 |
* bshum |
marks it as obsolete too |
| 12:32 |
bshum |
Well, I'll think about it, maybe after lunch. |
| 12:35 |
* jl- |
wishes he had another developer to work with here |
| 12:35 |
jl- |
thanks guys |
| 12:37 |
|
ryan joined #evergreen |
| 12:46 |
phasefx |
jl-: http://wiki.evergreen-ils.org/doku.php?id=advocacy:demonstration_data |
| 12:46 |
phasefx |
old, but maybe still useful |
| 12:48 |
jl- |
phasefx: interesting but this is not for generating copies? |
| 12:48 |
phasefx |
the bottom of the page is |
| 12:49 |
jl- |
indeed |
| 12:54 |
Dyrcona |
jl-: You noticed the python script linked at the bottom of the page? I knew there was something like that somewhere. |
| 12:56 |
jl- |
yes |
| 12:56 |
jl- |
the holy grail right at the bottom |
| 12:56 |
* phasefx |
had stored procedures that could generate fake data too, "but I have no idea where I put that stuff |
| 12:57 |
jl- |
let's see if I can make this work |
| 12:57 |
* jl- |
takes a snapshot just in case |
| 12:57 |
jl- |
gotta love virtualization |
| 12:57 |
|
akilsdonk joined #evergreen |
| 13:03 |
jl- |
SELECT id FROM biblio.record_entry ORDER BY RANDOM() LIMIT 1 |
| 13:03 |
jl- |
so the script will create random copies, but it won't create copies for each bib |
| 13:03 |
berick |
jl-: if the copy count is high enough, it will create copies for most bibs |
| 13:04 |
berick |
if you need copies for every bib, you could run the script then modify a few call numbers to link to any empty bibs |
| 13:04 |
jl- |
ok so if I have 600k bibs, I should set it to 1.2 million copies? |
| 13:05 |
jl- |
berick: it would be enough to have maybe 60-80% of my bibs covered |
| 13:05 |
berick |
yeah, probably |
| 13:06 |
berick |
if you're going to do that many, you may want to run them in batches |
| 13:06 |
berick |
of, say, 100k |
| 13:07 |
jl- |
berick: good idea, and maybe limit the id's to choose from |
| 13:07 |
|
bmills joined #evergreen |
| 13:08 |
jl- |
like a maxid of 300000 |
| 13:08 |
berick |
well, it will only chose a valid bib ID. the RANDOM() just affects the odering, not which ID to use |
| 13:09 |
jl- |
right but how can I set a ceiling for which bib id's to choose from |
| 13:09 |
jl- |
for my next sample data, it will have a different owner |
| 13:09 |
berick |
oh, yeah, if you need to set a ceiling, then sure |
| 13:09 |
jl- |
so from 1-300k library_owner 1 and then I'll run 300-600k with library_owner2 |
| 13:09 |
berick |
i misunderstood what you were trying to do |
| 13:09 |
berick |
gotcha, yeh |
| 13:09 |
jl- |
:) |
| 13:10 |
jl- |
this script makes me very happy |
| 13:16 |
mrpeters |
possible to accomplish alternate tpac skins without the ability to create new vhosts? have a customer who is super locked down on what hostnames the city will let them create and they don't see the need for a second one for the library. their goal is one hostname for the "in the building" TPAC and then their current one for the public, outside the building. |
| 13:18 |
tsbere |
mrpeters: Er, if they have two hostnames (internal vs external) then, er, you have two hostnames already? |
| 13:20 |
jl- |
SELECT id FROM biblio.record_entry WHERE id <= 1000 ORDER BY RANDOM() LIMIT 1 berick |
| 13:20 |
jl- |
ceiling |
| 13:20 |
jl- |
tested |
| 13:20 |
jl- |
;) |
| 13:21 |
bshum |
Heh |
| 13:21 |
berick |
jl-: cool, glad it's working |
| 13:21 |
jeff |
mrpeters: i think it might be possible with creative apache config, but i haven't tried yet. |
| 13:22 |
jeff |
mrpeters: that's not very helpful for an immediate answer, sorry |
| 13:22 |
jeff |
i think i've looked at this before, but it's been a while since i last looked. |
| 13:22 |
bshum |
mrpeters: Well, the city decides hostnames, so they control DNS. But you could always play ridiculous and add a custom vhost entry in their hosts file on the local machines |
| 13:22 |
jeff |
SetEnvIf wasn't helpful because it was a PerlVar being used... |
| 13:23 |
bshum |
"in the building" and all |
| 13:25 |
|
akilsdonk_ joined #evergreen |
| 13:26 |
eeevil |
