2009-12-10T00:13:19 *** dbs has quit IRC 2009-12-10T00:58:32 *** bshum has joined #evergreen 2009-12-10T00:59:01 *** bshum has left #evergreen 2009-12-10T01:05:55 *** tel1701 has joined #evergreen 2009-12-10T01:06:17 *** bshum has joined #evergreen 2009-12-10T01:07:35 *** bshum has left #evergreen 2009-12-10T01:09:44 *** tel1701 has quit IRC 2009-12-10T01:23:07 *** mck9 has left #evergreen 2009-12-10T02:35:14 *** eguest309 has joined #Evergreen 2009-12-10T02:49:48 *** pmplett has quit IRC 2009-12-10T03:14:18 *** [1]atz has joined #evergreen 2009-12-10T03:32:31 *** atz has quit IRC 2009-12-10T03:32:31 *** [1]atz is now known as atz 2009-12-10T03:37:23 *** eguest309 has left #Evergreen 2009-12-10T03:43:19 *** [1]atz has joined #evergreen 2009-12-10T03:56:19 *** [2]atz has joined #evergreen 2009-12-10T04:03:15 *** atz has quit IRC 2009-12-10T04:03:15 *** [2]atz is now known as atz 2009-12-10T04:14:25 *** [1]atz has quit IRC 2009-12-10T04:23:57 *** jmeeuwen has quit IRC 2009-12-10T04:36:26 *** gdunbar_ has joined #evergreen 2009-12-10T04:52:51 *** gdunbar has quit IRC 2009-12-10T05:48:40 *** pmplett has joined #evergreen 2009-12-10T06:10:07 @later tell dbs I tried Makefile.install with debian-lenny and couldn't find libmemcached-dev. Looks like it exists in squeeze and sid 2009-12-10T06:10:07 phasefx_: The operation succeeded. 2009-12-10T06:18:32 *** dbwells_ has joined #Evergreen 2009-12-10T06:22:11 @later tell dbs I was able to download the squeeze version of libmemcached2_0.31-1_i386.deb and libmemcached-dev_0.31-1_i386.deb and dpkg -i them with no dependency headaches. I forget how to mix in testing with stable using apt. Didn't try installing from source 2009-12-10T06:22:11 phasefx_: The operation succeeded. 2009-12-10T06:35:02 *** dbwells has quit IRC 2009-12-10T06:40:38 *** gmcharlt has joined #evergreen 2009-12-10T07:02:07 *** pmplett has quit IRC 2009-12-10T07:33:37 *** jmeeuwen has joined #evergreen 2009-12-10T07:40:20 *** gmcharlt has quit IRC 2009-12-10T07:45:20 *** gmcharlt has joined #evergreen 2009-12-10T07:45:57 *** gmcharlt has quit IRC 2009-12-10T07:46:07 *** gmcharlt has joined #evergreen 2009-12-10T08:05:46 *** mck9 has joined #evergreen 2009-12-10T08:09:34 *** natschil has joined #evergreen 2009-12-10T08:27:08 *** gdunbar has joined #evergreen 2009-12-10T08:43:24 *** gdunbar_ has quit IRC 2009-12-10T08:47:31 *** sfortin has joined #evergreen 2009-12-10T09:16:17 *** jenny has joined #evergreen 2009-12-10T09:23:33 weee.. no etch/lenny packages for libmemcached[-dev] 2009-12-10T09:23:39 * berick looks for sources 2009-12-10T09:25:24 can we just use libmemcache on etch/lenny? 2009-12-10T09:25:37 or does the api change 2009-12-10T09:26:12 different api 2009-12-10T09:26:38 well, macros or ifdef can handle that ... 2009-12-10T09:26:44 just sayin' ;) 2009-12-10T09:31:28 or configure/make/make install. just need to determine which is simpler to maintain in the long run.. 2009-12-10T09:34:35 debian testing definitely does have libmemcached 2009-12-10T09:34:47 it's installed on the testing box :) 2009-12-10T09:34:53 ii libmemcached-dev 0.31-1 Development files for libmemcached 2009-12-10T09:34:56 ii libmemcached2 0.31-1 A C and C++ client library to the memcached 2009-12-10T09:36:54 i have no idea which of the two is better/more desirable 2009-12-10T09:39:05 lame, you have to install memcached to build libmemcached. someone's been taking lessons from us... ;) 2009-12-10T09:40:02 i suppose that's part of the makefile anyway, though 2009-12-10T09:42:10 *** dbs has joined #evergreen 2009-12-10T09:46:03 berick pasted "libmemcached install" at http://paste.lisp.org/display/91877 2009-12-10T09:51:01 berick pasted "./configure error" at http://paste.lisp.org/display/91878 2009-12-10T09:51:14 if anyone has any thoughts... ^-- 2009-12-10T09:54:31 * dbs looks 2009-12-10T09:55:00 berick: that would be for debian I imagine? 2009-12-10T09:55:07 dbs: yep, lenny 2009-12-10T09:55:33 ah, just got phasefx's laters 2009-12-10T09:55:58 ah, didn't see those 2009-12-10T09:56:05 squeeze will be out by the time we roll another opensrf release anyway, right? 