2011-09-15T03:09:57 *** elene has joined #evergreen 2011-09-15T03:21:06 *** Callender has quit IRC 2011-09-15T03:22:17 *** afterl has quit IRC 2011-09-15T04:12:55 *** elene has quit IRC 2011-09-15T04:37:52 *** enhancin has joined #evergreen 2011-09-15T04:39:10 I am hoping someone has had this issue. I bet it's simple but I ran the autogen.sh script in /openils/bin but I'm getting errors in the staff client saying anything at /eg/... isn't found, and there is no eg folder in /openils/var/web ... 2011-09-15T04:40:30 *** elene has joined #evergreen 2011-09-15T04:55:45 *** elene has quit IRC 2011-09-15T04:58:27 *** enhancin has quit IRC 2011-09-15T05:54:47 *** elene has joined #evergreen 2011-09-15T06:17:32 *** elene has quit IRC 2011-09-15T06:57:23 *** elene has joined #evergreen 2011-09-15T07:35:18 *** plogik has quit IRC 2011-09-15T07:49:06 *** kmlussier has joined #evergreen 2011-09-15T08:04:26 *** elene has quit IRC 2011-09-15T08:37:01 *** mrpeters-isl has joined #evergreen 2011-09-15T08:42:53 *** Shae has joined #evergreen 2011-09-15T08:44:53 *** wolf29 has joined #evergreen 2011-09-15T08:48:18 Wgere is the publicly logged transcript of this irc channel? 2011-09-15T08:48:37 Where, not Wgere 2011-09-15T08:48:43 wolf29: evergreen-ils.org/irc.php 2011-09-15T08:50:41 phasefx: thanks. I have put up a box for a sip server. 2011-09-15T08:51:09 wolf29++ 2011-09-15T08:56:14 I want to set up a sudoer user for any developer who needs that level of access 2011-09-15T08:56:53 Fun - We are being told "hold available emails are going out too early!" 2011-09-15T08:57:14 I check the specific case. Item is captured at pickup library, flagged as on hold shelf. Next day it is in another library entirely and put into transit.... 2011-09-15T08:57:32 * tsbere suspects someone put an item into transit that shouldn't have been in transit 2011-09-15T08:59:43 tsbere: or the pickup lib checked it in on accident before they had the copy in hand 2011-09-15T08:59:57 ah, that may be what you meant, though 2011-09-15T09:00:10 A previous circulation had existed that started and ended at the pickup library for the hold, like the 4 previous circulations like it. 2011-09-15T09:01:12 But a number of our libraries "pre-sort" for other libraries. It is not impossible that an extra book got thrown in a bin. 2011-09-15T09:07:30 *** Callender has joined #evergreen 2011-09-15T09:10:17 *** Callender has quit IRC 2011-09-15T09:15:21 hey all, question about how Evergreen handles 856 tags and OPAC display. Say I have two libraries with identical bibs, but one points to "http://foo.com" and the other needs to point t "http://bar.com". Will Evergreen parse multiple 856 tags and only display the $u that corresponds with the org unit in the $9? 2011-09-15T09:19:05 mrpeters-isl: assuming the $9 of each is set to the library correctly, and you're searching /at that library/, yes, that's the point of located URIs 2011-09-15T09:20:04 perfect. i thought so, just was asked to verify. 2011-09-15T09:20:05 *** atz_ has joined #evergreen 2011-09-15T09:20:06 eeevil++ 2011-09-15T09:20:49 $9 can be any of shortname, org unit id, full library name from actor.org_unit, I assume? 2011-09-15T09:21:04 mrpeters-isl: shortname only, IIRC 2011-09-15T09:21:14 ah good to know 2011-09-15T09:23:41 *** atz__ has quit IRC 2011-09-15T09:24:21 mrpeters-isl: the release notes for 1.6.0.0 say: "Adding an 856$9 containing the short name of a location will..." 2011-09-15T09:24:49 thanks 2011-09-15T09:25:20 wasn't sure which version this was introduced. will be handy. 2011-09-15T09:28:39 *** yboston has joined #evergreen 2011-09-15T09:31:34 and here's a document orginally produced as a testing and best practices guide for located URIs: https://docs.google.com/document/d/1xhAnKg3CSc2Q7kIQYHnqUtGIxEr97EZhYWc5CjhhR6A/edit?hl=en_US (dbs will recongize it... ;) ) 2011-09-15T09:31:58 even better!! 2011-09-15T09:33:00 that document does not address the indicator requirements, however. but, feel free to pass it onto DIG 2011-09-15T09:35:20 hmm we havent been creating asset.uri entries i guess 2011-09-15T09:36:12 *** Callender has joined #evergreen 2011-09-15T09:37:04 was thinking just bare bib records were doing the trick for us 2011-09-15T09:42:54 they will -- if you have located uris turned on (or, not turned off) 2011-09-15T09:43:12 or, you're not setting the aformentioned indicators correctly 2011-09-15T09:43:16 or ... 2011-09-15T09:43:22 ok, cool 2011-09-15T09:43:40 ill do some experimenting. thanks for the advice. 2011-09-15T09:44:08 On the subject of grace periods, I haven't figured out how to best rig up the setting. Should it be YAOUS, or something more global? 2011-09-15T09:44:27 (or extend only when it ends on a closed date compared to whenever it contains a closed date) 2011-09-15T09:44:35 s/or/on/ 2011-09-15T09:45:11 *** jenny has joined #evergreen 2011-09-15T09:45:17 * tsbere currently has an if (1) that he changed to if (0) when testing the alternate mode 2011-09-15T09:45:30 *** kmlussier has quit IRC 2011-09-15T09:46:27 i'd vote for YAOUS 2011-09-15T09:46:50 with that, you can make it global or individual libraries can have their own policy 2011-09-15T09:48:28 Or I could go absolutly insane and add it to the recurring fine rule and the circ table, making it possible to decide based on anything you can use to decide which recurring fine rule to apply! But I am not that insane, so YAOUS or global setting will do for today ;) 2011-09-15T09:52:09 Also, should this be one setting ("grace period is for open days only") or two (add a "grace period auto-extends to end on an open day")? 2011-09-15T09:52:22 tsbere: IMO, an inheritable YAOUS that looks at circ.circ_lib would be most appropriate 2011-09-15T09:52:27 As in, should I allow for the current "grace period ignores closed days" behaviour 2011-09-15T09:53:15 tsbere: re "should I allow for the current...", I think no ... I think everyone considers that a bug 2011-09-15T09:55:28 except maybe libraries with drop-boxes 2011-09-15T09:56:08 phasefx: backdate! :) 2011-09-15T09:56:15 :) 2011-09-15T09:56:25 phasefx: but, why isn't it a bug for them? 2011-09-15T09:56:28 How about libraries with an AMH that accepts returns even when closed? 2011-09-15T09:56:44 * tsbere has at least one of those to worry about as it is 2011-09-15T09:56:51 eeevil: thinking being, if there's a drop-box, why should the library being closed excuse you from the due date? 2011-09-15T09:56:56 tsbere: still not sure how it's a bug? 2011-09-15T09:57:25 * phasefx is conflating things 2011-09-15T09:57:25 eeevil: They want the grace period to include the closed date because a patron, even when the library is closed, can return an item and have it marked as such in the DB. 2011-09-15T09:57:31 I think the AMH is a stronger argument ... 2011-09-15T09:58:19 tsbere: if there's a need to cover that case, then my assertion that "it's always a bug" is wrong in the real world, so I guess a YAOUS for that is useful 2011-09-15T09:59:05 Given how often said AMH acts up they aren't as likely to argue for the grace period not extending, for the record, but it was brought up when they first got the thing. 