2010-03-11T00:06:13 *** jamesrf has joined #evergreen 2010-03-11T00:28:51 *** jamesrf has quit IRC 2010-03-11T07:27:45 *** blackrock has joined #evergreen 2010-03-11T07:38:12 *** rsinger has joined #evergreen 2010-03-11T07:47:36 *** blackrock has left #evergreen 2010-03-11T07:55:13 *** mck9 has joined #evergreen 2010-03-11T08:13:52 *** sfortin has joined #evergreen 2010-03-11T08:23:32 *** gdunbar has joined #evergreen 2010-03-11T08:33:55 *** Melissa has joined #evergreen 2010-03-11T08:35:49 What does a Status of -1 mean in regards to viewing a hole for a particular patron? 2010-03-11T08:38:29 atheos: it means the hold and copy are in an inconsistent state, like the hold is captured, but the copy is checked out or some such 2010-03-11T08:39:26 berick - where can I see that in the database? would it be action.hold_transit_copy.copy_status ? 2010-03-11T08:40:18 asset.copy.status on the captured copy 2010-03-11T08:40:54 good deal, thanks berick 2010-03-11T08:41:20 np 2010-03-11T08:53:06 *** Meliss has joined #evergreen 2010-03-11T09:10:19 *** jenny has joined #evergreen 2010-03-11T09:23:16 *** sfortin has quit IRC 2010-03-11T09:28:26 morning, all 2010-03-11T09:28:47 I'm looking to build a staff client for EG 1.4.0.7 on Ubuntu Karmic... 2010-03-11T09:29:20 the wiki says that you run the ./configure like you would for a server build 2010-03-11T09:29:31 (which includes OpenSRF dependencies) 2010-03-11T09:29:40 is that the only way? :-) 2010-03-11T09:31:18 chrissharp123: another option is to grab the "build" folder from an install of the windows packaged version of the 1.4.0.7 client 2010-03-11T09:31:46 huh - well that's something I've not considered before 2010-03-11T09:32:00 jeff: thanks for the tip - I'll try that 2010-03-11T09:32:20 good luck! that method has worked well for me before. 2010-03-11T09:32:46 good - I may end up with more questions - but if that works - it's ideal - thanks again 2010-03-11T09:33:04 *** alxp has joined #evergreen 2010-03-11T09:36:07 *** bshum has joined #evergreen 2010-03-11T09:37:09 jeff: works like a charm - thanks again 2010-03-11T09:37:51 you're welcome! 2010-03-11T09:49:38 *** dbs_doh has joined #evergreen 2010-03-11T09:52:24 whee, brought the wrong laptop to work today 2010-03-11T09:54:08 Oops. 2010-03-11T09:57:21 berick: as I have no SVN access right now, you'll have to fix this typo yourself (heh) 2010-03-11T09:57:24 Open-ILS/web/templates/default/acq/picklist/user_request.tt2 2010-03-11T09:57:30 columnPickerPrefix='"acq.picklist.user_reqeust"' 2010-03-11T09:59:14 dbs_doh: doh! 2010-03-11T10:03:26 thought that might save you some pain :) 2010-03-11T10:08:35 too bad dynamic user settings no longer work in trunk w/ the advent of user setting types 2010-03-11T10:08:56 * berick makes note to add bug to launchpad 2010-03-11T10:10:23 I'm working on a new problem today, dealing with logging in via web browser to view reporter output, causing the staff user account to remain logged into the web browser and breaking the OPAC when visiting it. 2010-03-11T10:10:43 Anybody else know anything about that before I write up a new bug report? 2010-03-11T10:11:50 in what way does the opac break in that scenario? 2010-03-11T10:11:59 Well, so far 2010-03-11T10:12:09 It's leaving the staff user logged in 2010-03-11T10:12:24 So first of all, it exposes the My Account of that user in the OPAC 2010-03-11T10:12:28 i would actually expect that. 2010-03-11T10:12:29 * jeff nods 2010-03-11T10:12:31 Then, when you log out of it 2010-03-11T10:12:32 *** alxp has left #evergreen 2010-03-11T10:12:37 It returns a search depth of -1 2010-03-11T10:12:45 And this leaves the OPAC in a repeating loop 2010-03-11T10:12:54 oops. 2010-03-11T10:12:56 Where all searches fail and return you back to the start with -1 2010-03-11T10:13:13 Depth = -1 rather 2010-03-11T10:13:34 The only way to clear it so far for me was to clear my browser's history, etc. 2010-03-11T10:13:41 And then I could visit again as normal. 