2009-12-10T09:57:06 your paste looks good, berick; I'll replicate it on a test lenny system and tweak the Makefile, obviously 2009-12-10T09:58:17 or... could try to make opensrf work with either libmemcache and libmemcached, depending on what configure says. which would probably be best. 2009-12-10T10:01:58 just requested a code import for OpenSRF trunk into launchpad. this would have been a nice feature branch. 2009-12-10T10:04:57 dbs: should PKG_CHECK_MODULES in configure.ac be AC_CHECK_LIB instead? 2009-12-10T10:05:14 nope 2009-12-10T10:05:30 *** lisppaste3 has quit IRC 2009-12-10T10:05:49 you have pkg-config installed? 2009-12-10T10:06:00 ah 2009-12-10T10:06:06 lemme check 2009-12-10T10:06:19 *** Stuart has joined #evergreen 2009-12-10T10:06:19 http://coffeecode.net/archives/208-PKG_CHECK_MODULES-syntax-error-near-unexpected-token-DEPS,.html 2009-12-10T10:07:16 you'll need to rerun autogen.sh (the autoconf one) after that's installed 2009-12-10T10:07:37 heh, just come to that realization.. 2009-12-10T10:07:49 Hi, well I got Evergreen installed apart from not being able to login via srfsh I get the following message Unable to communicate with service open-ils.auth. Any ideas please. 2009-12-10T10:08:02 berick-- was installing pkg-config, then got sidetracked by lack of other packages 2009-12-10T10:08:10 thanks, dbs 2009-12-10T10:08:30 berick: "thanks, dbs" -- for screwing up a perfectly good build? :) 2009-12-10T10:08:35 *** lisppaste3 has joined #evergreen 2009-12-10T10:09:29 dbs: meh, screwing up perfectly good builds is a pastime of mine 2009-12-10T10:10:43 and I taught both of you everything you know about the sport 2009-12-10T10:12:23 yay, my dev brick is back in business 2009-12-10T10:12:54 Stuart: probably an error during services startup. check the osrfsys.log file for errors 2009-12-10T10:13:58 Stuart: did your run osrf_ctl.sh -l -a start_all ? 2009-12-10T10:14:04 Stuart: also check the hostname in the apps section of opensrf.xml (at the bottom). if you're using "osrf_ctl.sh -l ..." to start up, then it should be "localhost" 2009-12-10T10:14:17 *** [1]atz has joined #evergreen 2009-12-10T10:21:08 Stuart: there's a step by step sledgehammer here: http://evergreen-ils.org/dokuwiki/doku.php?id=troubleshooting:checking_for_errors 2009-12-10T10:23:43 * dbs needs to reboot - back in a sec 2009-12-10T10:23:47 *** dbs has quit IRC 2009-12-10T10:25:53 *** [1]atz has quit IRC 2009-12-10T10:26:09 *** dbs has joined #evergreen 2009-12-10T10:26:29 *** [1]atz has joined #evergreen 2009-12-10T10:29:57 *** [2]atz has joined #evergreen 2009-12-10T10:31:11 *** atz has quit IRC 2009-12-10T10:31:11 *** [2]atz is now known as atz 2009-12-10T10:36:59 *** [2]atz has joined #evergreen 2009-12-10T10:41:40 *** atz has quit IRC 2009-12-10T10:41:40 *** [2]atz is now known as atz 2009-12-10T10:47:54 *** [1]atz has quit IRC 2009-12-10T10:56:31 *** zed2 has joined #Evergreen 2009-12-10T10:57:47 Hello, I'm having the import problem with Czech letters in direct_digest.pl. How can I add the translation for the last letter in this name Štanský 2009-12-10T10:59:09 *** Stuart has quit IRC 2009-12-10T10:59:43 *** Meliss has joined #Evergreen 2009-12-10T11:00:16 It comes out in the bre format as ©ư which direct_ingest.pl does not like 2009-12-10T11:10:27 *** Meliss has quit IRC 2009-12-10T11:26:29 zed2 - there has been some discussion about this on the list recently 2009-12-10T11:27:50 You could try changing the entityize($xml) call in marc2bre.pl to entityize($xml, 'D') and see if that makes a difference 2009-12-10T11:34:36 zed2: also did some testing - running the input to marc2bre.pl through the NFD function of Unicode::Normalize also works, though we haven't tried with Czech specifically 2009-12-10T11:36:32 gmcharlt: that's what adding the 'D' parameter to entityize does, although dbwells has some records where NFD breaks things that work with NFC 2009-12-10T11:36:58 yup 2009-12-10T11:37:44 marc2bre.pl is already using the NFD, but I have somewhere more specific to look now 2009-12-10T11:48:58 I'll give {something} to the person that catches the NF-related error and re-runs the failed call after changing the NF of the record, so that whatever works is used 2009-12-10T11:49:18 in marc2[ab]re and direct_ingest 2009-12-10T11:49:40 well, if I turn off NFD and use NFC, it works for this record. 