2011-09-15T09:59:17 the description of that, IMO, should include something like "for use with an always-available AMH system" or the like 2011-09-15T09:59:34 if you go that route 2011-09-15T09:59:42 * tsbere was going to go with "the default is what it currently does, you have to turn something ON to get extending grace periods" 2011-09-15T10:00:07 AKA, don't change default behaviour when upgrading 2011-09-15T10:00:32 I could go either way on that ... extend-past-closed seems like a useful default as well, going forward (covers most complaint cases) 2011-09-15T10:00:57 So I add a default option to the seed values that turns that on at the top of the org tree. ;) 2011-09-15T10:01:08 *** jrodge01 has joined #evergreen 2011-09-15T10:03:36 Alternatively to two options, I define one integer option as a "mode". 0 is "do nothing when a grace period intersects a closed date", 1 is "extend the grace period when it ends on a closed date", and 2 is "extend the grace period when it contains a closed date". 2011-09-15T10:04:59 Might save headaches with "but I only want ends on closed dates for this library, why isn't it happening?" issues because the "extends across all closed dates" setting was turned on somewhere higher in the tree. 2011-09-15T10:06:43 *** Dyrcona has joined #evergreen 2011-09-15T10:06:48 * tsbere notes his latest comment on launchpad was truncated with a ... on the last line of the comment, in the middle of said line. 2011-09-15T10:09:47 *** dbs has joined #evergreen 2011-09-15T10:09:53 *** dbs has joined #evergreen 2011-09-15T10:10:07 eeevil: Any thoughts on "two bool settings" compared to "one mode setting"? 2011-09-15T10:15:49 tsbere: well, there's pretty clearly 3 modes, so we can use a bool where NULL is "what we do now" 2011-09-15T10:16:14 which satisfies the PoLS 2011-09-15T10:16:22 * tsbere isn't sure he likes the idea of a lack of setting doing something different than a false setting 2011-09-15T10:16:51 Because I don't know of any way to set a null via the library settings editor other than delete, which makes a setting higher up in the tree kick in 2011-09-15T10:17:07 point 2011-09-15T10:18:20 unless we invent a class (and underlying table) to support the "mode" version via a fmclass-type setting, 2-bools it is, I guess 2011-09-15T10:19:13 I could go either way on that, actually ... less settings is are better, but the maint overhead... 2011-09-15T10:19:27 (and the magic numbers that brings along) 2011-09-15T10:19:32 are cons 2011-09-15T10:23:49 impliment "radio" settings :D if this is set at a specific org, these others can't be 2011-09-15T10:24:22 * phasefx gets ready to dodge tomatoes 2011-09-15T10:27:22 Perhaps careful choice of boolean options can help with this. One "Grace periods auto extend" overall. One "When grace periods auto extend they do so for all closed dates they hit" to modify things when the first is turned on. 2011-09-15T10:28:20 +1 2011-09-15T10:28:52 still no way to opt out of the first if inherited 2011-09-15T10:28:58 Yes there is 2011-09-15T10:29:02 oh, doh 2011-09-15T10:29:04 The first can be opted out of by setting it to false 2011-09-15T10:29:05 ;) 2011-09-15T10:29:44 sounds good to me 2011-09-15T10:31:50 *** sal_ has joined #evergreen 2011-09-15T10:34:09 When I try to start the staff client on Linux, I get an error "bash: xulrunner: command not found" 2011-09-15T10:34:19 I went back over the steps and it doesn't look like I skipped any. 2011-09-15T10:36:25 jrodge01: Hmm, which linux distribution are you using? 2011-09-15T10:36:44 jrodge01: Perhaps you need to specify the exact xulrunner version... I think I've seen that with debian squeeze. 2011-09-15T10:37:16 jrodge01: tab-completion when typing xul might help 2011-09-15T10:42:15 phasefx: When do you expect to finish up looking at the library settings editor issues? Not sure how much I want to do in adding settings if it is going to conflict with that merge. 2011-09-15T10:43:35 tsbere: did you want to address the missing settings or do you want me try at it? 2011-09-15T10:43:51 * tsbere isn't even sure what settings are missing yet, has been fighting with other stuff 2011-09-15T10:43:59 I'm using debian squeeze. 2011-09-15T10:44:06 Let me get my master install here working with the new TT stuff so I can take a better look 2011-09-15T10:45:01 I tab complete and the same command I try pops up. 2011-09-15T10:45:25 keep hitting tab 2011-09-15T10:45:34 it may be stopping due to needing a final piece to be chosen 2011-09-15T10:45:40 tsbere: k. once lunch hits, I probably won't get to do any dev work other than bug fixing existing features until later this evening 2011-09-15T10:45:43 At which point, when typing xulrunner or the path? 2011-09-15T10:45:50 xulrunner 2011-09-15T10:47:40 It's probably something like xulrunner-1.9.1 or something akin to that. 2011-09-15T10:47:58 At least, that's what it was two nights ago when I installed master on squeeze.. 2011-09-15T10:55:05 Anyone had issues where the "Circ: Void processing fee on lost item return" YAOUS is not respected? 2011-09-15T10:55:42 i have it set to true, but it's not voiding the Lost Materials Processing Fee 2011-09-15T10:55:42 not respected as in set at multiple levels and the most specific one isn't being looked at, or not respected in that there is only one instance and it is ignored? 2011-09-15T10:55:46 We're not using that one, but I have related YAOUS that don't work consistently. 2011-09-15T10:55:52 mrpeters-isl: worked for me in master a few days ago 2011-09-15T10:55:59 we just have one a consortium setting for true 2011-09-15T10:56:11 jrodge01: What do you get when you run "which xulrunner"? 2011-09-15T10:56:19 just tried it on a test account, same behavior 2011-09-15T10:56:19 mrpeters-isl: no, actually, I'm thinking of a different setting 2011-09-15T10:56:35 Was just typing up a bug ticket on my issues with "Circ: Void lost item billing when returned" 2011-09-15T10:56:43 that seems to work ok 2011-09-15T10:56:51 that particular portion gets voided 2011-09-15T10:57:09 its just our $10 processing that consistently does not void along with it 2011-09-15T10:58:01 Hmm, ours don't get consistently voided, my current hypothesis was that it was either due to backdating or some quirk with another setting we have too "Circ: Restore overdues on lost item return" 2011-09-15T10:58:16 where the item didn't generate any overdues to restore 2011-09-15T10:58:28 moodaepo: I ran "whichxulrunner" and there was no output. 2011-09-15T10:58:54 eeevil: 2.2 beta in late October? 2011-09-15T10:59:07 jrodge01: You mean "which xulrunner" there is a space after which 2011-09-15T10:59:55 moodaepo: yes, I ran "which xulrunner", typo in IRC on my part 2011-09-15T11:00:31 What linux distro are you on? 2011-09-15T11:00:45 jrodge01: that means it is not in your execution path. you probably need to install it. 2011-09-15T11:01:08 Well, I just ran that command in Squeeze and it gave me nothing too. 2011-09-15T11:01:10 :( 2011-09-15T11:01:32 Is faiz's staff client problem possibly related to the 2.0.9 bug that prevents creation of org units? 