2010-03-11T10:13:47 But, that seems... bad. 2010-03-11T10:14:10 setting an explicit depth in the url and/or clearing cookies would probably fix it, also. both are just workarounds, of course. 2010-03-11T10:14:25 So I guess it would be best if the user was logged out automatically, rather than manually, but barring that, it would be better if the OPAC didn't get stuck in that broken loop. 2010-03-11T10:14:35 I actually tried that too. 2010-03-11T10:14:49 Changing the URL string to include a different depth (like 0,1, or 2) 2010-03-11T10:14:56 And it still went back to -1 brokenness. 2010-03-11T10:15:09 So clearing the cookie was the only way to fix it so far for us. 2010-03-11T10:15:42 I should preface this by stating that we're running 1.6.0.2 (modified) 2010-03-11T10:15:51 So I can't confirm if this problem persists elsewhere. 2010-03-11T10:16:04 But so far it's affected both our test and production server. 2010-03-11T10:16:20 sounds like a bug report and a feature request -- the feature request being a way to make logging out of the report view interface obvious/easy, or a way to have the act of logging into the reporter not carry over into the opac... 2010-03-11T10:16:32 but a clever individual in control of the browser can easily leverage one into the other. 2010-03-11T10:16:49 so perhaps back to the "make it obvious that you're logged in, and easy to log out" 2010-03-11T10:17:43 Cool, thanks for the suggestion on how to proceed Jeff. 2010-03-11T10:19:38 good luck. :) 2010-03-11T10:19:43 bug_reports++ 2010-03-11T10:27:45 Peachy, I just checked it against our old 1.4 server and the same problem is there too. 2010-03-11T10:33:00 Does anyone know how/who developed the Craftsman skin? 2010-03-11T10:34:34 bshum: the symptoms of this are similar to those of an earlier Safari bug, which was fixed in this changeset: http://svn.open-ils.org/trac/ILS/changeset/12607 2010-03-11T10:35:02 jeff: aha, interesting 2010-03-11T10:35:30 So they set it for the cookies to just expire, it seems, rather than be removed. 2010-03-11T10:38:22 bshum: logging into the reporter sets a ses cookie with a path of /, while logging in via the opac sets it at /opac/en-US/skin/default/xml/ -- that's why this is breaking. 2010-03-11T10:41:08 *** BobW has joined #evergreen 2010-03-11T10:43:38 *** brendan_bywater has joined #evergreen 2010-03-11T10:43:51 jeff: That sounds bad. 2010-03-11T10:44:57 when you click "log out" in the opac, it makes an api call to delete the session from the server, and it tries to delete the ses cookie in the client using javascript. due to the path mis-match, the ses= cookie with path=/ sticks around. 2010-03-11T10:45:33 later, when trying a search/etc, the client browser sees that it has a ses cookie, and tries to use it, but it's no longer valid... and the error handling at that point is poor, so you end up at the initial opac index page 2010-03-11T10:45:40 and as you saw, you can't search or log back in 2010-03-11T10:48:09 Hmm, that makes sense. 2010-03-11T10:48:29 a quick fix to both problems for you might be to change references of "ses" as a cookie name in OpenILS/WWW/Proxy.pm to reporter_ses or similar. 2010-03-11T10:49:14 oils_proxy_ses 2010-03-11T10:49:36 With the disclaimer of "a clever/determined individual in control of the browser after your report-loving user has departed can leverage one into the other", that would solve the bug and the feature request of "we don't want logging in to view reports to log you into the opac also" 2010-03-11T10:49:50 phasefx: good suggestion 2010-03-11T10:50:39 of course, it wouldn't surprise if the opac one day actually used Proxy 2010-03-11T10:50:40 but there remain underlying issues, such as "the opac should handle 'my session is no longer valid' without breaking" 2010-03-11T10:51:20 * phasefx nods 2010-03-11T10:51:24 at the same time, you could also tweak your proxy ses cookie expiry to be quite short, also. 2010-03-11T11:05:39 jeff, phasefx: Thanks for your help in investigating this issue. I'll check back with the others in our support chain before we try implementing these suggestions. 