2009-12-10T11:50:14 zed2: and the reverse is true for other characters, unfortunately :( 2009-12-10T11:50:44 The current character is C6B0 NFD converts to ©ư NFC to ©ư 2009-12-10T11:50:54 *** brendan_ga has quit IRC 2009-12-10T11:51:22 I'm not sure where wlayton is on his "destroy biblio_fingerprint.js" approach, I didn't quite understand what he was saying last night 2009-12-10T11:51:46 dbs: I think he ported the algo in the js to perl 2009-12-10T11:52:29 miker_: yeah, that's where he was heading. but when he said he put biblio_fingerprint.js into Ingest.pm, I got confused. I blame libmemcache* 2009-12-10T11:53:12 dbs: I think that's what he meant by that ... or, that's what I read 2009-12-10T11:54:46 zed2: http://zvon.org/other/charSearch/PHP/search.php says those are both correct, if {LATIN SMALL LETTER U WITH HORN} is the intended display character 2009-12-10T11:54:47 *** wlayton has joined #evergreen 2009-12-10T11:54:58 so ... it's all libmozjs's fault 2009-12-10T11:55:02 as usual 2009-12-10T11:55:23 hey all 2009-12-10T11:55:59 wlayton: speak of the devil! 2009-12-10T11:56:05 just to quickly clarify, I've tried _porting_ the code from biblio_fingerprint.js into Perl 2009-12-10T11:58:55 I'm far from a unicode expert, but from what I could tell, the JS code would chop off the most significant byte from a multi-byte character (whereas Perl doesn't) 2009-12-10T11:59:46 I have the fingerprint code in Perl and our records are now being run through direct_ingest.pl without issue. 2009-12-10T12:00:11 wlayton: well, it's more subtle than that ... it seems to be a problem with the way the data is being passed from perl to js 2009-12-10T12:00:46 because the MARC editor in the staff client handles any unicode, regardless of encoding, just fine 2009-12-10T12:01:10 well, regardless of the NF that the data is passed in 2009-12-10T12:01:34 miker_: I was definitely worried that I was throwing the baby out with the bath water... 2009-12-10T12:02:27 working code wins, and in practice I don't think anyone (other than me) has actually made use of biblio_fingerprint.js, as dbs has suggested 2009-12-10T12:03:05 and, more to the point, it's motivation for me to finish the expunging of open-ils.ingest via in-db ingest 2009-12-10T12:03:14 more to the point for me, anyway ;) 2009-12-10T12:03:18 throw out the baby! throw out the baby! it's keeping me up at night 2009-12-10T12:03:59 I can submit the code to the list for comment tonight. It's not very pretty, though... 2009-12-10T12:04:18 go for it 2009-12-10T12:04:21 wlayton++ 2009-12-10T12:04:53 *** sfortin has quit IRC 2009-12-10T12:05:35 it would be awesome if we had a QA team with a good set of sample data to run through the import process before and after to compare the results. I guess I should get a QA team proposal out there 2009-12-10T12:08:28 ... after which, a horde of QA team volunteers will magically do all of the work :) 2009-12-10T12:08:55 *** brendan_ga has joined #evergreen 2009-12-10T12:09:13 *** wlayton has quit IRC 2009-12-10T12:30:37 I guess I have 2 records that don't work for NFD or NFC 2009-12-10T12:31:50 *** plogk has joined #evergreen 2009-12-10T12:34:10 zed2: can you send those to the open-ils-dev mailing list to add to our sets of test data? 2009-12-10T12:40:27 okay. libmemcached now installs from source on everything except karmic 2009-12-10T12:41:23 * dbs wonders if we should strip out the gutsy & intrepid install targets 2009-12-10T12:41:40 Will do, I'll try to figure out the exact characters first. 