2012-01-11T00:21:17 *** PaintedBlue has joined #evergreen 2012-01-11T00:35:08 *** PaintedBlue has left #evergreen 2012-01-11T01:50:30 *** dbs has quit IRC 2012-01-11T02:09:12 *** mrpeters-isl has quit IRC 2012-01-11T03:22:58 *** mcmx has joined #evergreen 2012-01-11T03:24:09 Hi all, i was in here yesterday trying to get evergreen up and running and i had to go, however we establised that my database tables havent been created properly. Any advice on how to fix this? TIA 2012-01-11T03:32:22 *** mcmx_ has joined #evergreen 2012-01-11T03:35:35 *** mcmx has quit IRC 2012-01-11T03:53:01 *** mcmx_ has quit IRC 2012-01-11T06:59:22 *** tfaile has joined #evergreen 2012-01-11T07:30:28 *** dbs has joined #evergreen 2012-01-11T07:41:44 *** collum has joined #evergreen 2012-01-11T08:10:45 *** akilsdonk has joined #evergreen 2012-01-11T08:26:55 *** Dyrcona has joined #evergreen 2012-01-11T08:27:30 *** dbs has quit IRC 2012-01-11T08:31:36 *** _bott_ has quit IRC 2012-01-11T08:32:37 *** _bott_ has joined #evergreen 2012-01-11T08:38:14 *** kmlussier has joined #evergreen 2012-01-11T08:39:18 *** bshum_ has joined #evergreen 2012-01-11T08:39:53 *** bshum has quit IRC 2012-01-11T08:39:53 *** bshum_ is now known as bshum 2012-01-11T08:48:16 *** gdunbar has quit IRC 2012-01-11T08:55:47 *** mrpeters-isl has joined #evergreen 2012-01-11T09:00:14 *** bwicksall has joined #evergreen 2012-01-11T09:02:11 file this under "questions I have asked in this channel before but can't remember find the answer", but... 2012-01-11T09:02:26 what is the function of the OUS "charge lost on zero"? 2012-01-11T09:02:53 *** Meliss has joined #evergreen 2012-01-11T09:09:05 csharp: If the price of an item is 0, the default price that should be charged if it's marked lost. 2012-01-11T09:09:18 *** dzeiger has joined #evergreen 2012-01-11T09:09:21 csharp: it's a setting meant to solve a no-longer-present design constraint on asset.copy. it was not possible before (2.0?) to have a NULL copy price 2012-01-11T09:09:37 so there was no way to differentiate between no price and $0, except for the org setting 2012-01-11T09:09:44 kmlussier: berick: thanks! 2012-01-11T09:09:58 but, now you can have null prices, which is the prefered method of indicating "no price is set" 2012-01-11T09:11:09 berick: And yet libraries still set 0 when they don't "know" the price. Thus the setting is still useful. <_< 2012-01-11T09:12:22 tsbere: sure 2012-01-11T09:18:32 *** dbs has joined #evergreen 2012-01-11T09:47:33 *** jenny has joined #evergreen 2012-01-11T10:04:06 *** dzeiger has quit IRC 2012-01-11T10:20:41 *** dzeiger has joined #evergreen 2012-01-11T10:29:29 fixing https://bugs.launchpad.net/evergreen/+bug/914800 -- i seem to remember a JavaScript file, in addition to ils_events.xml where I might need to map this. Or am I dreaming that? 2012-01-11T10:29:31 Launchpad bug 914800 in Evergreen "Attempting to merge user in collections provides an unfriendly error message" (affected: 1, heat: 6) [Low,New] 2012-01-11T10:31:51 mrpeters-isl: Well, wherever the javascript needs to know about it. But I don't add everything to ils_events.xml these days. The system can handle *not* having entries there now ;) 2012-01-11T10:32:48 in master? 2012-01-11T10:32:55 so should i bother with this? 2012-01-11T10:33:14 * tsbere isn't even sure that adding things to ils_events.xml even *stops* those errors, instead of just changing them a little, as the error is *still* unexpected 2012-01-11T10:34:00 mrpeters-isl: What I am saying is "I think the fix is to go to the code that is throwing the failure and making it check for and report on MERGED_USER_IN_COLLECTIONS instead of throwing a network failure" 2012-01-11T10:35:19 yeah. i've done this before i know, i just can't find the old commits. 