2010-03-11T11:13:42 *** dbs_doh has quit IRC 2010-03-11T11:18:50 *** dbs has joined #evergreen 2010-03-11T11:19:01 ahhhhh 2010-03-11T11:30:26 *** BobW has quit IRC 2010-03-11T11:35:36 jeff: well, it would paper over the opac issue, but the reason we call the cookie "ses" is because interfaces that we hide behind that expect there to be a ses cookie 2010-03-11T11:35:56 jeff: for instance, you can go directly to the conify interfaces in a browser 2010-03-11T11:36:13 but if the cookie name changes, then that can't happen anymore 2010-03-11T11:37:01 we could change the path of the opac cookie, but it would reduce security in the other direction 2010-03-11T11:40:16 *** jamesrf has joined #evergreen 2010-03-11T11:44:34 yeah, i failed to consider the in-staff-client uses of /reporter/ and the proxy. 2010-03-11T11:45:22 oh, and the underlying interface wanting the ses cookie. i follow. 2010-03-11T11:48:54 the staff client will set the cookie also for use by cookie-using interfaces (the newer dojo interfaces, etc.) 2010-03-11T11:49:48 *** sfortin has joined #evergreen 2010-03-11T11:55:40 jeff: well, the in-SC use too ... because the proxy would pop up 2010-03-11T11:56:24 phasefx: dang, am I drunk already? Did you add an "Edit due time" option somewhere, or was it just adding the "Due time" column to the display? 2010-03-11T12:00:32 not for existing circs, I didn't 2010-03-11T12:01:59 there's a timepicker xul dialog (wraps the date and time picker xul widgets) we can use to replace some of these edit date textboxes 2010-03-11T12:02:41 phasefx: okay. I'm drunk. on one coffee! 2010-03-11T12:03:05 phasefx: cool, that might be my next experiment then. OH SO MANY TODOS. 2010-03-11T12:03:09 XUL_TIMESTAMP_DIALOG => util/timestamp.xul 2010-03-11T12:03:52 suitable for menu entries etc. where you don't want inline widgets 2010-03-11T12:13:58 *** brendan_bywater has quit IRC 2010-03-11T12:24:51 jeff: What if we moved the reports, etc. off our server and somewhere else. If the URL is different, would that mean that the session cookie be different? 2010-03-11T12:25:39 In that sense, maybe we have the OPAC be one URL set, and the report link as something else entirely. (not well versed on how the cookie operates) 2010-03-11T12:27:43 i don't think it would buy you more than it would break. 2010-03-11T12:28:39 the bug can be fixed. the feature request is a bigger issue. would having the session you create by viewing reports time out in 5 minutes meet your needs? 2010-03-11T12:29:25 what is the larger problem created by a reporter login also functioning as an opac login? 2010-03-11T12:29:32 (for you, in your environment) 2010-03-11T12:29:48 Just worried about making sure that there's no unintended consequences. 2010-03-11T12:30:06 So maybe setting the cookie's session time down would be best then. 2010-03-11T12:30:18 For me, I just hate to see bad things happen with the OPAC like that. 2010-03-11T12:30:36 I don't think any of our end users have complained about it. Yet. 2010-03-11T12:30:38 This only happens with staff, right? 2010-03-11T12:30:52 Right because any other login would fail on a reporter page 2010-03-11T12:30:57 Since they wouldn't have the permission to log in. 2010-03-11T12:31:10 We're actually experimenting with using the same method to secure our print notices too. 2010-03-11T12:31:49 That's part of what: http://svn.open-ils.org/trac/ILS/changeset/15777 2010-03-11T12:31:53 This changeset did for us. 2010-03-11T12:32:02 In limiting logins to specific org units 2010-03-11T12:32:22 So we could define limitations on what staff logins for reports or notices could be allowed access for. 2010-03-11T12:32:33 But then when we were testing it, we stumbled on the OPAC issues and got a little scared. 