2009-12-10T12:53:37 *** sfortin has joined #evergreen 2009-12-10T12:54:50 dbs: that bzr diff and patch -p0 magic worked well; I'm going to script it 2009-12-10T12:55:25 im wondering if someone might be able to help me out with some json grammar? I have the following query. I doesn't want to apply the group by clause. 2009-12-10T12:55:27 my $QUERY = {select => {ahr => ['usr']}, from => 'ahr', where => {capture_time => {between => [$start_date, $end_date]}},group_by => {ahr => ['usr']}}; 2009-12-10T12:55:58 * phasefx_ did bzr revert on the "merge tips" in his bzr-svn checkout, "though I imagine I could just let them accumulate. I'll edit trac tonight to hide those property changes. If we ever upgrade astinus to lenny, we'll get subversion 1.5 where those things will be invisible in any case 2009-12-10T12:59:01 dbs++ # magic QA hordes! 2009-12-10T13:03:49 plogk: hey, that's greek to me, but if you paste the SQL getting generated from that (perhaps at http://paste.lisp.org/new/evergreen), it may light some bulbs in here 2009-12-10T13:04:47 *** bshum has joined #evergreen 2009-12-10T13:19:26 phasefx: good idea, i'll play around with it abit more and post there if i can't make any headway. 2009-12-10T13:22:21 *** pmplett has joined #evergreen 2009-12-10T13:25:17 *** eguest309 has joined #Evergreen 2009-12-10T13:25:37 *** artunit has quit IRC 2009-12-10T13:28:39 *** jenny has quit IRC 2009-12-10T13:33:04 plogk: there is no expicity group by operator for json_query. adding distinct=>1 as a peer to select will do what you want, though 2009-12-10T13:34:09 miker_: that works, thx! 2009-12-10T13:34:20 the GROUP BY clause is calculated when you use at least one aggregate function in your select list, or when you ask for "distinct" 2009-12-10T13:38:54 miker_: thats good to know, thx.. I just came across the JSONTutorial.xml . Very handy! 2009-12-10T13:40:14 ahh, good 2009-12-10T13:41:14 there's a version on the project wiki covering the syntax available in EG 1.4 and 1.6, and it's advanced a good bit in trunk 2009-12-10T13:48:04 do you have any links? This was the only one i've stumbled across: http://www.open-ils.org/dokuwiki/doku.php?id=documentation:technical:jsontutorial 2009-12-10T14:14:24 *** brendan_ga has quit IRC 2009-12-10T14:18:49 plogk: that is indeed the one I'm talking about 2009-12-10T14:20:29 it's since gained several variations on the subselect theme, table alias contol and a few other niceties 2009-12-10T14:31:17 <_bott_> /me recalls a "Evergreen / SQL 101" program for EG2010 discussed here at some point and thinks it would be a wonderful submission for someone 2009-12-10T14:31:45 nudge, nudge. eh? 2009-12-10T14:31:54 <_bott_> Was is subtle enough? 2009-12-10T14:32:05 NO! I require a softer touch! 2009-12-10T14:32:14 dec 15, right? 2009-12-10T14:32:21 <_bott_> Yep 2009-12-10T14:32:30 <_bott_> but it only leaves you with 15 shopping days... 2009-12-10T14:32:48 <_bott_> ugh, that new math. That would be 10 2009-12-10T14:32:51 k ... might be good as a pre- or post- conf 2009-12-10T14:33:58 *** jenny has joined #evergreen 2009-12-10T14:36:44 *** sfortin has quit IRC 2009-12-10T15:06:35 *** sfortin has joined #evergreen 2009-12-10T15:06:54 *** bshum has quit IRC 2009-12-10T15:12:39 *** brendan_ga has joined #evergreen 2009-12-10T15:17:27 *** natschil has quit IRC 2009-12-10T15:18:20 nice job today berick 2009-12-10T15:18:38 *** stevenyvr has joined #evergreen 2009-12-10T15:22:09 berick++ # for whatever he did :) 2009-12-10T15:22:11 thanks brendan_ga. hope i didn't mumble too much, i was kind of nervous for some reason 2009-12-10T15:22:19 * phasefx_ needs a sanity check. "I have some interfaces with date filters, a start date and an end date, that ultimately generate SQL that look like: 2009-12-10T15:22:31 * berick is not on his A game today 2009-12-10T15:22:36 berick++ 2009-12-10T15:22:37 date BETWEEN '2008-12-10' AND '2009-12-09T23:59:59-05:00' 2009-12-10T15:23:05 what's worrying me is the timezone stuff 2009-12-10T15:23:58 phasefx_: why is it worrying you? 2009-12-10T15:24:28 miker_: because I was mixing things in my head.. events like fine generation and the like, whereas we're just concerned with information retrieval 2009-12-10T15:25:32 well, that BETWEEN clause is borken anyway ... the start is after the end 2009-12-10T15:25:44 2008 and 2009 2009-12-10T15:26:04 ah 2009-12-10T15:26:08 doh, of course 2009-12-10T15:26:15 okay, I think I know where my cognitive dissonance is happening.. what I'm really interested in is Calendar days 2009-12-10T15:26:33 does 23:59:59 make sense there? 