2012-01-11T10:35:21 *** jenny has quit IRC 2012-01-11T10:36:46 *** jenny has joined #evergreen 2012-01-11T10:37:18 searching git is absolutely miserable 2012-01-11T10:38:54 there are some tricks that make it less miserable. what are you trying to find? 2012-01-11T10:39:13 i know i've commited fixes to ils_events.xml and i just can't find any of them 2012-01-11T10:39:27 for example, i know i created the event for ACTOR_USER_DELETE_OPEN_XACTS 2012-01-11T10:40:02 mrpeters-isl: git blame ils_events.xml could help 2012-01-11T10:40:51 on a line-by-line basis, it tells you the author, date, and hash of the commit that last updated the line 2012-01-11T10:40:51 * tsbere had previously tried adding to ils_events.xml, got no better behavior, and then ditched them for handling code where the problem was, and thus is under the assumption that ils_events.xml doesn't do much for stopping errors 2012-01-11T10:41:10 found it 2012-01-11T10:41:11 http://git.evergreen-ils.org/?p=working/Evergreen.git;a=commitdiff;h=5f8445bb44e0493e7277a1467e8dfb5530cd4950 2012-01-11T10:44:26 this one doesnt look to be set up for handling the error at all though...at least it doesn't look like it 2012-01-11T10:46:17 mrpeters-isl: It is. The change to display.js does that check, though I usually just have things use the textcode instead of creating new ils_events.xml entries. 2012-01-11T10:46:57 * tsbere doesn't like all the hard-coded IDs for these events that have a theoretically unique ID in the form of their name 2012-01-11T10:47:05 ok. i'm interested in seeing how else i can do this then. 2012-01-11T10:47:56 do you have any past examples? 2012-01-11T10:51:18 mrpeters-isl: I have had examples in the past of "assuming that we *have* an ilsevent number breaks things" 2012-01-11T10:53:34 mrpeters-isl: I think ilsevent is set to '' (empty string) when there is no numeric code, which is defined and thus matches most "if it is defined...." checks. Using the textcode variable and the name of the event instead of the ilsevent and it's number then lets you do the same check. 2012-01-11T10:56:28 I'm sorry man, but I'm not following you at all. 2012-01-11T10:57:04 I don't understand how else I'm going to throw a clear error message without using ils_events.xml. That's the only way I've ever seen it done. If you have an example of how to do it another way, then maybe I can try that. 2012-01-11T10:57:09 in the past, all events were defined in that ils_events.xml file, but then we started creating some events on the fly in the code, and the staff client wasn't accomodating that originally 2012-01-11T10:57:24 mrpeters-isl: Instead of $blah->ilsevent and checking it is ####, check if $blah->ilsevent is defined. If so, check if $blah->textcode is the erroring textcode. 2012-01-11T10:58:39 mrpeters-isl: as far as I understand it, the only advantage to having an event in the ils_events.xml file is that you can leverage I18N for it, instead of doing the I18N in the code that receives the event 2012-01-11T10:58:43 mrpeters-isl: In the case of your example there with ACTOR_USR_DELETE_OPEN_XACTS, the ilsevent description isn't used in the error message. The *only* thing ilsevent inclusion gets you is the ilsevent *number*. It is using a language string for the error in disaply.js 2012-01-11T10:59:07 for proper spellings if display.js 2012-01-11T10:59:09 anyway 2012-01-11T10:59:11 <_< 2012-01-11T11:00:00 mrpeters-isl: Thus, instead of "switch(Number(robj.ilsevent))" you could just "switch(robj.textcode)" and look for "MERGED_USER_IN_COLLECTIONS", to the best of my knowledge. 2012-01-11T11:00:25 the Override prompt dialog does show the description for events 2012-01-11T11:00:38 Which is maybe the only useful place for them, then :P 2012-01-11T11:00:39 so I guess there's that 2012-01-11T11:00:53 * tsbere doesn't like that being in an XML either, thinks it should live in the DB 2012-01-11T11:01:03 if it is going to be kept around, anyway 2012-01-11T11:02:29 the cmd_patron_merge function looks nothing like cmd_patron_delete 2012-01-11T11:02:33 phasefx: I would also remain *much* happier if we just ditched the ilsevent numberic ids. :P 2012-01-11T11:02:45 tsbere: I'm with you 2012-01-11T11:02:49 so i don't see how i can use switch(rojb.textcode) 2012-01-11T11:02:56 natural keys for the win 2012-01-11T11:02:57 * tsbere has fixed code that assumed there was always an ID 2012-01-11T11:03:44 mrpeters-isl: I am providing examples without knowing *where* things are broken. You haven't told me what file/line things are erroring out on, for example, and I haven't tried to figure that one out myself either. 2012-01-11T11:04:18 i guess i assumed the bug does that? i understand if you havent looked at it. 2012-01-11T11:05:26 mrpeters-isl: The bug tells me "this error is being thrown". I don't see anything saying "this is apparently coming from function X in file Y not handling the error" 2012-01-11T11:06:15 I think it's pretty clear Actor.pm is throwing the error. 