2010-03-11T12:32:53 heh... one larger issue is that a OpenILS::WWW:Proxy login survives a browser exit, because the session cookie has an expiry of 1 hour -- it makes it a persistent cookie, stored on disk between browser sessions. I don't think that was intentional. :-) 2010-03-11T12:34:27 So our only recourse sounds like either inventing a procedure to remove the cookies from web browsers (just in case) or somehow changing how the session cookie is created in the first place? 2010-03-11T12:37:28 bshum: a greasemonkey script or bookmarklet that deletes or modifies the cookie might be a short-term workaround? Honestly, I've never had anyone complain about this here. Then again, not many people run reports here. 2010-03-11T12:38:26 *** brendan_bywater has joined #evergreen 2010-03-11T12:42:24 dbs: That's probably a fair assessment, we weren't sure if it was a truly bad "problem" in a live setting vs. a few of us tinkering in the back. 2010-03-11T12:42:38 dbs: However, a few of the staff machines are shared staff and OPAC for patrons. 2010-03-11T12:42:58 dbs: So I was just trying to sidestep potential problems. 2010-03-11T12:43:29 Use a different browser for the OPAC? 2010-03-11T12:44:12 or http://stackoverflow.com/questions/178263/javascript-bookmarklet-to-delete-all-cookies-within-a-given-domain 2010-03-11T12:51:19 *** pmplett has joined #evergreen 2010-03-11T12:52:08 *** natschil has joined #evergreen 2010-03-11T12:52:38 the opac bug should be fixed regardless. this isn't the first or the last time it'll cause a problem. 2010-03-11T12:53:09 if you're going to be using the oils proxy more for other things, 'fixing' the session cookie to not last between browser sessions would be highly recommended... 2010-03-11T12:53:31 and then you just need to remind/encourage staff to always exit both the staff client and any browser on shared workstations before walking away. 2010-03-11T12:54:37 it also might be a great idea for your staff/public shared machines to have different OS users. 2010-03-11T12:56:10 aggressive timeouts on the staff client might be a good failsafe, too 2010-03-11T12:56:57 different users on the workstation would sidestep the whole issue, no? 2010-03-11T12:57:23 that and making the cookies session-only 2010-03-11T12:57:25 need to make the browser in the client smarter about sessions too.. it'll happily shove a new stale session variable into a page on every page load 2010-03-11T12:59:10 as of IE7, session cookies are not properly cleared at browser close, fwiw 2010-03-11T12:59:45 I'll remove the -expires from the cookie creation for now 2010-03-11T13:00:21 Hopefully IE is banned from any machines dedicated to Evergreen, due to its awesome JS/DOM memory leakage & ability to bring down the operating system 2010-03-11T13:04:53 has to do its part to push folks to buy beefier computers so they can buy windows again 2010-03-11T13:05:16 *** frzosima has quit IRC 2010-03-11T13:05:40 *** frzosima has joined #evergreen 2010-03-11T13:29:25 *** jenny has quit IRC 2010-03-11T13:35:47 *** sfortin has quit IRC 2010-03-11T13:36:22 *** sfortin has joined #evergreen 2010-03-11T13:39:50 *** sfortin has quit IRC 2010-03-11T13:49:21 *** natschil has quit IRC 2010-03-11T13:49:36 *** agJohn has joined #evergreen 2010-03-11T13:51:54 *** natschil has joined #evergreen 2010-03-11T13:53:25 *** jenny has joined #evergreen 2010-03-11T13:56:40 *** natschil has quit IRC 2010-03-11T13:59:10 *** StephenGWills has joined #evergreen 2010-03-11T13:59:54 just when I thought evergreen was going to go legit, they pull