2009-12-10T15:26:59 what you probably want is just to send 2009-12-10 as the second date 2009-12-10T15:27:12 or rather, should Calendar Days encompass all timezones 2009-12-10T15:27:19 that sounds better to me 2009-12-10T15:27:25 2009-12-10 2009-12-10T15:27:32 or, date > x AND date < y 2009-12-10T15:27:43 rather, date >= x AND date < y 2009-12-10T15:29:33 in cstore (direct or json query) => {'-and':[{'date':{'>=':'2008-12-10'}},{'date':{'<':'2009-12-10'}}]} 2009-12-10T15:30:13 and that's preferable over between? 2009-12-10T15:30:23 well, it's exactly what you want 2009-12-10T15:30:25 *** eguest309 has left #Evergreen 2009-12-10T15:30:40 between will match 2009-12-10T00:00:00 2009-12-10T15:30:47 ah 2009-12-10T15:30:48 between is inclusive 2009-12-10T15:31:03 miker_++ # sanity restored 2009-12-10T15:32:08 * berick is still trying to figure out what's wrong with the first approach 2009-12-10T15:32:46 * phasefx_ needs to make a function that can increment a javascript date object by one day 2009-12-10T15:33:22 berick: because if you're collecting data from multiple OUs in different time zones, you can miss (or accidentally include) data ... 2009-12-10T15:33:33 or, I think that's phasefx_'s concern 2009-12-10T15:33:36 yeah 2009-12-10T15:33:49 though, if you're talking about a year worth of data... it's not a big deal :) 2009-12-10T15:33:54 a day, maybe 2009-12-10T15:34:12 an hour, for sure, but then you've got full timestamps for both and the problem goes away 2009-12-10T15:34:42 you know these folks will use interfaces like this for ad-hoc reporting 2009-12-10T15:34:50 not following. The conern is a payment for a branch that happened at 1am my time, but midnight for some other branch? 2009-12-10T15:35:05 and i want to see it, even though it's technically outside my selected time window? 2009-12-10T15:35:49 the widgets they have are datepickers, so I don't think they'll be thinking about timezones at all 2009-12-10T15:35:58 why should they? 2009-12-10T15:36:26 ohh the problem is relying on the server time zone 2009-12-10T15:36:35 instead of passing the client's timezone as part of the command 2009-12-10T15:36:37 that was where my sanity wavered.. they're not thinking about it, and I was worried it might affect them and what data got returned 2009-12-10T15:36:49 if they just have date pickers, then you need to send dates ... and that means either adding a day to BETWEEN, or using the -and mechanism 2009-12-10T15:36:50 *** StephenGWills has joined #evergreen 2009-12-10T15:37:54 or, cast what you're calling "date" here (which is surely a timestamptz) to a real date, and use BETWEEN 2009-12-10T15:38:09 so ... options abound ;) 2009-12-10T15:38:14 right, word 2009-12-10T15:38:54 widget.serialize(widget.value).split('T')[0] 2009-12-10T15:39:00 I'll keep between (easier to read), and just use the .value from datepicker elements (which returns YYYY-MM-DD), and increment the day for the second date 2009-12-10T15:39:09 should get you the date only for those dojo date pickers (was just fooling with those too) 2009-12-10T15:39:29 oh, or .value then :-) 2009-12-10T15:39:33 xul datepicker in this case. .value gets me YYYY-MM-DD, and .dateValue gets me a Date object 2009-12-10T15:39:40 ah 2009-12-10T15:39:49 *** jenny has quit IRC 2009-12-10T15:40:28 though I was formatting the second date as iso8601 when I was trying to include the 23:59:59 madness 2009-12-10T15:42:17 *** sfortin has quit IRC 2009-12-10T15:42:52 looks like I can pass in any value to Date's .setDate() and it'll wrap into subsequent months as needed 2009-12-10T15:43:11 koha had this multi-TZ problem also... and we basically set up different apache vhosts in each TZ needed for a consortium. 