mrpeters: and, strictly speaking, anyone can point a name at any IP as long as they own the name. so if a completely separate name is what they want, they could go register a different name and point it at the opac IP (assuming, of course, there's no transparent proxy in front that will filter at the protocol level) |
| 13:29 |
bshum |
Heh, I imagine it's not a huge stretch to refactor how the lib_ips.txt file works too, and change it to pass tpac variables for templating purposes. |
| 13:29 |
mrpeters |
they don't already have 2 hostnames...thats the problem |
| 13:29 |
mrpeters |
the hosts file isnt a bad idea though |
| 13:30 |
mrpeters |
think we need to just beat on the city some more and explain why its needed |
| 13:31 |
dbwells |
mrpeters: there's also always the option of running the second vhost on the same hostname, different port. That path would come with its own set of complications, though. |
| 13:31 |
mrpeters |
they have patrons getting mad that they can't get to certain databases that only work inside the library, so they want to go with different skins to avoid that |
| 13:31 |
tsbere |
mrpeters: THAT sounds like an easier problem to solve |
| 13:32 |
jl- |
berick: so if I want to give ~100k bibs I'll set the num_copies to 200k and also for num_volumes? not sure what the difference is |
| 13:32 |
jl- |
*give 100k bibs a copy |
| 13:32 |
|
ldwhalen joined #evergreen |
| 13:32 |
mrpeters |
you wouldn't go with alternating skins, tsbere, depending on your location? |
| 13:32 |
tsbere |
mrpeters: Use rewrite rules to set an ENV variable, then decide if you want to show the databases based on the value of said variable in the templates. One set of templates! |
| 13:32 |
mrpeters |
hmmm thats not a bad idea either |
| 13:32 |
tsbere |
(or just tell the templates how to check relevant ENV vars directly, I guess) |
| 13:34 |
berick |
jl-: each volume can have multiple copies. if you set num_vols and num_copies the same, each volume will have one copy (which is fine) |
| 13:35 |
berick |
the fewer volumes you set, the more the copies will clump around bibs, which lowers the overall coverage |
| 13:35 |
berick |
so, yeah, set them both to 200k for 100k bibs |
| 13:36 |
jl- |
lgotcha |
| 13:36 |
jl- |
thanks for writing that script |
| 13:37 |
berick |
jl-: my pleasure. I wrote it for myself ;) |
| 13:37 |
bshum |
mrpeters: tsbere: So maybe something like IP condition checking like in http://www.netshinesoftware.com/blog/restricting-access-to-a-url-by-ip-address-using-mod-rewrite/ |
| 13:37 |
bshum |
Where you write something like RewriteCond %{REMOTE_ADDR} !^(A\.B\.C\.D)$ |
| 13:38 |
bshum |
And then set an apache variable to use a template override |
| 13:38 |
bshum |
Or other variable to define whether or not to show a link actually. |
| 13:39 |
tsbere |
bshum: RewriteCond whatever followed by RewriteRule . - [E=ENVVAR:ENVVALUE] |
| 13:39 |
tsbere |
bshum: Then ENV.ENVVAR in templates is set to ENVVALUE |
| 13:39 |
bshum |
Right |
| 13:39 |
tsbere |
bshum: Or if it is simple enough SetEnvIf would work too, I think |
| 13:40 |
bshum |
So wrap the link in like... [% IF ENVVALUE %] put <a>link here</a> [% END %] or some such? |
| 13:40 |
tsbere |
bshum: [% IF ENV.ENVVALUE %] |
| 13:40 |
tsbere |
But yea |
| 13:40 |
bshum |
And that'll only show the link if the apache variable is set |
| 13:40 |
bshum |
tsbere++ # apache rewrites are so great |
| 13:40 |
jeff |
for "different behavior within the same template dir based on some variable", we do that with our youth stuff |
| 13:41 |
tsbere |
Or, perhaps, say <a href="[%- IF ENV.ENVVALUE -%]realurl[%- ELSE -%]onlyavailablewithinthellibrarypageurl[%- END -%]">link text</a> |
| 13:43 |
jeff |
we also have a conditional include based on physical_loc: https://github.com/tadl/Evergreen_templates_tadlskin/blob/9c94d5359cc6db927579badc072ac9d2d126d8ea/templates_tadlskin/opac/home.tt2#L3 |
| 13:43 |
jeff |
IF ctx.physical_loc == '23'; INCLUDE "opac/parts/inlib_nav.tt2"; END; |
| 13:44 |
eeevil |