2012-01-11T11:06:21 from, the bug 2012-01-11T11:06:39 Actor.pm throws events. The error is "the client isn't handling it" 2012-01-11T11:07:10 and to my knowledge ils_events.xml handles the error codes 2012-01-11T11:07:42 but you're telling me it shoudlnt be used. i'm completely confused. i've always seen errors mapped with ils_events. 2012-01-11T11:09:16 mrpeters-isl: ils_events handles certain kinds of error messages 2012-01-11T11:09:19 This is not one of them 2012-01-11T11:09:47 so why don't we handle all of the error codes in one place? 2012-01-11T11:09:53 The *staff client* code that is not expecting that event code needs to be told how to handle it 2012-01-11T11:10:17 mrpeters-isl: Because not all ils_events are errors to begin with, and not every code path gets the same "this is why this is happening" text, so on and so forth 2012-01-11T11:11:15 there's quite a few similar things handled in ils_events, so at some point, this was proper practice 2012-01-11T11:11:50 you say you handle it different, and you don't have any example branches where you've done this that i could look at? 2012-01-11T11:12:38 mrpeters-isl: Let me put it a different way. There are a lot of things in ils_events that, as far as I can tell, *only exist to get ilsevent numeric ids*. I am *against* using numeric IDs when we have *unique names* for these things. The staff client code that handles the specific events tends to call staff client language strings, not the ils_event descriptions, for error messages. 2012-01-11T11:13:33 Thus, unless this needs to show up on an override dialog, where phasefx already mentioned that those descriptions show up, I would argue it doesn't need a numeric ID, but even if you *do* add it to the xml file *special handling code is needed in the staff client* 2012-01-11T11:15:21 ok. this is beyond my skillset then. i need to walk away before i get frustrated. i'll just keep an eye on the bug for a fix if someone has time. 2012-01-11T11:15:36 heh, http://forum.kaspersky.com/lofiversion/index.php/t138878.html (Evergreen, Kaspersky, and receipt printers) 2012-01-11T11:15:57 mrpeters-isl: http://git.evergreen-ils.org/?p=working/Evergreen.git;a=blobdiff;f=Open-ILS/xul/staff_client/server/patron/display.js;h=3eba4f4d4d3333294724841aba0cb4737bfe886e;hp=2e17b8bb1b4e09c4868f8d774331e72e5b3737cd;hb=5f8445bb44e0493e7277a1467e8dfb5530cd4950;hpb=bbc6b3f9fafd0d591ae8b258650bc50ed273be84 2012-01-11T11:16:25 tsbere: yeah, i understand where it needs to go. I just don't know enough javascript to understand what that's doing. 2012-01-11T11:16:58 names > numbers, yes indeed 2012-01-11T11:18:59 *** AaronZ-PLS has joined #evergreen 2012-01-11T11:26:17 *** frank___ has joined #evergreen 2012-01-11T11:27:52 Hi everybody, I'am in trouble with my circulation policies, I have three types, but when I check out a book of each kind of, the systems takes just the default circulation policie, does someone know what the problem is? 2012-01-11T11:36:22 *** graced has joined #evergreen 2012-01-11T11:45:02 *** matt_carlson has joined #evergreen 2012-01-11T11:49:20 *** sal_ has joined #evergreen 2012-01-11T11:58:44 Circulation-type question: Is there a script I can run that will move an item (in my 2-branch test system) from "in transit" (where it seems to be stuck) to whatever status will make it loanable? 2012-01-11T11:59:07 sal_: it has to be checked in at the destination 2012-01-11T11:59:22 Ah. I think I can do that ;-) 2012-01-11T11:59:37 you could probably write a srfsh script to do it 2012-01-11T12:00:10 have a look at reshelving_complete.srfsh 2012-01-11T12:00:35 i'm betting you could write something like that to check in any items with the in transit status 2012-01-11T12:00:50 Thanks. I will. 2012-01-11T12:01:05 (practice with Perl!) 2012-01-11T12:01:44 *** matt_carlson has quit IRC 2012-01-11T12:03:24 *** matt_carlson has joined #evergreen 2012-01-11T12:08:11 frank___: I think folks might need more details 2012-01-11T12:11:12 sal_: On master you could set transit suppression for the entire org tree and no transits will ever be generated normally too 2012-01-11T12:14:59 phasefx: yes, well, I have thre