2009-12-10T15:43:22 so doing a.setDate(29) when a is in february will go to march 2009-12-10T15:43:57 phasefx_: heh ... musta been talking to mysql ab when they wrote that... 2009-12-10T15:44:00 * miker_ ducks 2009-12-10T15:44:32 * phasefx_ grins, "No, I'm sure mysql would have truncated it to 28 2009-12-10T15:45:05 so then each client just connects to the vhost appropriate for their TZ... so that was easy enough for staff 2009-12-10T15:45:21 but it got to be trickier w/ patron stuff 2009-12-10T15:51:56 miker_: what the heck you ducking for? you think somebody is going to defend mysql? 2009-12-10T15:52:50 *** sfortin has joined #evergreen 2009-12-10T15:56:25 well, there is atz over there... 2009-12-10T15:56:32 * miker_ really ducks now 2009-12-10T16:02:36 *** dbwells_ is now known as dbwells 2009-12-10T16:06:35 too much? 2009-12-10T16:10:30 * dbs is busy folksonomically tagging #code4lib nicks for later easier retrieval 2009-12-10T16:10:59 wisdom of the dbs 2009-12-10T16:11:28 miker_: refresh my memory... joins in a pcrud search clause.. supported? 2009-12-10T16:12:38 Hello all. I have hopefully a simple question about billing types. Are the first 9 default types system-level types, and therefore not editable or selectable in most normal cases? 2009-12-10T16:12:44 *** pmplett is now known as pmp_afk 2009-12-10T16:14:24 dbwells: we reserve the first 100 for system-use, and are explicitly not offering them in the bill patron dialog 2009-12-10T16:14:26 dbwells: I believe that's what we found. They're applied by the system (e.g. when a book becomes lost the corresponding billing type is automatically applied). 2009-12-10T16:15:07 first 100 = .id <= 100 2009-12-10T16:15:21 the sequence should start above that for user added billing types 2009-12-10T16:15:56 phasefx_: dbs: thanks guys. Just making sure my experience was the intended reality. 2009-12-10T16:16:16 *** jmeeuwen_ has joined #evergreen 2009-12-10T16:16:18 *** jmeeuwen has quit IRC 2009-12-10T16:22:37 dbwells: how's your migration going? 2009-12-10T16:22:52 if you're too busy to talk, I would certainly understand :) 2009-12-10T16:23:50 dbs: It has begun in earnest this morning. Around noon we gave the "no more edits to anything" notification to our staff. 2009-12-10T16:24:36 dbwells++ # good luck! 2009-12-10T16:24:45 dbs: We are migrating bib and item data over the weekend to hopefully allow cataloging operations to resume on Monday. 2009-12-10T16:24:56 sounds very familiar. 2009-12-10T16:25:23 dbs: Patron and circ stuff will be next weekend. Assuming all goes as planned. 2009-12-10T16:25:28 * dbs recalls composing a very very bleary message at 3:00-ish AM on Monday morning - hopefully all goes smoothly for you :) 2009-12-10T16:27:32 dbs: thanks. I think we are probably over-prepared, but you never know! We have done several partial test migrations and one full test. We did circ training today on the test instance and didn't have *too* many problems. 2009-12-10T16:29:56 *** sfortin has quit IRC 2009-12-10T16:30:01 back for a moment to billing types, I have another question then about bug handling conventions. 1.6 has a minor bug in that manual billing doesn't work until you add at least one non-system billing type. This has been fixed in trunk (a 'Misc' type with ID 100 is in the seed file). How are these types of bugs best reported? 2009-12-10T16:31:26 Somewhat ironically, I think this was just asked about on the list! 2009-12-10T16:32:55 if no one is inspired to correct it immediately (after you mentioning or after seeing the recent emails), I think a bug in Launchpad would be great, or maybe an amendment to the upgrade/install instructions 2009-12-10T16:32:59 bugs.launchpad.net/evergreen 2009-12-10T16:33:34 *** jenny has joined #evergreen 2009-12-10T16:35:01 *** dchristens has joined #evergreen 2009-12-10T16:35:17 dbs: phasefx_: ok, thanks again guys. I wasn't sure there was any special way to report a "bug already fixed in trunk" type of bug. Maybe once I check out launchpad a status like that will already exist. 