2011-09-15T11:01:40 Tab-completion set it as "which xulrunner-1.9.1" 2011-09-15T11:02:11 we have that same thing enabled, bshum 2011-09-15T11:02:42 it's almost it goes down the list and then stops after the item price is voided 2011-09-15T11:03:45 In related news, we never really came up with a process for dealing with releases that are known to be broken. Seems like advertising a known broken out-of-the-box release isn't a good policy 2011-09-15T11:04:47 dbs: I thought that was a 2.0.8 bug that was fixed in 2.0.9? 2011-09-15T11:05:10 bshum: this could be more evidence of my incredible shrinking brain 2011-09-15T11:05:12 But yeah, announcing bad releases might be good. 2011-09-15T11:06:29 dbwells' patch from 2011-08-16? 2011-09-15T11:07:46 yeah. 2011-09-15T11:07:57 * dbs slinks off 2011-09-15T11:09:52 Drycona: What's the package name to install? Just "aptitude install xulrunner" ? 2011-09-15T11:10:18 bshum: i've at least tracked down the code we need to be looking at 2011-09-15T11:10:37 Application/Circ/Circulate.pm 2011-09-15T11:11:21 line 3043 starts that block 2011-09-15T11:12:01 mrpeters-isl: I started a bug ticket on this - https://bugs.launchpad.net/evergreen/+bug/851000 2011-09-15T11:12:24 i suck at perl, but lets put our heads together and figure this out 2011-09-15T11:13:01 I only mentioned my problems though, I guess we can add the other YAOUS to it for lost processing fee too. Or should that be a separate bug? (they're all kind of related under the umbrella of lost YAOUS not working quite consistently) 2011-09-15T11:13:28 jrodge01: probably. you could search for xulrunner. 2011-09-15T11:13:39 ill create a new bug for my situtation, but optimally i create the bug and include a fix 2011-09-15T11:14:51 Drycona: Is this a normal problem with Debian when installing Evergreen? 2011-09-15T11:14:52 circ.void_lost_proc_fee_on_checkin ist he api call 2011-09-15T11:17:22 Dyrcona: is there a particular version I need? 2011-09-15T11:17:50 * mrpeters-isl worries that there is an open-ils.circ.checkin in the logs, but nothing on the void_lost_proc_fee 2011-09-15T11:18:24 jrodge01: 1.9.2 or 3.6.22 are recommended for Evergreen 2.1 or later. I think they'll work with 2.0 also. 2011-09-15T11:18:55 aha! here's why it failed (for me) bshum 2011-09-15T11:19:04 mrpeters-isl: is your ID different? 2011-09-15T11:19:07 3/4 2011-09-15T11:19:13 nooo much more simple 2011-09-15T11:19:26 the two items that didn't get their processing fee voided are owned by libraries other than where it was checked in 2011-09-15T11:19:35 thus, no permission to void the fee 2011-09-15T11:19:43 if they own the item, it works fine 2011-09-15T11:19:48 Interesting.... 2011-09-15T11:20:09 jrodge01: it is not normal 2011-09-15T11:20:34 jrodge01: however, the default install for Evergreen doesn't really set it up for desktop use 2011-09-15T11:20:52 mrpeters-isl: Maybe that's my problem too, (checking) 2011-09-15T11:22:07 jrodge01: http://packages.debian.org/squeeze/xulrunner-1.9.1 is probably what you want 2011-09-15T11:22:23 mrpeters-isl: Hmm, which permission do you think that would be? VOID_BILLING (if so ours is set at consortium depth...) 2011-09-15T11:22:49 umm hang on 2011-09-15T11:23:20 Sure. 2011-09-15T11:23:24 phasefx: Force-pushed a hopefully fully-fixed library settings editor branch. Rebased and all. 2011-09-15T11:25:13 * tsbere suspects he had a partial paste at one point 2011-09-15T11:26:11 tsbere: sweet, will check it 2011-09-15T11:26:22 hmm bshum having a little trouble finding what permission it would be 2011-09-15T11:26:24 I get 222 settings before and after it now. :D 2011-09-15T11:26:52 rgrepping for OILS_SETTING_VOID_LOST_PROCESS_FEE_ON_CHECKIN and 'circ.void_lost_proc_fee_on_checkin' 2011-09-15T11:27:09 mrpeters-isl: Maybe that one doesn't exist yet. 2011-09-15T11:27:24 Wouldn't be the first time. 2011-09-15T11:27:36 Dyrcona: What is the staff client exactly? What I'm trying to do is set up the server so I can access the webinterface from a different computer. 2011-09-15T11:28:16 jrodge01: The staff client is a xulrunner application that allows one to perform Evergreen staff functions. 2011-09-15T11:28:39 jrodge01: For instance, cataloging, circulation, create patrons, etc. 2011-09-15T11:29:41 mrpeters-isl: Those look like settings though, rather than permissions? 2011-09-15T11:29:54 yeah 2011-09-15T11:30:01 its not a permission, i know 2011-09-15T11:30:08 i was just hoping to find a hit elsewhere on one of those 2011-09-15T11:30:15 Ah 2011-09-15T11:30:16 but im only finding them in Circulate.pm and Const.pm 2011-09-15T11:30:54 mrpeters-isl: Circulate.pm handles checkin so I'd only expect to see it there. 2011-09-15T11:31:09 yeah, agreed 2011-09-15T11:31:20 * tries to unwrap YAOUS... yet another optional user setting? 2011-09-15T11:31:27 close! 2011-09-15T11:31:29 org unit setting 2011-09-15T11:31:31 mrpeters-isl: do they exist in your actor.org_unit_settings? 2011-09-15T11:31:43 Yet Another Org Unit Setting. 2011-09-15T11:31:48 Still one of my favorite Evergreen acronyms. :) 2011-09-15T11:32:00 yeah, they do for the org unit 1 2011-09-15T11:32:16 so, that trickles all the way down, as i read it 2011-09-15T11:32:27 *** bjwebb has joined #evergreen 2011-09-15T11:32:27 *** bjwebb has joined #evergreen 2011-09-15T11:32:44 s/SVF/SVWTF/ 2011-09-15T11:33:57 mrpeters-isl: So just that I get my head around this... if the settings are engaged, the perl leads us to checkin_handle_lost_now_found later on 2011-09-15T11:34:11 bshum: look at Circulate.pm 2011-09-15T11:34:27 it has a whole block of code for what to do with a lost item that's checked in 2011-09-15T11:34:40 Right that's where I am. 2011-09-15T11:34:42 throw the alert, check YAOUS, permissions, etc. 2011-09-15T11:35:03 Right, and then pass to that other function either 3 or 4 (for the bill type I guess?) 2011-09-15T11:35:11 so, if check in library OU = owning library OU my consortium "void" rule works 2011-09-15T11:35:13 And then that deals with voiding. 2011-09-15T11:35:23 voiding-- 2011-09-15T11:35:43 if != then it fails, because the user doesn't have permission to void the bills (via the check-in), at the other OU 2011-09-15T11:35:52 which actually, is exactly how we want it 2011-09-15T11:36:16 our policy states that once you've been billed, it's at the discresion of the library who owns the item whether or not they wish to enforce the bill 2011-09-15T11:36:33 *** bjwebb_ has joined #evergreen 2011-09-15T11:36:46 so an item transits out, gets lost, then brought back -- it's not up to the library where the check-in occurs how to handle the bills because it's not their item 2011-09-15T11:37:05 so evergreen++ for doing exactly what I think it should 2011-09-15T11:37:40 for you, we just need to find the permission that is needed to void those bills so you can make it a consortium setting 2011-09-15T11:38:04 Sounds like it. 2011-09-15T11:38:15 Dyrcona: So the staff client isn't necessary to access the functionality? 2011-09-15T11:38:20 diggin through logs to try to find it 2011-09-15T11:38:41 jrodge01: It depends on the functionality. What are you trying to do? 2011-09-15T11:39:02 jrodge01: In general, library staff need the staff client. 