types of users, each one has it´s own circ duration rule, and I created the Circulation Policy Configuration for each one too, for example, I have the user student, it has to have a check out with 7 days and 2 renewal, when I try to check out a book for this user, the due date is for 14 days with 3 renewal, and if I check in the Check duration rule the one that is taken is the default rule 2012-01-11T12:15:50 tsbere: Hmm. On the one hand, I like the idea of having transit for "realism" (hence 2 branches). OTOH, it's a pain to deal with. Will see how annoyed I get ;-) 2012-01-11T12:19:13 sal_: They go away when they reach the other branch and are checked in 2012-01-11T12:20:07 tsbere: that they do. But that means I have to remember to check them in using the other branch's SIP id... 2012-01-11T12:20:18 (or staff client) 2012-01-11T12:20:24 sal_: If this is for testing, well, oh well. Not much you can do there ;) 2012-01-11T12:21:19 tsbere: oh, it's entirely for testing. My "library" consists of a couple of VMs, and my items are barcodes and RFID tags on cardboard. (With cover art ;-) ) 2012-01-11T12:21:55 * tsbere doesn't do RFID testing, but does test barcodes 2012-01-11T12:22:03 My old test barcodes were on actual books. Now I use fake items. 2012-01-11T12:26:03 * dbs wonders how gmcharlt is doing with his governance to-dos 2012-01-11T12:29:23 * sal_ is testing peripheral software against EG, so RFID comes in handy... Trying to teach a British self-checkout system about Evergreen :-) 2012-01-11T12:33:48 dbs: shooting for Friday 2012-01-11T12:35:00 gmcharlt: i'm shooting, but mostly aiming at my own skull 2012-01-11T12:37:32 *** AaronZ-PLS has quit IRC 2012-01-11T12:38:04 dbs: s/for Friday/for Friday so that at least Monday gets hit, not Tuesday morning/ ;) 2012-01-11T12:44:01 *** akilsdonk has quit IRC 2012-01-11T12:49:48 *** kmlussier_ has joined #evergreen 2012-01-11T12:49:48 *** kmlussier has quit IRC 2012-01-11T12:49:49 *** kmlussier_ is now known as kmlussier 2012-01-11T13:03:11 *** graced has quit IRC 2012-01-11T13:09:51 *** graced has joined #evergreen 2012-01-11T13:13:22 *** AaronZ-PLS has joined #evergreen 2012-01-11T13:34:36 "Ain't no fun waitin' 'round for a script to run." 2012-01-11T13:39:44 *** artunit has quit IRC 2012-01-11T13:40:07 better waiting for a script to run than doing it all manually? 2012-01-11T13:40:19 because, hey... you can do something else. :-) 2012-01-11T13:41:11 * csharp recommends Frozen Bubble 2012-01-11T13:42:10 * Dyrcona has been looking at tickets on his consortium's internal ticket system, and rather arbitrarily closing a lot of old ones. 2012-01-11T13:43:13 * Dyrcona needs another dev VM. %) 2012-01-11T13:43:57 * tsbere has 2 dev VMs, one for 2.0 and the other for 2.1+ 2012-01-11T13:44:16 It is nice to close our internal tickets saying: Oh, look, according to Launchpad that code was merged into master. :) 2012-01-11T13:53:59 kmlussier: if you're here, you might try doing "cpan Business::OnlinePayment::PayPal" to solve your problem 2012-01-11T13:54:15 maybe that is still yet to be added to the dependencies? 2012-01-11T13:54:39 mrpeters-isl hasn't seen my reply to his email on the list, yet. :) 2012-01-11T13:54:40 *** matt_carlson has quit IRC 2012-01-11T13:55:23 mrpeters-isl and Dyrcona: Thanks! I'll give it a try. 2012-01-11T13:55:29 ah yep you beat me to it! 2012-01-11T13:55:42 I'm going to add it to the cpan modules to install. 2012-01-11T13:55:48 Dyrcona: i'll sign off 2012-01-11T13:55:53 if needed 2012-01-11T13:57:06 pile on top of https://bugs.launchpad.net/evergreen/+bug/914482 ? 2012-01-11T13:57:07 Launchpad bug 914482 in Evergreen "Make Debian Squeeze use PostgreSQL 9.1 by default" (affected: 1, heat: 6) [Undecided,New] 2012-01-11T13:57:46 * dbs still doesn't know how eeevil didn't get whacked by the 8.4 vs. 9.1 thing, given that libpq-dev was getting installed first 2012-01-11T13:57:53 dbs: maybe, though I do like to keep things separate. 2012-01-11T13:58:25 Dyrcona: yeah. I like keeping things separate, too, except there are a crapload of separate bugs festering away 2012-01-11T13:58:32 dbs: I did have some fun installing the 9.1 server on Ubuntu. Had to install it separately because the Makefile insisted on using squeez-backports. 