me back in! 2010-03-11T13:59:58 :) 2010-03-11T14:01:01 has anyone ever dealt with moving 8.2 data to 8.3 and had tsearch2.so lib functions missing? :( 2010-03-11T14:01:02 what you talking about Willis^H^Hs? :) 2010-03-11T14:02:24 where 8.2 and 8.3 are postgreSQL flavors, btw. 2010-03-11T14:05:17 *** jeff has quit IRC 2010-03-11T14:13:13 StephenGWills: I believe sitka has successfully upgraded from 8.2 to 8.3 2010-03-11T14:14:47 ok, thanks. Don't mind me, I just needed a shoulder to cry on. 2010-03-11T14:15:40 *** jeff has joined #evergreen 2010-03-11T14:37:38 Was the requirement to add contrib/tsearch2.sql to the target 8.3 database first, before restoring the 8.2 database into the 8.3 database? 2010-03-11T14:37:51 * dbs vaguely recalls something like that. jamesrf ? 2010-03-11T14:42:40 dbs: I think we had to do that when we migrated our 8.2 test database to 8.3 2010-03-11T14:43:01 It feels like ages ago now 2010-03-11T14:43:54 Wouldn't you add all those contrib files to the database first. 2010-03-11T14:44:02 Then restore the 8.2 into 8.3 2010-03-11T14:44:06 To "upgrade" it 2010-03-11T14:55:09 *** jenny has quit IRC 2010-03-11T15:32:54 bshum I am all over adding contrib files to 8.3 which has reduced the errors considerably. 2010-03-11T15:33:02 thanks 2010-03-11T15:37:03 StephenGWills: Neat. 2010-03-11T15:39:24 *** Bildsheim has joined #evergreen 2010-03-11T15:44:34 sorry i don't recall, branflakes did the db wrangling for our upgrade 2010-03-11T15:44:39 *** pmplett is now known as pmp_afk 2010-03-11T15:45:59 *** Bildsheim has quit IRC 2010-03-11T15:46:55 er nm sounds like you figured it out 2010-03-11T15:49:12 *** jamesrf has quit IRC 2010-03-11T15:54:27 *** frzosima has quit IRC 2010-03-11T15:58:57 *** natschil has joined #evergreen 2010-03-11T16:00:19 thanks jamessrf... am well on the way. 2010-03-11T16:05:15 *** jenny has joined #evergreen 2010-03-11T16:25:34 *** natschil has quit IRC 2010-03-11T16:26:47 *** Melissa has quit IRC 2010-03-11T16:53:35 *** jamesrf has joined #evergreen 2010-03-11T17:06:48 *** bshum has quit IRC 2010-03-11T17:20:43 Whoa, MVLC is running trunk? That's pretty hardcore 2010-03-11T17:20:55 (even on a test environment) 2010-03-11T17:25:52 dbs: MVLC? 2010-03-11T17:26:21 Merrimack VLC... recent post from Jason Stephenson on open-ils-dev 2010-03-11T17:27:52 Ah. Nice. I keep thinking about setting up a trunk server but never getting around to doing the deed. 2010-03-11T17:31:41 I have a couple of quick questions about setting up fines/default fines. 1) How can I setup different default fines for different item types. Example: DVD-$20, Hardcover-$15, Juv Paperback-$10 etc. I just see one default value for ALL items. I could create a new Billing Type but that can only be applied manually(?) 2010-03-11T17:32:27 These default fines would be for lost/replacement costs. 2010-03-11T17:35:20 *** pmp_afk has quit IRC 2010-03-11T17:43:28 *** dbs has quit IRC 2010-03-11T17:47:12 *** jenny has left #evergreen 2010-03-11T17:51:41 2) With the default fine as it is right now, it seems like the fine applied is the replacement cost of the item as recorded in the price field and the default fine is applied only if the item price is null. Is there anyway to actually have the default fine be applied and over ride the entry for item price? 2010-03-11T18:32:56 *** pmp_afk has joined #evergreen 2010-03-11T19:08:37 *** leed- has joined #evergreen 2010-03-11T20:12:13 *** StephenGWills has quit IRC 2010-03-11T20:42:33 *** pmp_afk has quit IRC 2010-03-11T20:47:25 *** jamesrf has quit IRC 2010-03-11T21:13:43 *** pmp_afk has joined #evergreen 2010-03-11T22:01:37 *** jeff_ has quit IRC 2010-03-11T22:02:18 *** jeff_ has joined #evergreen 2010-03-11T22:40:37 *** jeff_ has quit IRC 2010-03-11T22:53:01 *** Guest24487 has joined #evergreen 2010-03-11T22:56:09 *** pmp_afk has quit IRC