2011-09-15T11:39:05 I'm tasked with learning the ins and outs. 2011-09-15T11:39:28 Okay, so does xulrunner need to be installed on the evergreen/opensrf server, or another computer? 2011-09-15T11:39:44 jrodge01: either / or 2011-09-15T11:39:52 and / and 2011-09-15T11:40:01 jrodge01: On the staff's computer. If you build a custom client, then the installer will get it for you. 2011-09-15T11:40:54 Alright. So on my current computer I installed xulrunner. 2011-09-15T11:41:02 How do I access the system...? 2011-09-15T11:41:30 Dyrcona: meh ... 2.2 beta whenever we can ... but anything's possible 2011-09-15T11:41:42 dbs: not a fan of SVF? 2011-09-15T11:42:01 jrodge01: http://paste.lisp.org/display/124672 2011-09-15T11:42:24 eeevil: no, just another relatively complex thing to grok. my brain is too full 2011-09-15T11:42:27 jrodge01: if you installed Evergreen server from source, those commands will build a client. 2011-09-15T11:43:41 jrodge01: do you need a windows client too? 2011-09-15T11:44:20 sal: yes 2011-09-15T11:44:58 Dyrcona: "make: *** No rule to make target 'rigrelease'. Stop. 2011-09-15T11:45:12 try make rigbeta 2011-09-15T11:45:19 (instead of rigrelease?) 2011-09-15T11:45:23 jrodge01: post the output of the pwd command. 2011-09-15T11:45:24 Same. 2011-09-15T11:45:32 What version of evergreen? Some of those make targets don't work in earlier versions. 2011-09-15T11:45:42 oh, that's true, too. 2011-09-15T11:46:00 * Dyrcona is just so used to do everything from master branch. 2011-09-15T11:46:18 * have only been building since 2.0.6 2011-09-15T11:46:28 I'm going for 2.0 2011-09-15T11:46:37 Hmm. 2011-09-15T11:46:42 If it's suggest I go for 2.1, I can try that out. 2011-09-15T11:47:24 jrodge01: are you building from source, or from the zip files on the download page? 2011-09-15T11:47:44 Source. 2011-09-15T11:48:05 okay, then you will have to compile your own Windows client... 2011-09-15T11:48:09 Goes to check a 2.0 box 2011-09-15T11:48:43 If its easier to set up using a zip, I'll just set it up again. 2011-09-15T11:49:03 dbs: my brain is too full even for tsbere's transit suppression settings 2011-09-15T11:49:05 * compiling from source is gooooood for you ;-) 2011-09-15T11:49:35 jrodge01: sal_: In case you need our notes from community client building: http://open-ils.org/dokuwiki/doku.php?id=scratchpad:packaging_the_staff_client_for_windows 2011-09-15T11:49:36 phasefx: take a break. 2011-09-15T11:49:38 brain keeps flashing tl;dr 2011-09-15T11:50:19 there's no reason to compile the staff client from source, right? 2011-09-15T11:50:19 bshum: thanks. That helped me build my master windows client ;-) 2011-09-15T11:50:43 dbs has a point actually, you could just symlink 2011-09-15T11:50:46 for 2.0, just symlink 2011-09-15T11:50:48 To 2.0.9 or whatnot 2011-09-15T11:51:14 * dbs doesn't want to scare jrodge01 off with too much insanity. We have to save the insanity for later. 2011-09-15T11:52:18 jrodge01: If you install from one of the tarballs, you should be able to download and use the corresponding windows client. 2011-09-15T11:52:51 jrodge01: dbs is right, you only need to build a client from source in special cases. (I have one of those special cases, so I always build from source.) 2011-09-15T11:53:50 I think it's the windows client that I have to build separately. Although maybe I've been doing extra work for nothing ;-) 2011-09-15T11:56:52 We merge some local branches that affect the client, so we build clients from source. 2011-09-15T11:59:23 * dbs wonders how confused jrodge01 is at this point 2011-09-15T11:59:40 *** jrodge01 has quit IRC 2011-09-15T12:00:29 apparently quite a bit. 2011-09-15T12:03:06 * didn't mean to scare him (her?) off... 2011-09-15T12:03:28 To his original question though with xulrunner vs. xulrunner-1.9.1 or whatnot... maybe we should remove that step for running the staff client on Linux from the wiki install steps. 2011-09-15T12:03:33 *his/her 2011-09-15T12:04:02 btw, should I have to build a windows client when compiling from the master? or does that get generated as part of the make install? 2011-09-15T12:04:24 Or you know... use the README as single-source like dbs always tries to remind us ;) 2011-09-15T12:04:39 (since that step for staff client on linux is not in the READMEs either) 2011-09-15T12:04:45 the xulrunner line isn't in the README 2011-09-15T12:05:24 sal_: You have to explicitly build the client from master when building from source. 2011-09-15T12:05:37 otoh, that's how I test whether I have a good build--if I can (1) run the client, and (2) add the SIP permission group. 2011-09-15T12:05:39 i.e from a git checkout 2011-09-15T12:05:55 grabbing 0622 and 0623 2011-09-15T12:06:02 sal_: and instructions on how to do it (including making a windows installer, if you want it) are here: http://evergreen-ils.org/dokuwiki/doku.php?id=mozilla-devel:building_the_staff_client 2011-09-15T12:08:15 gmcharlt_: thanks! My real question was whether "make STAFF_CLIENT_STAMP_ID=rel_name install" made the windows client too, and I had just been wasting time :-) 2011-09-15T12:08:38 I think not, but... 2011-09-15T12:08:45 fwiw, for the to-be-written xulrunner README instructions, i would recommend "Download xulrunner tarball from mozilla.org - yada yada" as current distros ship xulrunner that's too new for us (and xulrunner-old packages are disappearing from repos for new distros) 2011-09-15T12:08:53 sal_: No, you have not been wasting time. 2011-09-15T12:08:54 sal_: definitely not 2011-09-15T12:09:53 * thinks that the documentation around building Evergreen is *really* good. I've run into very few issues, other than with autogen-related snafus. 2011-09-15T12:10:23 Speaking of those steps for building your own client, nsis is not installed by default with an Evergreen install. Is that intended and we should always install separately as desired? (or is that the purpose behind the strange error I got when trying to build a client without nsis installed?) 2011-09-15T12:10:57 strange error? 2011-09-15T12:11:05 Erm... warning maybe 2011-09-15T12:11:18 I get the "nsis is not installed error", but it's not strange. 2011-09-15T12:11:33 * tsbere wrote that check in because nsis isn't installed by default 2011-09-15T12:11:34 * always forgets that she needs it until building the Windows client ;-) 2011-09-15T12:11:35 That's the one, guess it shouldn't bother me, that's how it ought to work. 2011-09-15T12:14:26 My thought was most people wouldn't be building the client, so nsis should not be installed as a core dep; kind of like we don't install postgresql server by default anymore 2011-09-15T12:14:54 dbs: maybe in developer instructions for source README? 2011-09-15T12:15:37 sal_: maybe. we could add optional targets to Makefile.install if they aren't already there 2011-09-15T12:18:29 is there any way to edit an existing record via the patron registration web interface (/eg/actor/user/register)? 2011-09-15T12:22:37 *** plux has joined #evergreen 2011-09-15T12:23:59 jeffdavis: IIRC you could pass a param like ?id= and it calls up the existing record - but "Search Patron" should bring that up for you (unless you're on an old version of Evergreen?) 