2012-01-11T13:59:10 Dyrcona: yeah, I thought it was weird that APT_BACKPORT_TOOL had squeeze-backports hardcoded, but I'm not an ubuntu guy 2012-01-11T13:59:14 *** frank___ has quit IRC 2012-01-11T13:59:44 I think we should probably add a different target for 9.1 on ubuntu. 2012-01-11T13:59:45 (git blame says I'm to blame for it) 2012-01-11T14:00:04 * dbs thinks we should split the Makefile.installs up by distro 2012-01-11T14:00:19 with common sections include'd where possible 2012-01-11T14:00:31 that sounds like a good project. 2012-01-11T14:00:52 Lots of other projects do that, and configure figures out which one to run, Qt comes to mind. 2012-01-11T14:01:07 well 2012-01-11T14:01:32 Makefile.install currently comes before configure for us, but I guess if we had a robust enough configure.ac 2012-01-11T14:10:05 *** akilsdonk has joined #evergreen 2012-01-11T14:10:50 *** jeff has quit IRC 2012-01-11T14:13:46 Now, I'm waiting on aptitude search to finish so I can see if Precise has the paypal module in a deb.--Lucid doesn't. 2012-01-11T14:14:53 Nope. Looks like they package Business::PayPal::API instead. 2012-01-11T14:17:33 *** jeff has joined #evergreen 2012-01-11T14:17:33 *** jeff has joined #evergreen 2012-01-11T14:19:43 * tsbere grumbles about broken master SQL 2012-01-11T14:21:31 * Dyrcona wonders what "broken master SQL" means in this context. 2012-01-11T14:22:59 In this case it means "using oils_i18n_gettext in schema.config when it isn't *created* until 2 files later", I think 2012-01-11T14:23:14 * tsbere blames senator 2012-01-11T14:26:37 * tsbere also blames berick 2012-01-11T14:26:44 @blame 2012-01-11T14:26:45 csharp: Error: "blame" is not a valid command. 2012-01-11T14:28:15 tsbere: will fix 2012-01-11T14:28:48 *** edoceo has joined #evergreen 2012-01-11T14:29:40 Event validation failed with Can't locate object method "simple_scalar_request" via package "open-ils.cstore" (perhaps you forgot to load "open-ils.cstore"?) -- what does that mean? 2012-01-11T14:29:46 *** edoceo has left #evergreen 2012-01-11T14:29:53 *** edoceo has joined #evergreen 2012-01-11T14:31:54 *** yboston has joined #evergreen 2012-01-11T14:34:08 tsbere++ 2012-01-11T14:34:24 :q 2012-01-11T14:34:30 senator: want to just move them to 950.data.seed-values.sql ? 2012-01-11T14:34:30 dbs: I got thwocked by the squeeze bug when trying to build the master yesterday on precise pangolin. Also had some missing libraries for opensrf. 2012-01-11T14:34:35 * berick wonders if that will break other things, though 2012-01-11T14:34:53 * tsbere wonders why :q was on his clipboard. Not that he intended to paste it into IRC either, for that matter 2012-01-11T14:35:03 * sal_ still hasn't gotten it working, but hopes to get back to it at some point today 2012-01-11T14:35:49 sal_: What bug are you referring to, the APT_BACKPORT_TOOL? 2012-01-11T14:35:58 *** jbondharris has joined #evergreen 2012-01-11T14:36:13 berick: since i18n isn't generally applied to csp anyway, i figured i'd just not use it 2012-01-11T14:36:37 Trying to build postgres 9.1 on precise. 2012-01-11T14:36:49 sal_: yeah, clearly there needs to be a fix for ubunt-ites 2012-01-11T14:36:58 Either the README is unclear, or the makefile is screwed up. Or both. 2012-01-11T14:37:09 dbs: will be willing guinea pig if you need one. 2012-01-11T14:37:13 sal_: just install the server from aptitude or apt-get, that's what I did. 2012-01-11T14:37:22 luckily, git blames tsbere for that one (install_pgsql_server_debs_91 uses APT_BACKPORT_TOOL) 2012-01-11T14:37:24 :) 2012-01-11T14:38:00 I suppose if we were actually smart, we would detect the OS ourselves and dispatch the appropriate command. 2012-01-11T14:38:01 senator: well, ideally they would be, since they're staff-visible labels, but I'm fine w/ just fixing it for now (a la the surrounding code). 2012-01-11T14:38:12 touble is, we'd need a different target since precise doesn't need backports to install 9.1 2012-01-11T14:38:14 i18n-ize it! 2012-01-11T14:38:16 Good afternoon - quick question when someone has a moment: when building the database, (eg_db_config.pl), should the "hostname" be the FQDN? 2012-01-11T14:38:21 *** JohnC has joined #evergreen 2012-01-11T14:38:28 * berick was wondering if dbs would jump in :) 2012-01-11T14:38:33 Dyrcona: you wouldn't need a different target, you could detect whether to use backports or not 2012-01-11T14:38:41 there's a solid vote, then 2012-01-11T14:38:52 I may do a rollback to opensrf and try to install again... My database seems to be hosed, or something's wrong with new opensrf, since I fail the login test in srfsh... 