mrpeters: you could use mod_proxy to redirect to an external resource, make those external resources be vhosts on the same server, and user the server's /etc/hosts to point back at itself... |
| 13:45 |
eeevil |
IOW, just standard vhost config with mod_proxy in front to internally redirect based on something in the original URL |
| 13:45 |
|
dMiller__ joined #evergreen |
| 13:46 |
eeevil |
hrm... have to make sure /all/ links are host-relative, though. unsure if that's likely in tpac |
| 13:48 |
|
hbrennan joined #evergreen |
| 13:56 |
|
bmills joined #evergreen |
| 14:03 |
jl- |
berick: so for my example source 2 | 10 | System Local | f | t would probably be good? |
| 14:03 |
jl- |
for the bibs |
| 14:05 |
berick |
jl-: yep. any of the existing sources will work. |
| 14:05 |
berick |
anyone else having google issues? |
| 14:06 |
berick |
google.com, googleapis.com crapping out on me |
| 14:06 |
jl- |
even source 3? (transcendent) the purpose will be to have different library scope search |
| 14:06 |
jl- |
with different owners |
| 14:06 |
berick |
ah, yes, don't use a transcendent source -- sorry, forgot about those |
| 14:06 |
jl- |
google works fine for me |
| 14:07 |
berick |
hmm |
| 14:14 |
|
ldwhalen joined #evergreen |
| 14:14 |
berick |
no one on twitter complaining... i must be in a Googmuda Triangle. |
| 14:15 |
bshum |
Seems fine for us too |
| 14:15 |
berick |
ooh, someone on twitter finally said something. i'm not alone ;) |
| 14:16 |
* berick |
feels better |
| 14:18 |
* jeff |
is having a difficult time saying "Goodmuda" without a vague accent |
| 14:20 |
* jeff |
ponders the impact of regional googleapis / google outage vs AWS region outage |
| 14:20 |
jeff |
difficult to measure when they're not happening |
| 14:20 |
jeff |
though i'd say the googleapis/google scenario is slightly less challenging to measure/simulate |
| 14:28 |
csharp |
@who is complaining on twitter? |
| 14:28 |
pinesol_green |
rangi is complaining on twitter. |
| 14:31 |
|
ldwhalen joined #evergreen |
| 14:38 |
jl- |
trying to delete some records.. DELETE FROM biblio.record_entry WHERE ID BETWEEN 200000 AND 250000; returns DELETE 0 |
| 14:38 |
jl- |
max(id) is 300000 and there are no gaps |
| 14:38 |
berick |
jl-: it sets the deleted=true column on the bib when a DELETE is performed on bibs |
| 14:38 |
berick |
it doesn't actually DELETE anything |
| 14:38 |
berick |
hence the 0 |
| 14:39 |
jl- |
hmm |
| 14:39 |
jl- |
how do I actually delete anything then with sql? |
| 14:39 |
jl- |
in that table |
| 14:39 |
Dyrcona |
jl-: You want the short answer? |
| 14:40 |
jl- |
sure |
| 14:40 |
jl- |
whatever is feasible |
| 14:40 |
Dyrcona |
jl-: Drop the database and start over. |
| 14:41 |
jl- |
lol |
| 14:42 |
* Dyrcona |
is not kidding. |
| 14:42 |
Dyrcona |
At this point, you're learning Evergreen. |
| 14:42 |
Dyrcona |
The easiest thing is to wipe out the database, or even your whole VM, and start over. |
| 14:43 |
jl- |
vm sure but we do have a dev server that has 600k formatted records |
| 14:43 |
Dyrcona |
There are several things in Evergreen that are never really deleted from the database: bibliographic records, call numbers/volumes, copies, patrons, and a few others. |
| 14:43 |
jl- |
that I loaded with batches of 20 k |
| 14:43 |
jeffdavis |
Here's a long answer: you could drop the protect_bib_rec_delete rule on biblio.record_entry, delete as needed, and then re-enable that rule. But that tends to be an error-prone process, and other tables have similar protections... |
| 14:43 |
|
rjackson-isl joined #evergreen |
| 14:44 |
Dyrcona |
Yeah, there's a lot more to it than just dropping the rule. |
| 14:44 |
jl- |
ok seems to have a lot of consequences |
| 14:44 |
jeffdavis |
Arguably going through the process is a good way to learn more about what happens. But yeah, I prefer the short answer too. :) |
| 14:44 |
Dyrcona |
Well, if you don't have volumes and copies, it is a little easier. |
| 14:45 |
berick |
could also just leave the records there and adjust your queries to use WHERE deleted=FALSE (admittedly, I have no idea what your overall goal is) |