2011-09-15T12:24:38 berick: practical_examples++ 2011-09-15T12:24:39 *** akilsdonk has joined #evergreen 2011-09-15T12:28:21 dbs: this is 2.0.8; no search option (at least not within the web interface itself) and ?id= doesn't seem to work 2011-09-15T12:28:41 so maybe just not possible on this version 2011-09-15T12:29:01 jeffdavis: okay, wasn't sure if you were inside the staff client and just trying to refer to the single-pane vs. seven-pane editor 2011-09-15T12:29:06 mrpeters-isl: Hmm, I tried returning a "lost" item to the circulating library who owned the copy as myself (a super admin with all permissions) and the lost fee remained on the patron's record still. 2011-09-15T12:29:28 dbs: yeah no sorry, I want to look at the edit stuff with firebug is all :) 2011-09-15T12:29:29 mrpeters-isl: That leaves me wondering if it's really permission related as you suggested. 2011-09-15T12:30:05 jeffdavis: ?usr= 2011-09-15T12:30:22 jeffdavis: reading register.js, var userId = cgi.param('usr'); 2011-09-15T12:30:42 (well, &usr=) 2011-09-15T12:30:59 slick! I missed that in register.js, thanks 2011-09-15T12:41:34 you also might want &ws=something 2011-09-15T12:41:45 ie: a valid workstation 2011-09-15T12:50:16 mrpeters-isl: Yep, I just found a case where the lost bill is still on the patron's record even though the patron, item, and checkin were all the same lib. So I guess my issue is different than the one you described :( 2011-09-15T12:54:35 hey, is anyone doing apt upgrades on evergreen-ils.org periodically? sys admin-y maintenance :) 2011-09-15T12:56:05 I think we're not using the packaged version for dokuwiki; how about wordpress? 2011-09-15T12:56:22 * tsbere can honestly say he has never updated via apt on evergreen-ils.org 2011-09-15T12:56:37 * bshum looks at moodaepo... 2011-09-15T12:56:39 * tsbere also doesn't think he has shell access there, so tis a moot point for him ;) 2011-09-15T12:57:02 tsbere: would you like to be a community sys admin for it? :D 2011-09-15T12:57:18 Not really. I have a hard enough time remembering to update our servers. ;) 2011-09-15T12:58:06 * phasefx just can't remember if anyone committed to maintaining it 2011-09-15T13:07:48 *** Dyrcona has quit IRC 2011-09-15T13:17:16 phasefx: no one really committed to it, though it's sort of a group admin situation 2011-09-15T13:17:36 I guess it would kind of fall to me/GPLS to maintain the server/ODS 2011-09-15T13:17:39 *** jenny1 has joined #evergreen 2011-09-15T13:17:44 s/ODS/OS/ 2011-09-15T13:17:50 *** jenny1 has left #evergreen 2011-09-15T13:18:07 * csharp unless moodaepo just wants to ;-) 2011-09-15T13:18:39 csharp: better than a server/DOS 2011-09-15T13:18:59 heh 2011-09-15T13:19:15 *** jenny has quit IRC 2011-09-15T13:20:51 phasefx: did you have a particular concern? or were you just thinking of it? 2011-09-15T13:31:26 *** kmlussier has joined #evergreen 2011-09-15T13:34:10 eeevil / dbwells / anyone else who cares: This is what I have so far for grace period extension: http://git.mvlcstaff.org/?p=tsbere/ILS.git;a=shortlog;h=refs/heads/grace_period_extend 2011-09-15T13:34:13 *** artunit has quit IRC 2011-09-15T13:35:32 *** matt_carlson has joined #evergreen 2011-09-15T13:38:23 *** bjwebb_ has quit IRC 2011-09-15T13:39:14 tsbere: casual reading of the patch, it looks reasonable indeed 2011-09-15T13:39:14 *** bjwebb has quit IRC 2011-09-15T13:43:14 I am thinking I should switch to just copying $due instead of using $due_dt, if I am going to do all the math in seconds anyway (I originally was doing a days => 1 addition) 2011-09-15T13:45:17 tsbere: not looking at the code, but $due is the epoc-time of $due_dt, right? 2011-09-15T13:45:24 yea 2011-09-15T13:45:39 yeah, seems like a performance win to me 2011-09-15T13:45:49 (not huge, but it adds up) 2011-09-15T13:45:59 Hmmm 2011-09-15T13:46:05 esp checkin-time fine calc 2011-09-15T13:48:36 Hmmm 2011-09-15T13:48:40 strftime 2011-09-15T13:48:47 I am using that on due_dt 2011-09-15T13:53:26 Looks to be a PITA to do without due_dt. May be better off not changing the usage of due_dt. 2011-09-15T13:54:10 csharp: nothing specific, just foreboding 2011-09-15T13:54:58 way awesome, tsbere 2011-09-15T13:55:27 eeevil: If I wasn't using strftime I would go for using a copy of $due. Since I am using strftime.... 2011-09-15T13:56:33 tsbere: have you looked at the "feature" ;) that causes grace days to not be honored when you do a backdated checkin? 2011-09-15T13:57:12 mrpeters-isl: That is on my list of things to look into, although the extending working makes that less important, IMO 2011-09-15T13:57:34 i agree 2011-09-15T13:57:40 it will take care of many of the cases 2011-09-15T13:57:44 tsbere: ahh... well, you could use transforms in the where clause there to turn the columns into epoc seconds (extract(epoc from )), but I'm not sure how messy that would be with CDBI 2011-09-15T13:58:08 Or I can be lazy and leave the *working* code alone ;) 2011-09-15T13:58:25 tsbere: there's always that, yes ;) 2011-09-15T13:58:33 working code wins! 2011-09-15T13:58:42 On checkin shouldn't be that big of a deal if you are running the fine generator nightly anyway. The longer it is overdue the more likely it is this never runs because there is at least one billing already. 2011-09-15T14:00:14 tsbere: am I right in understanding that AUTOUPDATE_HOST no longer needs to be specified during Evergreen installation, and just using during staff client compilation suffices? 2011-09-15T14:00:59 phasefx: I personally set it at configure time ;) But you can use it at staff client compilation time as well, if you aren't letting install do the final run of that for you. 2011-09-15T14:01:17 tsbere: has that always been true? or is true in 2.1? 2011-09-15T14:01:31 I think it was true from the first time automatic updates got merged in 2011-09-15T14:01:37 cool deal, thanks 2011-09-15T14:02:31 bye! 2011-09-15T14:02:41 *** wolf29 has quit IRC 2011-09-15T14:04:59 mrpeters-isl: Here is a question. Should backdated checkins obey automatically extended grace periods, or just the originals? ;) 2011-09-15T14:05:20 *** matt_carlson has quit IRC 2011-09-15T14:08:46 let me ask around with the people who initially reported it and get back to you 2011-09-15T14:11:31 mrpeters-isl: Right now grace periods both don't auto extend *and* are ignored on backdated checkin, so... 2011-09-15T14:14:17 tsbere: grace period code looks good. I am a little concerned about the 7 day limit, though. It made good sense when checking for only dow closings (to prevent infinite loops if the library is set to closed every day), but between semesters, we sometimes close for more than a week, and I suspect the same is true of many academics. 2011-09-15T14:15:10 not sure how high, it should go, though 2011-09-15T14:15:15 dbwells: I picked 7 days based on the disabled code limiting to 7 days 2011-09-15T14:15:24 ok, what do i have broken to cause this? http://pastie.org/2538894 2011-09-15T14:15:31 everytime I try to edit marc in staff client 2011-09-15T14:15:38 tsbere: right, but the disabled code only checked dow of week closings, right? 2011-09-15T14:15:50 I have no clue what the disabled code was trying to do 2011-09-15T14:15:57 I largely deleted it and started over ;) 2011-09-15T14:17:08 Right, being half-done and all. I think it was meant as an infinite loop protection (if we are closed every day, we're sunk). 