2012-01-11T14:39:05 I'm all for separating the makefiles by os. :) 2012-01-11T14:39:24 other than actor.usr_standing_penalty where else would the staff client be seeing a user as "in collecitons" 2012-01-11T14:39:30 "make install" - "Okay, let's figure out what distro you're on and proceed accordingly" :) 2012-01-11T14:39:37 dbs++ 2012-01-11T14:39:39 * tsbere once had the idea of "Makefile.install can detect what OS it is on, calls Makefile.install.OS afterwards....." but never made it work 2012-01-11T14:39:48 im trying to merge 2 users and get the "MERGED_USER_IN_COLLECTIONS" but can't find anything to prove that 2012-01-11T14:40:15 make -f Open-ILS/src/extras/Makefile.ubuntu_precise 2012-01-11T14:40:27 and so on would be a nice intermediate step. 2012-01-11T14:41:20 fine, but tsbere has to test it 2012-01-11T14:41:29 jbondharris: the hostname should be whatever works to connect to the server. if localhost works, that's fine. 2012-01-11T14:41:34 senator++ 2012-01-11T14:41:46 senator: that'll teach you to do the right thing 2012-01-11T14:42:18 no good deed and all.. 2012-01-11T14:42:44 *** matt_carlson has joined #evergreen 2012-01-11T14:42:47 :-) 2012-01-11T14:43:34 * tsbere pushes his work to working, then goes to test senator's fix 2012-01-11T14:43:42 jbondharris: specifically, whatever you use for with "psql -U evergreen -h evergreen" is what you use with eg_db_config.pl 2012-01-11T14:43:46 BTW berick: http://git.evergreen-ils.org/?p=working/Evergreen.git;a=shortlog;h=refs/heads/user/tsbere/login_command_line_options 2012-01-11T14:43:50 Drycona: So, the client needs the hosts name, and not localhost, I would want to use "Library", or "Library.rosaryhs.local", 2012-01-11T14:44:18 berick: And I already spotted a typo. <_< 2012-01-11T14:44:22 (in my branch) 2012-01-11T14:44:24 jbondharris: if you mean the evergreen staff client when you say client, it never connects directly to the database. 2012-01-11T14:44:58 Drycona: Yes, that is what I mean. I am being somewhat imprecise. 2012-01-11T14:45:06 jbondharris: what dbs said: whatever works with psql, works with eg_db_config.pl. 2012-01-11T14:45:08 berick: force-pushed the fix. <_< 2012-01-11T14:45:25 tsbere: the thing i pushed to working is the just-dont-use-i18n version. the use-i18n-successfully version is yet to come, with an LP bug 2012-01-11T14:45:44 senator: Then I will wait for the latter, I guess 2012-01-11T14:46:42 tsbere: nice! 2012-01-11T14:48:15 *** dzeiger has quit IRC 2012-01-11T14:49:03 As for backport versus non-backport tool for installing the server, it looks like the consistent thing to do would be to add a install_pgsql_server_backport_debs_91 target that uses APT_BACKPORT_TOOL and change the existing target to use APT_TOOL. 2012-01-11T14:49:14 That would be consistent at least with the client targets. 2012-01-11T14:49:25 is money.collections_tracker for collection agency, or just a log of someone collecting money? 2012-01-11T14:49:58 ah, definetly collection agency 2012-01-11T14:50:40 dbs: OK, now I am going to be obtuse: "localhost" worked on my previous install, and the host's "name" (Library) did not. Is that because that's what I put in for host in the eg_db_config command? If I put in 2012-01-11T14:50:54 "Library" would that have worked as well? 2012-01-11T14:51:25 Dyrcona++ # yep, matches my thinking. at least for the short term 2012-01-11T14:51:59 dbs: I am doin g anew install on a clean machine, so I want to get this right the first time, and not create wierd problems by having a non-standard host name based on my lack of experience. 2012-01-11T14:52:00 I'll do that in a few minutes and update the collab branch. I'm responding to an email about email at the moment. 2012-01-11T14:52:02 jbondharris: the hostname for the database connection doesn't have anything to do with the hostname for the Apache cnnection 2012-01-11T14:52:30 dbs: OK, so I can use localhost without fear of creating problems... 