| 14:45 |
Dyrcona |
Well, 600,000 is a lot of junk records to be keeping around for no useful purpose. |
| 14:49 |
|
ningalls joined #evergreen |
| 15:00 |
jl- |
hmm is there a function to limit search results in the opac? |
| 15:00 |
jl- |
by default |
| 15:00 |
jl- |
going through 1m + records can be slow |
| 15:02 |
eeevil |
jl-: tl;dr: yes |
| 15:02 |
jl- |
is that a setting in the staff client? where? |
| 15:03 |
eeevil |
it's not. it's inherent in how search works |
| 15:03 |
eeevil |
search is paged (well, superpaged) ... it's ... complicated ;) |
| 15:04 |
eeevil |
search is not a simple SQL statement in evergreen |
| 15:10 |
jl- |
ok |
| 15:10 |
|
Dyrcona1 joined #evergreen |
| 15:13 |
Dyrcona |
Stupid wireless. |
| 15:17 |
* csharp |
wants an ethernet drop beside every power outlet |
| 15:18 |
* Dyrcona |
has 3 ethernet drops by his power, 1 in use by his desk phone. |
| 15:18 |
* Dyrcona |
also has an ethernet cable that is plugged into one of the drops just lying on his desk doing nothing. |
| 15:18 |
* Dyrcona |
should probably use it. :) |
| 15:18 |
tsbere |
csharp: And of course, once you get that you will find that you are invariably trying to use one that has no patch to a switch at the other end or is on the wrong vlan. ;) |
| 15:19 |
tsbere |
csharp: Assuming they aren't already all in use, anyway |
| 15:20 |
Dyrcona |
Whatever happened to network over power lines anyway? |
| 15:22 |
jeff |
Dyrcona: HomePlug: http://en.wikipedia.org/wiki/HomePlug |
| 15:24 |
bshum |
jl-: One quick way to limit a search in the OPAC is to make use of the logc variable. Like hostname/eg/opac/home?locg=100 (or whatever the ID is of the library you want to filter down to) |
| 15:24 |
bshum |
*locg, not logc |
| 15:25 |
bshum |
That'll apply a search filter to the OPAC results for that given org unit |
| 15:26 |
|
kbeswick joined #evergreen |
| 15:30 |
jl- |
bshum: makes sense, a search filter for copies and library owner? |
| 15:34 |
jl- |
org unit = library_owner |
| 15:34 |
jl- |
gotcha |
| 15:36 |
|
dMiller__ joined #evergreen |
| 15:38 |
* rangi |
waves from the equinox offices |
| 15:39 |
* bshum |
waves back at rangi |
| 15:39 |
bshum |
Have you posted a pic of your new laptop? |
| 15:39 |
rangi |
not yet |
| 15:40 |
rangi |
have now, on google+ |
| 15:40 |
jeff |
/msg rangi okay, now remember... the gear is hidden under the ping pong table. be sure to put it in place and ensure that the unit has power before you leave. |
| 15:40 |
rangi |
heh |
| 15:40 |
* jeff |
ducks |
| 15:50 |
eeevil |
rangi++ # and sorry I couldn't be there for the NZ beer :( |
| 16:00 |
|
JTaylor_ joined #evergreen |
| 16:01 |
gmcharlt |
rangi: howdy! |
| 16:05 |
|
dbs_mob joined #evergreen |
| 16:06 |
dbs_mob |
did someone say NZ beer? |
| 16:06 |
* dbs_mob |
hops on a plane to Atlanta |
| 16:06 |
* gmcharlt |
is imaging not just one dbs, but a mob of them |
| 16:06 |
* bshum |
should have just bought that ticket south. |
| 16:07 |
dbwells |
gmcharlt: I always imagine that as well :) |
| 16:28 |
|
tspindler left #evergreen |
| 16:31 |
|
mdriscoll left #evergreen |
| 16:38 |
bshum |
"Utopic Unicorn" really? |
| 16:38 |
bshum |
I mean unicorns sure. But utopic? |
| 16:40 |
Dyrcona |
Whatever. |
| 16:40 |
* Dyrcona |
starts installing OpenBSD..... |
| 16:42 |
Dyrcona |
They should have chosen Uguisu: "Their guano is used in face creams!" |
| 16:54 |
|
mrpeters left #evergreen |
| 17:03 |
gsams |
I've got an odd issue with expired holds on 2.3.5. We have an issue that is happening where items that have hit the hold expiration boundary are cancelling automatically with an odd timestamp, for instance 4/22/2014 12:6 PM |
| 17:04 |
gsams |
I don't feel like it should be cancelling the hold off of the patron account until after we check it in with the clear shelf expired holds flag enabled, but they are being cancelled it looks like |
| 17:05 |
csharp |
a unicorn is the first imaginary animal they've used |
| 17:05 |
bshum |
csharp: Indeed. |
| 17:05 |
bshum |