2009-12-10T16:35:45 I will also reply to Ryan on the list if nobody else is yet. 2009-12-10T16:36:05 my reply should show up eventually :-/ 2009-12-10T16:36:39 but if you want to reference the bug number, that might be a good thing to do 2009-12-10T16:36:59 phasefx_: cool, thanks. I am 9MB of my pathetic 300MB email quota, so I need to do some housekeeping before I send anything :( 2009-12-10T16:38:30 Hi folks. Finally freed up some time again from $other_work to get back to learning Evergreen :-D Happy day! 2009-12-10T16:39:01 (well... for me - you'll probably get sick of my questions :-) 2009-12-10T16:39:01 yay 2009-12-10T16:39:22 yay :-} 2009-12-10T16:39:26 heh 2009-12-10T16:39:47 dchristens++ # good to see you back 2009-12-10T16:40:08 Hey dbs - good to be back! How goes? 2009-12-10T16:40:40 *** gdunbar has quit IRC 2009-12-10T16:40:41 fantastic! although we're still not at 1.6.0.0 official yet. looking forward to getting there eventually. 2009-12-10T16:41:02 dchristens: you see the Table of Contents stuff I added to your OpenLibrary plugin? 2009-12-10T16:41:34 Hrm... feels close. I'm still at 1.4.0.4; should I wait to upgrade? 2009-12-10T16:41:54 dbs++ ToC - sweet! 2009-12-10T16:42:28 "we're" meaning Conifer, 1.6.0.0 is officially out 2009-12-10T16:43:02 Appears I've got some catching up to do.... 2009-12-10T16:53:01 OK I've traced my record conversion problem to getting the 'fingerprint' field 2009-12-10T16:53:28 How do I turn on all of the debugging code that is already in biblio_fingerprint.js 2009-12-10T16:59:50 zed2: sorry that we didn't link you to the previous conversation, but we knew that... :( 2009-12-10T17:00:33 zed2: thread started here: http://markmail.org/message/gbmdb3apj6x374dr 2009-12-10T17:01:51 dbs: ok, the OpenLibrary table of contents thing is very, very cool. :-) 2009-12-10T17:06:38 dchristens: hey, you started it! 2009-12-10T17:06:40 * dbs heads home 2009-12-10T17:06:48 *** dbs has quit IRC 2009-12-10T17:10:53 "my" cataloguers want to keep track of both an item's list-price and discount-price (aka what they actually paid). Would it make more sense to add a new field to asset.copy, or monkey around with stat cats? Or is there another option? 2009-12-10T17:13:57 dchristens: copy stat cats can't be freeform. we have placed our "cost" data in a copy note with title "Cost" 2009-12-10T17:14:29 dchristens: but we would prefer to see the schema extended. there was some talk about that maybe-happening as part of Acq work, but I haven't inquired about it post-migration. 2009-12-10T17:14:49 dchristens: and in my case, it isn't out catalogers but our auditors. 2009-12-10T17:15:42 dchristens: (and to expand on my stat cat statement, "can't be freeform" means you'd have to configure a value for every possible cost.) 2009-12-10T17:15:55 jeff: thanks. heh - I think the cataloguers are asking me because the auditors are asking *them* :-) I'll see if I can convince them that a copy note will do the trick - thanks again 2009-12-10T17:16:37 dchristens: then reporting on that later can be tricky, mostly if they aren't consistent with the copy note titles. 2009-12-10T17:16:49 like "Cost " and "cost" and " COST ", etc. 2009-12-10T17:17:18 (not quite sure why my brain insists on thinking that stat cats *are* free-form. D'oh.) 2009-12-10T17:17:26 patron stat cats are. 2009-12-10T17:17:30 copy stat cats are not. 2009-12-10T17:17:34 aha 2009-12-10T17:17:46 notes sux0r :) that's why they're de-emphasized. I'd like to see a toggle for patron and item stat cats for freeform vs controlled 2009-12-10T17:18:11 heh 2009-12-10T17:21:30 alternately or additionally, maybe some non-controlled stat cats where a datatype can be enforced (numeric only, for example) 2009-12-10T17:21:54 if I were to (locally) modify the asset.copy schema, I'd then need to make corresponding changes to the edit-item-attribute screen... I'm a little worried about taking our installation off the beaten track... 