2011-09-15T14:17:08 It *looks* like the disabled code was doing "for every closed date by hours of operations after the due date extend the grace period a day" 2011-09-15T14:17:25 Mine is also meant as infinite loop protection 2011-09-15T14:18:12 The disabled code would stop when it found an open day after the due date or after 7 days, ignoring the length of the grace period 2011-09-15T14:18:18 I understand that, but when checking dow, we only ever need to check 7, but I think we need to check more if closed dates are in the mix. 2011-09-15T14:19:00 So do I harcode a bigger number, or make it an OU setting with a hardcoded bigger default? :p 2011-09-15T14:20:06 tsbere: the intent of the old code was to, yes, find any HOO closings after we hit a closed date (or hit none of them) ... up to 7, because then we're just always closed 2011-09-15T14:20:53 s/harcode/hardcode/ <_< 2011-09-15T14:21:15 I would think the value could be fairly high. It is only going to hurt lazy libraries who are closed a lot :) 2011-09-15T14:22:04 How about 31 days? If you can't find enough days to eat up your grace period in an entire month........... 2011-09-15T14:22:37 i wanna work there 2011-09-15T14:22:47 Or we could only increment $count for dow closings, to keep the original protection, and thereby get around any code-based policy exceptions. 2011-09-15T14:22:54 what we want to do, I think, is look at each day in turn. if it's closed, do what your OU settings say to do, and move on. if we check each day for HOO settings, and they're all closed (not 7 total, but all days) then we can stop 2011-09-15T14:22:56 (the library who isn't open enough to eat up grace days) 2011-09-15T14:23:37 mrpeters-isl: I work at a "library" that is "closed" 24/7 ;) 2011-09-15T14:23:57 (we exist in our own system as a location, but we never "open") 2011-09-15T14:24:03 *** moodaepo_ has joined #evergreen 2011-09-15T14:24:06 but if you're HOO-closed 6 days a week, and closed-date closed for 3 weeks on that normally-open day, we should keep looking for open days ... perhaps up to a year, if we need an upper bound 2011-09-15T14:25:05 I concur 2011-09-15T14:25:05 tsbere: but, in your case, if you're HOO-closed 24/7 then you'll stop after the first week of overdue 2011-09-15T14:25:31 atheos: hey, do you guys still use pfsense? I'm wondering if we ever learned whether it mangled staff client javascript or not 2011-09-15T14:25:38 so we'll need to use a hash instead of a count for HOO-closed, but I think that should be an easy-ish change 2011-09-15T14:25:39 I am updating the contributors page and wanted to confirm that it's ok to add tsbere as an active contributor, I sure hope so : ) 2011-09-15T14:25:48 I might also clarify when I say dow closings, I am talking about HOO closings 2011-09-15T14:25:49 phasefx: think they do 2011-09-15T14:26:02 *** Dyrcona has joined #evergreen 2011-09-15T14:26:07 eeevil: I have an alternate idea. Working on it now. 2011-09-15T14:26:11 phasefx - yes, but we bypass content filtering (proxy) on all of our staff client workstations. 2011-09-15T14:26:14 Well, alternate-ish idea 2011-09-15T14:26:28 atheos: ah, gracias 2011-09-15T14:26:38 phasefx - haven't tested it in a long time, so I'm not certain if it's still an issue. 2011-09-15T14:27:06 I know websense will definitely mangle things if it sees enablePrivilege, for example 2011-09-15T14:27:16 it seems like every time we add a new library their content filtering is getting in the way 2011-09-15T14:27:27 soon as they exclude our IP's its much better 2011-09-15T14:28:29 * moodaepo_ will update the page hearing no objections in 5....4...3..2.1 2011-09-15T14:29:06 *** matt_carlson has joined #evergreen 2011-09-15T14:31:39 tsbere: this is minor, but if we are considering performance tweaks, I would try adding multi-day closed dates as one lump (or at least a sub loop) rather than look it up multiple times. 2011-09-15T14:31:50 just an idea 2011-09-15T14:33:56 content_filtering-- 2011-09-15T14:34:20 dbwells: No clue how to begin on that right now, honestly. Hadn't given that much thought. 2011-09-15T14:37:08 *** moodaepo_ has quit IRC 2011-09-15T14:49:04 *** sal_ has quit IRC 2011-09-15T14:49:59 *** matt_carlson has quit IRC 2011-09-15T14:54:29 *** sal_ has joined #evergreen 2011-09-15T14:56:04 dbwells: Waiting on your transit suppression notes, BTW. Figured I should hold off on anything significant on that branch until I see what you have to say. 2011-09-15T14:57:03 tsbere: they are in progress. Your examples helped a lot, so I have less to say. 2011-09-15T14:57:44 *** _bott_ has quit IRC 2011-09-15T14:58:03 * tsbere is glad his examples were useful 2011-09-15T14:58:10 *** _bott_ has joined #evergreen 2011-09-15T14:58:48 * tsbere found a bug in his in-progress changes to the grace period extension code. <= and >= do very different things......... 2011-09-15T15:01:58 tsbere++ I found the examples very useful too. 2011-09-15T15:03:21 today i am killing zombie hacks 2011-09-15T15:03:55 such as a hack that lived far too long, resulting in a need for a 400 MB report each night. 2011-09-15T15:04:21 (the report continued to grow over time the longer the hack lived) 2011-09-15T15:05:03 That's a big report. 2011-09-15T15:05:10 yeah :P 2011-09-15T15:05:26 big and wasteful 2011-09-15T15:08:47 any way to tie an old SVN commit number to a git commit? trying to track down an old patch...wasn't sure how those migrated. 2011-09-15T15:09:51 mrpeters-isl: svn and trac are still up if you just need to see the changeset 2011-09-15T15:09:52 *** akilsdonk has quit IRC 2011-09-15T15:10:03 oh are they? ok cool 2011-09-15T15:12:11 phasefx: remember this at all? http://svn.open-ils.org/trac/ILS/changeset/20282 2011-09-15T15:12:22 same rogue print button is on the transit list 2011-09-15T15:12:40 there is one that works, one that says We are unable to Print or print Preview this page. 2011-09-15T15:13:08 mrpeters-isl: also this works, with some low rate of false positives possible: git log --grep=@20282 2011-09-15T15:13:14 for future reference 2011-09-15T15:13:23 eeevil / dbwells : Force-pushed a new version of grace_period_extend (rebased it too). Skips looping at all if all 7 days are closed (no days to generate fines for anway, right?), will loop for 366 checks, and should use any closed date info it finds (multi or not) as it goes along from each call. 2011-09-15T15:14:00 note that the 366 checks may not be 366 days. A multi-day closed date can use one loop but add several days. 2011-09-15T15:14:07 (have to be in the right branch though, so not ideal i guesS) 2011-09-15T15:14:19 mrpeters-isl: I remember it. the default interface changed from an html one to a xul one, and you can't print a xul interface like that 2011-09-15T15:14:19 thanks senator 2011-09-15T15:14:44 phasefx: yeah, so i was thinking i'd comb through and try to find others to get rid of the redundant/broken print buttons 2011-09-15T15:15:30 shouldn't be any more, but you're welcome to look 2011-09-15T15:15:42 the transit list has one, at least in 2.0.4 2011-09-15T15:15:55 