gsams: Was that supposed to be 12:6 or 12:06? |
| 17:05 |
gsams |
12:6 is the timestamp it is showing |
| 17:06 |
csharp |
eww |
| 17:06 |
gsams |
yeah, I'm worried about that |
| 17:06 |
gsams |
We have a few like that in the holds shelf browser, but most of them are normal timestamps |
| 17:07 |
csharp |
gsams: is it entered that way in the database? |
| 17:07 |
csharp |
seems like postgres would reject an invalid time |
| 17:07 |
csharp |
the expire time comes from the client machine (e.g., the patron's web browser when placing the hold) |
| 17:08 |
csharp |
they used to be able to edit that expire date, but that was pre TPAC |
| 17:08 |
csharp |
(that is to say that I don't know if they can edit it now) |
| 17:09 |
gsams |
the database shows a timestamp of 2014-04-22 12:06:56.199783-05 |
| 17:09 |
bshum |
And that's "expire_time" or "shelf_expire_time" |
| 17:10 |
bshum |
Cause those are different things |
| 17:10 |
gsams |
That's actually the cancel_time in the database |
| 17:10 |
gsams |
the expire_time shows null |
| 17:10 |
csharp |
hmm |
| 17:10 |
bshum |
null expire_time isn't wholly unusual |
| 17:10 |
gsams |
but the reason clearly shows as hold shelf expiration |
| 17:11 |
csharp |
gsams: okay - well, then we know it's a display issue |
| 17:11 |
csharp |
oh - hold shelf expiration, not hold expiration |
| 17:11 |
gsams |
yes, sorry if I mispoke at some point |
| 17:12 |
bshum |
I don't have a 2.3 system handy anymore |
| 17:12 |
gsams |
it does have a shelf_expire_time as well |
| 17:12 |
csharp |
gsams: they aren't cancelling automatically - it probably means that a staff member entered the "Clear Shelf Expired Holds" interface |
| 17:12 |
bshum |
But yeah, what csharp says |
| 17:12 |
bshum |
Maybe they went to the wrong menu option |
| 17:12 |
bshum |
And that cancels holds |
| 17:12 |
gsams |
wouldn't that remove it from the hold shelf browser? |
| 17:13 |
csharp |
just clicking on that menu item will automatically cancel all holds past the configured shelf expire time |
| 17:13 |
csharp |
we've greyed it out (ala bshum's patch) |
| 17:13 |
csharp |
or phasefx's patch via bshum |
| 17:13 |
bshum |
It'll probably leave them on the shelf browser though. |
| 17:13 |
gsams |
hmmm. |
| 17:13 |
bshum |
Till checked in |
| 17:14 |
bshum |
But yeah, use of the menu option, even accidentally could mess things up. |
| 17:14 |
gsams |
that might explain it then, I'll have to find out if someone has been pushing buttons |
| 17:14 |
csharp |
that menu item should really be removed - this is going to happen over and over |
| 17:14 |
gsams |
are you talking about the clear these holds button on the clear shelf expired holds interface? |
| 17:14 |
csharp |
gsams: if you have a bunch of items showing with that same cancel timestamp, that's almost certainly what happened |
| 17:14 |
csharp |
"what does this button do" BOOM |
| 17:15 |
bshum |
csharp: I'm not sure how I feel about removing someone else's "feature" but it does seem to cause more pain than expected in my opinion. |
| 17:15 |
gsams |
yeah that is what it looks like then |
| 17:15 |
csharp |
do we know whose feature it was? |
| 17:15 |
bshum |
csharp: Not off the top of my head, but it probably came into being during the big rewrites of 2.0/2.1 |
| 17:16 |
* csharp |
requests a "drop my database" menu item button and makes everyone else just live with it ;-) |
| 17:16 |
gsams |
csharp++ |
| 17:16 |
bshum |
"Clear Hold Shelf" process is noted in the long, long list of 2.0 features |
| 17:17 |
bshum |
Back then, we weren't as good with tracking new features with LP, etc. |
| 17:17 |
bshum |
Might only exist as some note in SVN and the memory of the original devs |
| 17:17 |
csharp |
@fix that problem |
| 17:17 |
pinesol_green |
csharp: Mr. Spock: Something fascinating just happened. |
| 17:17 |
* csharp |
wants a @fix command that "fixes" whatever we want |
| 17:18 |
gsams |
I don't suppose there is a good way to see who might have accidentally hit the button? |
| 17:18 |
csharp |
gsams: I think only logs will tell |