2012-01-11T14:53:18 jbondharris: yep. that will create your database and configure your opensrf.xml file to connect to the database accordingly 2012-01-11T14:54:02 (if you branch out into a multi-server situation, then you'll want hostnames that can resolve to the right database server, but for a single-server situation you're good) 2012-01-11T14:54:03 dbs: Thanks! 2012-01-11T14:59:13 tsbere: https://bugs.launchpad.net/evergreen/+bug/914990 2012-01-11T14:59:14 Launchpad bug 914990 in Evergreen "Fix oils_i18n_gettext() usage at the wrong time and place" (affected: 1, heat: 6) [Undecided,New] 2012-01-11T15:04:11 senator: From what I know, if it isn't in 950.data.seed-values.sql then oils_i18n_gettext is useless clutter. At least in the SQL. 2012-01-11T15:05:14 which means that 012.schema.vandelay.sql, 200.schema.acq.sql, and 400.schema.action_trigger.sql may have useless cruft oils_i18n_gettext calls 2012-01-11T15:07:11 tsbere: that's probable, but fairly easy to convert that from being "cruft" to being consumed 2012-01-11T15:07:35 dbs: Either way, upgrade scripts aren't looked at either, to my knowledge <_< 2012-01-11T15:07:45 tsbere: they don't matter 2012-01-11T15:08:11 tsbere / dbs: that satisfies me as to what i was asking about in the LP bug. thanks 2012-01-11T15:08:54 950.data.seed-values.sql (and soon others) gets parsed and a POT file gets generated; that POT file gets translated in PO files; those PO files generate a SQL file that inserts all of the strings into the database 2012-01-11T15:09:20 * senator saves that remark for future reference 2012-01-11T15:10:09 and not in a transaction, so any existing strings fail harmlessly preserving the local customization / don't get updated with the correct translation depending on how you look at it, and new strings get added in 2012-01-11T15:11:25 Updated my branch with the new install_pgsql_server_backport_debs_91 and install_pgsql_server_debs_91 targets. 2012-01-11T15:11:26 I guess we'll need append mode or something for db-seed-i18n.py and just loop over the available SQL files that hold data 2012-01-11T15:11:35 Havent' tested it, though. 2012-01-11T15:12:06 dbs: May as well use the manifest file for the looping, I think, if we go that route. Not sure of the best way to handle that. 2012-01-11T15:13:00 tsbere: did you publish your i18n build process improvement branch ? 2012-01-11T15:13:38 dbs: The one that doesn't work? <_< 2012-01-11T15:13:42 hah 2012-01-11T15:14:32 *** matt_carlson has quit IRC 2012-01-11T15:14:42 berick: https://bugs.launchpad.net/evergreen/+bug/915004 - if you wanted to jump on that ;) 2012-01-11T15:14:42 Launchpad bug 915004 in Evergreen "Support auto-login in staff client" (affected: 1, heat: 6) [Wishlist,New] 2012-01-11T15:14:58 * tsbere isn't sure if there is any desire to backport a feature like that on berick's part for 2.1.whatever support in the future 2012-01-11T15:16:40 no new features! 2012-01-11T15:17:30 tsbere: yep, will look at that. /me hacks dbs repositories and ports it into every branch 2012-01-11T15:17:51 no, no backported needed for me. my desire is just for dev 2012-01-11T15:18:33 berick: I bet our libraries will love the feature. If only so they can remove the sticky notes holding passwords. <_< 2012-01-11T15:18:42 hah 2012-01-11T15:28:40 berick++ 2012-01-11T15:33:56 *** jenny has quit IRC 2012-01-11T15:39:26 * dbs looks at marcxml-uris and wonders "why aren't you giving me the URI listed at the SYS level when I'm searching at the BR level, baby?" (rel_2_1_ 2012-01-11T15:40:45 http://laurentian.concat.ca/opac/extras/unapi?id=tag:evergreen-opac:biblio-record_entry/2507031/LUSYS&format=marcxml-uris&locale=en-US for SYS level 2012-01-11T15:40:53 http://laurentian.concat.ca/opac/extras/unapi?id=tag:evergreen-opac:biblio-record_entry/2507031/OSUL&format=marcxml-uris&locale=en-US for BR level 2012-01-11T15:41:08 * dbs rages at the dying of the light 2012-01-11T15:42:27 * dbs guesses that the staged search portion of the patch got into rel_2_1, but perhaps a corresponding SuperCat.pm patch did not 2012-01-11T15:42:35 * dbs will check it out laterzzzz 2012-01-11T15:45:41 *** matt_carlson has joined #evergreen 2012-01-11T15:46:56 *** dbs has quit IRC 2012-01-11T15:54:45 *** youdonotexist has joined #evergreen 2012-01-11T15:58:04 *** Meliss has quit IRC 2012-01-11T15:58:25 What does this mean? Event validation failed with Can't locate object method "simple_scalar_request" via package "open-ils.cstore" (perhaps you forgot to load "open-ils.cstore"?) 