2009-12-10T17:22:26 phasefx++ yep, that would do it :-) 2009-12-10T17:22:48 * dchristens looks around for his magic wand.... 2009-12-10T17:24:51 * phasefx_ can think of kludgy options as well. "but throwing by things into notes, you can at least migrate the data to a better scheme later. Maybe run strict recurring maintenance reports to check for inconsistent use of those notes 2009-12-10T17:25:16 * phasefx_ twists his head sideways and rereads that sentence 2009-12-10T17:26:27 one major downside with copy notes at the moment, you can't create them in batch during the same step as item creation 2009-12-10T17:26:47 heh - enforce policy by making the record creator fix the mistakes over and over until they learn.... 2009-12-10T17:27:41 make all data fields work as tags/folksonomy.. converge on correctness through wisdom of the crowds ;) 2009-12-10T17:27:50 :-) 2009-12-10T17:27:59 *** r123 has quit IRC 2009-12-10T17:28:00 tag: overdue tag: not overdue tag: really 2009-12-10T17:31:07 Hmm. What is "deposit_amount" used for in asset.copy? Could I, er, "repurpose" that? 2009-12-10T17:33:21 *** gmcharlt has quit IRC 2009-12-10T17:33:38 it serves double-duty at the moment, for deposits and rentals. Could be better labeled as checkout fee or something. If the deposit flag is set, then the patron gets the money back afterward 2009-12-10T17:33:54 so no, you can't repurpose it 2009-12-10T17:34:04 in the past, you could have 2009-12-10T17:34:07 :-) 2009-12-10T17:35:51 I've been trying to think of sneaky ways someone could use child tables to extend the schema in a good way, but.. :-/ 2009-12-10T17:36:21 a child table inherits all the columns from the parent table, and can have columns of its own 2009-12-10T17:36:36 and you can reference parent rows through the child table 2009-12-10T17:37:03 oooh.... cool. 2009-12-10T17:37:11 so, for example, we could extend asset.copy with asset.copy_local, and tweak the fm_IDL.xml file to use that table instead 2009-12-10T17:37:20 but I'm not sure what it's buying us 2009-12-10T17:37:44 future-proofing? 2009-12-10T17:38:13 but if you're doing upgrades, we're never going to blow away your tables anyway.. so you just have to watch out for field name collisions, which this doesn't solve 2009-12-10T17:38:29 maybe manageability 2009-12-10T17:39:00 your customizations are table definitions instead of alter table statements 2009-12-10T17:39:20 I like that 2009-12-10T17:39:53 if something blows away your custom fm_IDL.xml, your data is still there and safe 2009-12-10T17:42:46 I think I'll add the field+tweak edit-item-attribs, and leave myself a REALLY BIG NOTE. 2009-12-10T17:42:55 Thanks for going 'round this with me :-) 2009-12-10T17:43:11 another thought.. you could add a trigger to the copy notes table to look for certain title values and enforce rules based on that 2009-12-10T17:43:34 oooh... yep 2009-12-10T17:44:19 I'd recommend a prefix for your local columns 2009-12-10T17:44:40 phasefx++ 2009-12-10T17:44:49 will do. Thanks again 2009-12-10T17:45:15 copy editor is pretty hackish, hope it doesn't give you pain :) 2009-12-10T17:45:20 I'll mull this over overnight, before I do anything drastic :-) 2009-12-10T17:45:51 *** dchristens has left #evergreen 2009-12-10T17:46:40 *** zed2 has left #Evergreen 2009-12-10T17:47:44 @later tell dchristens example here of tweaking the copy editor and item lists: http://svn.open-ils.org/trac/ILS/changeset/15043/ 2009-12-10T17:47:44 phasefx_: The operation succeeded. 2009-12-10T17:49:02 @later tell dchristens though that's for a bool, so not the best example 2009-12-10T17:49:02 phasefx_: The operation succeeded. 2009-12-10T17:55:26 *** jenny has left #evergreen 2009-12-10T19:32:53 *** pmp_afk is now known as pmplett 2009-12-10T20:12:48 gah ... sorry berick. yes, cstore searches allow joins 2009-12-10T20:13:14 in the "extra" hash, add a "join" key, and stick in that what you would for json_query 2009-12-10T20:41:06 *** stevenyvr has left #evergreen 2009-12-10T21:38:37 *** dbs has joined #evergreen