| 17:18 |
csharp |
the user id will be in the log messages |
| 17:18 |
bshum |
Probably right, holds don't have audit trails by default. |
| 17:19 |
gsams |
mkay, that'll have to do then |
| 17:19 |
bshum |
At least you have the exact moment when the hold was cancelled. |
| 17:19 |
pinesol_green |
Incoming from qatests: Test Success - http://testing.evergreen-ils.org/~live/test.html <http://testing.evergreen-ils.org/~live/test.html> |
| 17:19 |
gsams |
I think I know what happened though, and timestamps are enough for any local damage that might have been done. |
| 17:19 |
bshum |
That'll help with log sifting |
| 17:19 |
gsams |
csharp++ |
| 17:19 |
gsams |
bshum++ |
| 17:19 |
gsams |
many thanks! |
| 17:20 |
|
mmorgan left #evergreen |
| 17:23 |
gsams |
turns out it was me! Ha! |
| 17:23 |
gsams |
I was answering questions for our incoming library and cancelled holds by accident |
| 17:24 |
gsams |
that sounds like a thing I would do while doing something totally innocuous. |
| 17:24 |
gsams |
so... what is this change that can disable this "feature" that I might be able to implement? |
| 17:28 |
* bshum |
points gsams at http://irc.evergreen-ils.org/evergreen/2013-09-05#i_27907 |
| 17:32 |
gsams |
bshum++ will give that a go |
| 17:34 |
bshum |
gsams: Should be as simple as dropping that in and then restarting staff clients to see the menu option grey out |
| 17:36 |
gsams |
bshum: that did the trick, thank you very much. Saving me from a lot of headaches down the line. |
| 17:36 |
gsams |
bshum++ |
| 17:37 |
gsams |
phasefx++ # for helping disable the clear shelf expired holds option back in September of 2013 |
| 17:41 |
bshum |
csharp: I'm building a new trusty demo to poke at the ubuntu makefile you setup for Evergreen. |
| 17:41 |
bshum |
OpenSRF seems fine with the revisions I threw in my working branch |
| 17:41 |
bshum |
Or at least, the basic test passes |
| 17:42 |
csharp |
cool |
| 17:42 |
csharp |
yeah - I was updating the Evergreen makefile to be able to test the opensrf makefile ;-) |
| 17:43 |
csharp |
beyond the math test, that is |
| 17:43 |
bshum |
Hehe, yeah |
| 17:43 |
bshum |
On paper, all the changes you suggest look alright. |
| 17:43 |
bshum |
I'm still not entirely sure what gmcharlt was warning us to watch for, some change in perl I think. |
| 17:43 |
bshum |
But I'm sure we'll hit something new |
| 17:47 |
gmcharlt |
bshum: upshot - hash key sort randomization caused failures in Koha's test cases of its copy of QueryParser |
| 17:48 |
gmcharlt |
that may (or may not) apply to Evergreen |
| 17:48 |
gmcharlt |
the issue manifested as a given query randoming resulting in different parse trees |
| 17:49 |
gmcharlt |
hopefully isn't not widespread |
| 17:50 |
bshum |
Guess we'll see. |
| 17:52 |
gmcharlt |
speaking of which, a minor example in OpenSRF - https://bugs.launchpad.net/opensrf/+bug/1285915 |
| 17:52 |
pinesol_green |
Launchpad bug 1285915 in OpenSRF "t/09-Utils-JSON.t can fail on Perl 5.18" (affected: 1, heat: 6) [Low,New] |
| 18:03 |
|
jwoodard joined #evergreen |
| 18:08 |
|
vlewis joined #evergreen |
| 18:41 |
|
berick joined #evergreen |
| 18:49 |
hbrennan |
Hey all - Just returned from lunch to discover we're running on offline mode (messed with db, jabber prob) |
| 18:50 |
hbrennan |
One of our offline circ stations is giving a print_tree error at the end of a transaction and won't print a receipt... These transactions are still being logged though, right? |
| 18:50 |
hbrennan |
This particular circ station normally required a hit of the 'OK' to actually print receipts, while our other stations automatically spit it out |
| 18:57 |
hbrennan |
nm. Figured it out |
| 18:57 |
hbrennan |
Answer is yes, the transactions are still logged |
| 18:57 |
hbrennan |
I pulled a bshum (conversation with self) |
| 18:58 |
gsams |
hbrennan++ |
| 18:58 |
gsams |
I did that a few days ago I think |
| 18:58 |
hbrennan |
Sometimes you just have to say things "out loud" |
| 18:59 |
hbrennan |
Then panic because no one is there to help, and then figure it out |