ah 2011-09-15T15:16:22 We have a complaint that you can't print from checkin. They want to be able to. 2011-09-15T15:16:43 hm 2011-09-15T15:16:48 * mrpeters-isl tries 2011-09-15T15:17:15 yeah "Print List" at the bottom worked 2011-09-15T15:17:22 from checkin 2011-09-15T15:17:55 * tsbere didn't try, he just noted the report 2011-09-15T15:18:02 yeah, have them try that 2011-09-15T15:18:52 * tsbere has been focusing on things like "grace periods aren't working!" because of the closed day eating it 2011-09-15T15:22:09 tsbere: looking 2011-09-15T15:22:25 phasefx: how does "browser.xul" come into play? 2011-09-15T15:22:35 apparently that's where the rogue print button lives 2011-09-15T15:22:48 along with back, reload, forward, etc. 2011-09-15T15:23:15 mrpeters-isl: There is a param that can be passed to disable the print button (or one you dont pass, I forget which way it goes) 2011-09-15T15:23:24 As well as a number of the other buttons 2011-09-15T15:23:35 Different buttons use different parameters 2011-09-15T15:23:57 yeah, wondering if its that SVn link i pasted earlier 2011-09-15T15:24:29 browser.xul is essentially a wrapper around what amounts to an iframe. It's intended to be used with embedded HTML pages (so you can have back/foward/print). It's not really needed for xul pages, but that can happen out of laziness (when replacing an html interface with a xul interface) 2011-09-15T15:25:10 or when the entry point for the xul interface is an html interface 2011-09-15T15:25:15 (old Local Settings) 2011-09-15T15:26:23 tsbere: other than the comment that says "We limit to 7 potential days below", that is better than I'd imagined. rock, sir. docs can say "if you are closed for a year, use the frickin backdate please." 2011-09-15T15:29:46 ok so the real fix might be just not using browser.xul to wrap the transit_list.xul 2011-09-15T15:32:01 mrpeters-isl: but the easiest fix is that no-print/disable-print parameter 2011-09-15T15:32:37 eeevil: I WILL FIX THE COMMENT. Because I hate knowingly putting comments in that lie. ;) 2011-09-15T15:32:40 i dont recall that one 2011-09-15T15:35:16 tsbere: :) 2011-09-15T15:35:35 * eeevil awaits dbwells' eyeballs 2011-09-15T15:37:25 * dbwells is trying to write something intelligent about transit suppression needs 2011-09-15T15:40:40 * tsbere force-pushed the comment fix, not wanting a full commit just for that. 2011-09-15T15:47:44 *** plux has quit IRC 2011-09-15T15:51:30 *** dbs has quit IRC 2011-09-15T15:58:08 tsbere: at first (and second) glance the grace-period code looks good. I'll do my best to test it tomorrow against some of our data and report back. 2011-09-15T16:02:42 *** mrpeters-isl has quit IRC 2011-09-15T16:04:04 *** dbs has joined #evergreen 2011-09-15T16:04:04 *** dbs has joined #evergreen 2011-09-15T16:18:34 eeevil: I want to split this grace period code out to a function that can be used when backdating as well. Any thoughts on where to put it? (the backdating point I want to use it with happens in CircCommon.pm) 2011-09-15T16:19:26 *** kmlussier has quit IRC 2011-09-15T16:20:57 tsbere: if you can pull all the inputs out, AppUtils might be a good spot. save an extra cross-service call 2011-09-15T16:26:30 eeevil: action.pm doesn't currently use AppUtils. Should I make it do so? 2011-09-15T16:26:40 (and I can easily handle the inputs) 2011-09-15T16:30:55 eeevil: Actually, looking, I would have to add a few things to AppUtils. There any problem with me adding it to CircCommon (I consider fines to be circ related) and pulling that in instead? Looks like it is a tad closer on "bringing things in" than AppUtils. 2011-09-15T16:31:54 Yarrrggghhh! 2011-09-15T16:34:05 Eureka! 2011-09-15T16:34:09 identifier|isbn:9780590129275 2011-09-15T16:34:22 Now I know what I need to fix! 2011-09-15T16:38:12 *** jenny has joined #evergreen 2011-09-15T16:38:16 *** jenny has left #evergreen 2011-09-15T16:39:35 Dyrcona: is that good or bad? (or just loud? :-) 2011-09-15T16:39:37 tsbere: as long CircCommon.pm doesn't register any methods, no objections (don't want to pollute the open-ils.storage namespace) 2011-09-15T16:39:49 sal_: just loud. 2011-09-15T16:39:59 eeevil: It looks to define a few functions, tis about it. 2011-09-15T16:40:06 sal_: it could be bad that I'm starting to understand the search code, though. :) 2011-09-15T16:40:26 Dyrcona: oh, boy... 2011-09-15T16:43:03 * Dyrcona is trying to teach z39.50 search some new tricks. 2011-09-15T16:43:40 Dyrcona: you'll like config.search_aliase (IIRC) then :) 2011-09-15T16:44:09 in fact, there should be an isbn alias in the stock seed data 2011-09-15T16:44:41 so, in CQL (SRU) you can say: isbn=9780590129275 2011-09-15T16:44:52 which is translated to: isbn:9780590129275 2011-09-15T16:45:04 which is translated to: identifier|isbn:9780590129275 2011-09-15T16:45:19 eeevil: My goal is to make the ISBN search from z39.50 do the same as ISBN quick search when searching the local catalog. 2011-09-15T16:45:28 2-step translation ... second "which is" does not replace the first 2011-09-15T16:46:30 Dyrcona: in master, just map the appropriate attr (in one of the z-related config files) to isbn ... that'll do what I said up there 2011-09-15T16:46:39 Dyrcona: we're using the isbn index in SRU. what are you trying to do that's different? 2011-09-15T16:47:54 Long story short, make the ISBN field use the identifier|isbn: search when search the local catalog from the z39.50 search page. 2011-09-15T16:48:20 * Dyrcona didn't know that's what he really wanted until the Eureka! above. 2011-09-15T16:49:18 Dyrcona: okay, so what eeevil said and what we're doing 2011-09-15T16:50:09 oils_z3950.xl maps attr 7 to eg.isbn already. 2011-09-15T16:51:16 it's like somebody wanted to expose ISBN's on the standard Bath profile :) 2011-09-15T16:52:25 I'm not sure you understand (or at least I feel confused at this point). 2011-09-15T16:52:52 It sounds like I don't understand. That's not what you want? 2011-09-15T16:53:26 When staff do a z39.50 search on a 13 digit variant of a 10-digit isbn, and check Local Evergreen catalog, they want it to match the local record even if the MARC only has the 10-digit isbn. 2011-09-15T16:53:42 Is that what you understood me as asking? 2011-09-15T16:55:09 Dyrcona: like http://zed.concat.ca/opac/extras/sru?version=1.1&operation=searchRetrieve&query=eg.isbn=9780131855250 2011-09-15T16:55:47 (you never actually said anything about 10-digit vs. 13-digit ISBNs before, at least not while I was in channel) 2011-09-15T16:55:54 yeah, but from the staff client. 2011-09-15T16:56:03 on our own catalog. 2011-09-15T16:56:07 it doesn't do that now. 2011-09-15T16:58:17 however, I think I'm starting to get it. 2011-09-15T16:58:25 config.metabib_search_alias is the table you're looking for 2011-09-15T16:58:52 we have eg.isbn | identifier | 18 for eg.isbn 2011-09-15T16:59:32 which is in the stock 950.data.seed-values.sql file 2011-09-15T16:59:53 so... either you don't have that, or your 10-digit/13-digit normalization wasn't normalizing? 2011-09-15T17:00:51 dbs: I have that. 2011-09-15T17:01:16 when I search the 13-digit isbn I posted earlier in isbn quick search, it finds the right record. 