2012-01-11T15:58:40 I'm getting that in some perl code when calling OpenILS::Application::AppUtils::simplereq() 2012-01-11T16:03:27 *** akilsdonk has quit IRC 2012-01-11T16:04:16 *** collum has quit IRC 2012-01-11T16:11:18 *** jenny has joined #evergreen 2012-01-11T16:11:27 *** senator has quit IRC 2012-01-11T16:12:40 *** senator has joined #evergreen 2012-01-11T16:25:59 *** bshum has quit IRC 2012-01-11T16:25:59 *** bshum has joined #evergreen 2012-01-11T16:26:23 edoceo: I suspect that you are calling something incorrectly, or that the apputils function in question is broken but not used anywhere. Or both. 2012-01-11T16:33:30 *** bwicksall has quit IRC 2012-01-11T16:40:15 *** matt_carlson has quit IRC 2012-01-11T16:41:04 *** fortin has joined #evergreen 2012-01-11T16:42:57 *** kmlussier has quit IRC 2012-01-11T16:47:43 *** bwicksall has joined #evergreen 2012-01-11T16:50:18 *** jbondharris has quit IRC 2012-01-11T16:53:52 tsbere: AppUtils->simplereq() is used all over the place. 2012-01-11T16:54:03 * tsbere didn't look 2012-01-11T16:54:13 * Dyrcona didn't have to. 2012-01-11T16:56:18 *** yboston has quit IRC 2012-01-11T16:56:20 Thinks 0663 is going to have to be split into two scripts. 2012-01-11T16:57:30 *** matt_carlson has joined #evergreen 2012-01-11T17:02:18 *** sal_ has quit IRC 2012-01-11T17:02:31 *** matt_carlson has quit IRC 2012-01-11T17:04:50 And, here we go, again...... 2012-01-11T17:07:50 *** jamesrf has quit IRC 2012-01-11T17:08:30 *** jamesrf has joined #evergreen 2012-01-11T17:30:34 *** Dyrcona has quit IRC 2012-01-11T17:35:04 *** matt_carlson has joined #evergreen 2012-01-11T17:45:16 *** jenny has quit IRC 2012-01-11T18:28:08 *** matt_carlson has quit IRC 2012-01-11T18:44:38 *** graced has quit IRC 2012-01-11T19:13:40 *** tecoripa has joined #evergreen 2012-01-11T19:14:43 I'm looking at the template toolkit *.tt2 files: what is the "ctx." variable that apeears everywhere? What is it used for, where is it defined? 2012-01-11T19:15:14 does it stand for "context"? 2012-01-11T19:18:12 *** tecoripa has left #evergreen 2012-01-11T19:42:55 *** youdonotexist has quit IRC 2012-01-11T20:19:26 *** dbs has joined #evergreen 2012-01-11T20:20:56 @later tell tecoripa yes, ctx means "context" and it's the object that gets loaded up with values from the Perl module that handles the page request; the template can then manipulate and display those values 2012-01-11T20:20:56 dbs: The operation succeeded. 2012-01-11T20:24:28 *** artunit has joined #evergreen 2012-01-11T21:15:32 oh, and of course http://laurentian.concat.ca/opac/extras/unapi?id=tag:evergreen-opac:biblio-record_entry/2507031/CONS&format=marcxml-uris&locale=en-US returns the SYS-level URI 2012-01-11T21:20:01 * dbs finds no corresponding commit to adjust O:W:SuperCat for located URIs in holdings XML. hmm. 2012-01-11T21:30:19 ah, O:A:SuperCat. that's where we need to add the org_unit_ancestors(search_ou)-derived list of OU ids for the URIs 2012-01-11T21:30:36 * dbs gets the feeling nobody else is using located URIs. heh 2012-01-11T22:32:30 dbs: We tried, found they are inconsistent between JS and T pacs, and didn't. I think. 2012-01-11T22:35:48 tsbere: s/between JS and T pacs// 2012-01-11T22:36:35 dbs: While that may be true, I was stating what I believe was our reason to not use them. Inconsistent elsewhere was not, to my knowledge, a factor ;) 2012-01-11T22:41:21 we're using them 2012-01-11T22:43:53 oh nice, just saw https://bugs.launchpad.net/evergreen/+bug/898427 2012-01-11T22:43:54 Launchpad bug 898427 in Evergreen "holdings_xml-uris is broken as is holdings_xml-full" (affected: 1, heat: 5) [Undecided,New] 2012-01-11T22:44:16 * dbs stops trying to develop his own fix, will try out jamesrf's 2012-01-11T22:47:01 * dbs had just gotten to the point of thinking "WTF isn't the block that creates the URIs select stmt getting parsed ever?" 2012-01-11T22:56:09 * dbs cleverly broke marcxml-uris when he was fixing authority browsing, apparently. 2012-01-11T23:00:59 jamesrf++ # works nicely 2012-01-11T23:14:53 oh goody. use_label is getting the href as a value in asset.uri when there's a perfectly good $y subfield 2012-01-11T23:36:57 *** dbs has quit IRC