| 18:59 |
gsams |
I find it very effective myself |
| 18:59 |
gsams |
nothing like being under fire |
| 19:00 |
hbrennan |
Luckily it wasn't me who broke it |
| 19:01 |
hbrennan |
Welp, with EG down I think I'll take a break and get some sunshine while delivering piles of electronics to recycling! |
| 19:48 |
|
dcook joined #evergreen |
| 20:00 |
gsams |
Got a request from our migrating library, they desperately need a way to search call numbers that isn't the bib call number option or a shelf browse. |
| 20:01 |
gsams |
unfortunately bib call number has nothing to target (so it appears) and shelf browse isn't a good option for them |
| 20:02 |
gsams |
they have all of their call numbers set to generic FIC/dewey style call numbers no matter where they are at, and if a patron approaches their staff with a call number they have to be able to search in a way that lists all the possibilities |
| 20:04 |
gsams |
if anyone has options that might be of use for this, I would greatly appreciate any help or pointing in the right direction to make this happen. |
| 20:44 |
|
kmlussier joined #evergreen |
| 20:49 |
|
gsams joined #evergreen |
| 20:58 |
|
_bott_ joined #evergreen |
| 21:09 |
bshum |
hbrennan: Ha, I'm glad I can be an example :) |
| 21:10 |
hbrennan |
bshum: You're a great one! |
| 21:12 |
bshum |
I just type a lot now. Used to be you couldn't get me to say anything at all. My boss had to order me to say or ask at least one thing in IRC every day at the beginning when I was still a newbie. |
| 21:13 |
hbrennan |
That's a great idea |
| 21:13 |
* kmlussier |
tries to imagine a world where bshum needs to be forced to talk in IRC. |
| 21:13 |
hbrennan |
haha |
| 21:13 |
hbrennan |
I type really fast, so a lot more often comes out than needs to |
| 21:13 |
hbrennan |
wow, how many grammar issues are in that? |
| 21:14 |
hbrennan |
Worse sentence ever |
| 21:14 |
hbrennan |
And there's the proof |
| 21:14 |
gsams |
haha |
| 21:14 |
kmlussier |
hbrennan: My typos in IRC are terrible. |
| 21:14 |
gsams |
bshum: I should probably follow that idea, ask a question in IRC at least once per day |
| 21:15 |
bshum |
When I used to play online games, I got into the habit of at least capitalizing the first letter and putting a period at the end of each sentence. My guildmates were snobs. :) |
| 21:15 |
gsams |
I'm a bit picky about my typing |
| 21:15 |
gsams |
though I do allow dropped caps and lack of periods |
| 21:15 |
hbrennan |
I'm very impressed with the typing skills of EGers in here |
| 21:16 |
hbrennan |
. |
| 21:16 |
gsams |
I just like to make sure that my message is clear. |
| 21:16 |
kmlussier |
I'm very impressed with the typing skills of my 9 yo. She can type nearly as fast as I can. |
| 21:16 |
hbrennan |
Now I'm self-conscious |
| 21:16 |
gsams |
haha |
| 21:16 |
bshum |
gsams: It's good to ask questions, I'm happy that I've evolved to be able to help answer a few from time to time. |
| 21:17 |
gsams |
bshum: I've learned quite a bit every time I have asked a question. Except my last one, which I imagine might be a while. |
| 21:18 |
gsams |
I'm still learning what can be changed easily and what might take actual development |
| 21:18 |
hbrennan |
This is the best place to play devil's advocate |
| 21:18 |
bshum |
That tends to be a fairly complicated question indeed. |
| 21:19 |
gsams |
aye, I like that the people on here are willing to share their experience and opinions about various questions I've asked |
| 21:19 |
bshum |
gsams: Bib call number browse sucks. I hate it and we removed it from our catalog a couple years back when we first went live on TPAC. |
| 21:19 |
bshum |
Mostly because in our consortium, hardly anyone used bib call numbers. |
| 21:20 |
gsams |
bshum: I'm going to remove it to, but the new library needs a call number search that isn't bib related or shelf browsing |
| 21:20 |
bshum |
What's wrong with the shelf browser? :) |
| 21:20 |
gsams |
haha |
| 21:20 |
gsams |
I've never had a problem with it |
| 21:20 |