2011-09-15T17:01:56 when I search that isbn the cataloging module's z39.50 search page, with local catalog checked, it doesn't find the record. 2011-09-15T17:05:18 light dawns on dbs 2011-09-15T17:05:54 the z39.50 _search_ page, as in, "Import via z39.50" 2011-09-15T17:06:06 yes, sorry, i wasn't being clear enough. 2011-09-15T17:06:18 it's late, and my brain is a bit rattled. 2011-09-15T17:06:26 I should have picked up on that, I was glazing over key pieces of info 2011-09-15T17:06:47 I think it might need a change is Z3950.pm, but I'm not sure. 2011-09-15T17:06:56 heh. You could add your own Evergreen instance as a z39.50 target, and mask the "Local Evergreen Catalog" target :) 2011-09-15T17:07:10 *** matt_carlson has joined #evergreen 2011-09-15T17:07:20 They also want us to make changes to loading records found in the local instance. 2011-09-15T17:07:24 Would complicate matters slightly ;) 2011-09-15T17:08:08 tsbere: let's get the searching figured out, first... 2011-09-15T17:08:16 it's an older request. 2011-09-15T17:09:19 i'm going to try putting in identifier|isbn into the search hash generated ins the zstyle search methods or thereabouts. 2011-09-15T17:09:22 I would prefer to not break the newer request's ability to be completed when fixing the older one ;) 2011-09-15T17:10:12 we can tell them search by isbn or retrieve records differently: choose one. 2011-09-15T17:12:58 *** Shae has quit IRC 2011-09-15T17:20:17 *** enhancin has joined #evergreen 2011-09-15T17:21:58 dbs: Thank you for your help the other day. The config files worked 2011-09-15T17:22:13 yay! 2011-09-15T17:22:18 We need to add a check for that 2011-09-15T17:23:02 Yeah, but now I'm getting messages about /openils/var/web/eg missing (which it is unless it's a rewrite rule). I tried running autogen but nothing was changed really 2011-09-15T17:29:16 *** Dyrcona has quit IRC 2011-09-15T17:29:38 Am I missing something? =/ 2011-09-15T17:29:43 *** akilsdonk has joined #evergreen 2011-09-15T17:31:08 *** sal_ has quit IRC 2011-09-15T17:31:09 enhancin: yeah, your Apache config sounds messed up. it should be a rewrite for the template toolkit stuff 2011-09-15T17:31:15 * dbs needs to hit the road again 2011-09-15T17:31:57 Alright I'll check through the configs again for apache 2011-09-15T17:32:00 thanks dbs 2011-09-15T17:33:08 *** dbs has quit IRC 2011-09-15T17:41:28 Does anyone have the /eg rewrite rules? I can't find it anywhere in the eg.conf or eg_vhost.conf. I see the Location /eg directive but it doesn't seem to be working 2011-09-15T17:43:58 enhancin: in master? 2011-09-15T17:44:05 *** matt_carlson has quit IRC 2011-09-15T17:44:57 berick: i'm not sure...i'm just not getting my /eg rewritten somehow... 2011-09-15T17:45:10 i found rewritemap directive about the /openils/bin/openurl_map.pl 2011-09-15T17:46:22 enhancin: i mean, what code/version are you using? 2011-09-15T17:46:51 *** sal_ has joined #evergreen 2011-09-15T17:46:56 Evergreen 2.0.9 2011-09-15T17:48:19 enhancin: does your startup.pl have: 2011-09-15T17:48:19 use OpenILS::WWW::EGWeb ('/openils/conf/oils_web.xml'); 2011-09-15T17:49:26 yeah let me go through thre oils_web.xml 2011-09-15T17:57:12 everything looks good there...what are the web_dir and base_dir values? both of them supposed to be the web root? 2011-09-15T17:57:58 web_dir is the same as DocumentRoot in apache 2011-09-15T17:58:14 we should probably remove the /openils/lib/perl5 thing from startup 2011-09-15T17:58:16 base_path is the base url page /eg 2011-09-15T17:58:26 phasefx: yeah ;) 2011-09-15T17:58:29 nevermind yeah i see it 2011-09-15T17:58:42 * phasefx leaves that as an exercise for the log trawler 2011-09-15T17:58:49 enhancin: what exactly do you mean by /eg is not getting rewritten? 2011-09-15T17:59:20 to be clear, if you point a browser directly at /eg, it should return a 404 2011-09-15T17:59:26 *** yboston has quit IRC 2011-09-15T18:00:06 Awesome, yeah. I'm going to go through every single one of my config files and make sure they're all correct. i feel silly 2011-09-15T18:00:25 i must have taken the ones that weren't from evergreen. thanks guys 2011-09-15T18:00:34 some day i'll get used to this thing and possibly help someone else =P 2011-09-15T18:00:44 ;) 2011-09-15T18:01:12 *** enhancin has quit IRC 2011-09-15T18:13:21 tsbere: the staff client auto-update, it doesn't handle xulrunner upgrades whether it's doing things incrementally or fully, is that correct? Or will a full update pull over a new xulrunner too? 2011-09-15T18:13:44 phasefx: It should xulrunner update on *both* :D 2011-09-15T18:13:55 ah, I had it in my head that it would not 2011-09-15T18:14:00 that it was just chrome files 2011-09-15T18:14:49 thanks 2011-09-15T18:15:45 *** akilsdonk has quit IRC 2011-09-15T18:16:03 The new layout has xulrunner INSIDE the app dir, instead of outside. That allows xulrunner to update itself. 2011-09-15T18:16:46 ah, gotcha 2011-09-15T18:29:31 how would you guys answer this, "Evergreen Roadmap gives information about Acquistions and Serials -- are these usable? 2011-09-15T18:29:38 yes. 2011-09-15T18:30:10 for comparison, they're on Horizon 7 2011-09-15T18:48:46 I would give you information from a former Horizon 7 customer.....if I thought we ever USED Horizon 7 Acquisitions or Serials 2011-09-15T19:01:42 *** sal_ has quit IRC 2011-09-15T19:10:39 maybe I'll let dbs answer that one :) 2011-09-15T19:26:30 *** bjwebb has joined #evergreen 2011-09-15T19:29:44 *** bjwebb_ has joined #evergreen 2011-09-15T19:57:32 *** banjolibrarian has joined #evergreen 2011-09-15T19:58:23 *** bjwebb_ has quit IRC 2011-09-15T19:58:34 Can anyone tell me where the "profile group" values are stored? In other words, where do I find out what each number refers to (51 for 'adult patron', for example)? 2011-09-15T19:59:21 Forgot to add - in version 1.6.1.8 2011-09-15T20:03:23 permission.grp_tree 2011-09-15T20:04:50 tsbere: Thanks much! 2011-09-15T20:21:00 eeevil / dbwells: Pushed a commit to my grace period extend branch as a "make grace periods obey backdated checkin" 2011-09-15T20:55:20 *** banjolibrarian has quit IRC 2011-09-15T21:20:12 *** atz__ has joined #evergreen 2011-09-15T21:23:43 *** atz_ has quit IRC 2011-09-15T21:27:25 *** agJohn has quit IRC 2011-09-15T21:34:37 *** bjwebb has quit IRC 2011-09-15T21:37:19 *** artunit has joined #evergreen 2011-09-15T22:12:57 *** elene has joined #evergreen 2011-09-15T22:13:26 *** banjolibrarian has joined #evergreen 2011-09-15T22:29:34 *** banjolibrarian has quit IRC 2011-09-15T22:32:31 oh good. i just went from 0.38s to 0.06s per request -- amazing what re-using an https connection will do for performance. 2011-09-15T22:39:21 heh 2011-09-15T22:40:23 * tsbere has finished his "make backports obey grace periods" attempt and re-worked his "suppress transits" based on dbwells' input 2011-09-15T22:40:28 er 2011-09-15T22:40:32 s/backports/backdates/ 2011-09-15T22:40:49 * tsbere shouldn't have relied on any form of autocomplete when it prompted on that one 2011-09-15T22:55:11 *** dbs has joined #evergreen 2011-09-15T23:00:15 *** elene has quit IRC 2011-09-15T23:00:53 *** elene has joined #evergreen 2011-09-15T23:51:35 *** dbs has quit IRC