bshum |
It'll still get you to the bib you're searching around, in theory. |
| 21:20 |
kmlussier |
gsams: Are they looking for a simple list of call numbers? |
| 21:21 |
gsams |
they need search results related to a call number |
| 21:21 |
bshum |
gsams: Ooh, pretty, I haven't looked at your catalog lately. I like the NTLC logo. |
| 21:21 |
kmlussier |
Oh, so not a browse list. |
| 21:21 |
gsams |
bshum: Thanks! It's actually one of the first things the new library helped with |
| 21:22 |
gsams |
kmlussier: yeah, normal search results based on a call number basically. |
| 21:22 |
kmlussier |
gsams: When I saw your e-mail, I thought you were looking for something like this: http://masslnc.cwmars.org/node/2590. Which we never funded, but the specs are there if anyone wants to run with it. :) |
| 21:23 |
gsams |
all of their fictional materials use the same "FIC ABC" structure so they need to be able to search by that |
| 21:24 |
gsams |
kmlussier: That is actually a pretty neat idea. I wish that was what I was looking for, though I don't know that I have the chops to implement that sort of features haha |
| 21:25 |
bshum |
Hmm |
| 21:25 |
gsams |
DVDs, Easy Books, Adult Fiction, Young Adult, etc. They apparently never thought they would need more descriptive call numbers until it was too large a project. |
| 21:26 |
gsams |
And of course, no one stores call number data in 092/099 in this group |
| 21:28 |
gsams |
Where does it expose the Volume information so that cnbrowse can work? Would it be possible to use the same system and remove the browse part? |
| 21:29 |
bshum |
So all that gets handled in Open-ILS/src/perlmods/lib/OpenILS/WWW/EGCatLoader/Search.pm |
| 21:30 |
bshum |
(I started looking at it from another angle, seeing where item_barcode as a qtype was expressed, but I can see cnbrowse in there too) |
| 21:33 |
gsams |
hrm |
| 21:37 |
gsams |
I think I'm going to look at this tomorrow, and maybe that will work better for me |
| 21:38 |
gsams |
at the moment, I feel too uncomfortable with the idea of hacking a call number search into this |
| 21:39 |
bshum |
It seems more new featurish to me. |
| 21:39 |
bshum |
aka, some sort of development. |
| 21:39 |
bshum |
But honestly, search has always been a weird place to look for things. Maybe there's something more obvious I'm missing. |
| 21:39 |
bshum |
(irony) |
| 21:41 |
bshum |
So... they need to be able to search by call numbers (on the volume level) by themselves and get a giant list? Or is this like, an additional filter. |
| 21:41 |
bshum |
I mean, how would you narrow a search from FIC ABC to something more relevant. |
| 21:43 |
gsams |
Basically yes, and I'm not sure how they would narrow things down from there |
| 21:43 |
gsams |
I'm planning on upgrading after the migration which would reimplement shelving locations at least, but I'm not sure otherwise. |
| 21:49 |
gsams |
It does sound like a feature request to me as well though |
| 21:49 |
bshum |
Indeed. |
| 21:50 |
bshum |
I can't find anything in QueryParser.pm for volume/call_number type search filters |
| 21:50 |
bshum |
locations, sure; lassos, you betcha; lucky?? okay... |
| 21:50 |
bshum |
But nope, nothing for raw CN lookups as a filter choice. |
| 21:52 |
gsams |
yeah, I was afraid of that too |
| 21:53 |
gsams |
I wasn't sure how it was pulling up the shelf browser, but I had a feeling it wouldn't be compatible(easily) with a regular search |
| 21:53 |
bshum |
It's probably adaptable. |
| 21:54 |
bshum |
But yeah, new feature, whee. |
| 21:54 |
bshum |
If it was me, I'd push them away from call number searching and finding more ways of narrowing or mapping their data to other areas. |
| 21:54 |
bshum |
:D |
| 21:55 |
gsams |
aye, I'm going to have a talk with their director about that, considering |
| 21:58 |
jeff |
hrm |
| 22:08 |
gsams |
bshum++ thanks for taking a look with me on that |
| 22:09 |
gsams |
I'm going to go home and see if there is something I can do with fresh eyes tomorrow. |
| 22:50 |
|
kmlussier left #evergreen |
