2010-04-27T01:07:51 *** mck9 has left #evergreen 2010-04-27T01:12:34 *** pmplett has quit IRC 2010-04-27T03:40:14 *** atz has quit IRC 2010-04-27T03:41:02 *** atz has joined #evergreen 2010-04-27T08:00:31 *** mck9 has joined #evergreen 2010-04-27T08:51:46 *** dkyle-mac has joined #evergreen 2010-04-27T09:08:23 *** jenny has joined #evergreen 2010-04-27T09:14:08 *** StephenGWills has joined #evergreen 2010-04-27T09:14:15 *** leed has quit IRC 2010-04-27T09:15:09 *** leed has joined #evergreen 2010-04-27T09:23:46 *** sfortin has joined #evergreen 2010-04-27T09:32:20 haha thanks for the lolz berick 2010-04-27T10:02:19 *** dbs has joined #evergreen 2010-04-27T10:14:30 *** moodaepo has joined #evergreen 2010-04-27T10:20:50 *** bshum has joined #evergreen 2010-04-27T10:21:03 * phasefx is going to have download some more MIDI's now 2010-04-27T10:26:52 phasefx: http://www.kids-and-games.com/midi/T/TARKUSG.MID is one of the best MIDI conversions I've heard 2010-04-27T10:32:35 *** sfortin has quit IRC 2010-04-27T10:33:37 *** Callender has quit IRC 2010-04-27T10:34:09 *** sfortin has joined #evergreen 2010-04-27T10:47:53 *** brendan_bywater has quit IRC 2010-04-27T10:57:10 *** Callender has joined #evergreen 2010-04-27T11:00:06 *** brendan_bywater has joined #evergreen 2010-04-27T11:05:43 _bott_: thanks to you and Marla for the wiki index page to presentations - I added my lightning talk slides there, too 2010-04-27T11:07:05 <_bott_> of course, my ulterior motive is, that I won't need to maintain evergreen2010.org indefinitely 2010-04-27T11:08:01 dumb question..... the gateway log is owned by www-data and the osrf.log is owned by opensrf. are these the logs of the http-transport and xmpp connectors respectively? 2010-04-27T11:30:28 _bott_++ 2010-04-27T11:33:03 StephenGWills: basically, yes 2010-04-27T11:44:01 thanks 2010-04-27T11:50:51 ah yes, I was looking for slides on evergreen2010.org! time to peruse the wiki :) 2010-04-27T11:51:48 wjr: http://evergreen-ils.org/dokuwiki/doku.php?id=eg10 2010-04-27T11:51:56 dbs: thanks =) 2010-04-27T11:52:10 I think _bott_ did 99% of the work :) 2010-04-27T11:52:21 _bott_: also thanks ;) 2010-04-27T11:52:54 <_bott_> I'm just putting content out there. Others did the work 2010-04-27T11:55:19 :P 2010-04-27T12:08:32 *** davidWalker has joined #evergreen 2010-04-27T12:09:29 hi all, which version of the staff client works with the development or demo server? 2010-04-27T12:11:54 davidWalker: checking the /xul directory will tell you what builds a server supports. demo.gapines.org is running 1.2, and supports up to rel_1_2_4_0 2010-04-27T12:12:13 dev.gapines.org is running 1.4 and supports up to 1_4_0_7 2010-04-27T12:12:25 for demo, you can use this one: http://www.evergreen-ils.org/downloads/evergreen-setup-rel_1_2_4_0.exe 2010-04-27T12:12:40 actually, dev.gapines.org might be running something closer to trunk? checking... 2010-04-27T12:13:55 nevermind, it appears i was correct the first time. dev claims to be running 1.4, and supports builds up to rel_1_4_0_7. 2010-04-27T12:14:21 jeff, phasefx: thanks! 2010-04-27T12:14:38 is 1.6 significantly different from 1.4 ? 2010-04-27T12:14:50 wjr: and... you were AT the conference, so the slides are also on your flash drive -- but not the lightning talks. 2010-04-27T12:14:56 and I'm sure dev is running branch rel_1_4 from some nebulous point in time, and the 1.4.0.7 symlink was added for convenience 2010-04-27T12:15:08 phasefx: *nod* 2010-04-27T12:15:17 version call returns 1-4-0 2010-04-27T12:16:57 davidWalker: running a virtual server could let you test 1.6 2010-04-27T12:17:09 I'd say there are big differences between 1.4 and 1.6 2010-04-27T12:17:12 1.6 is significantly different from 1.4, yes 2010-04-27T12:34:58 *** r123 has joined #evergreen 2010-04-27T12:46:56 phasefx: yeah, I've got a group here looking to evaluate Evergreen, and I'd like to avoid having everyone install a virtual server locally. One of our guys tried setting Evergreen up on a virtual Red Hat server, but couldn't get it running, unfortunately. 2010-04-27T12:47:31 so I was looking at using client against the dev server as Plan B 2010-04-27T12:47:42 another option would be to get a virtual server up and running somewhere where they can all access it. 2010-04-27T12:48:32 jeff: yeah, that was what the guy who tried to install it on Red Hat was trying to do 2010-04-27T12:49:23 smart guy, very capable of this stuff. But ran into problems I think with OpenSRF, or something like that 2010-04-27T12:52:32 there are ready-made virtual server images, so it'd just be a matter of networking, not installing Evergreen 2010-04-27T12:55:49 urgh, miker_ ... looks like we might have missed bringing http://svn.open-ils.org/trac/ILS/changeset/15825/branches/rel_1_6_0/Open-ILS/src/sql into 1.6.0.3-1.6.0.4-upgrade 2010-04-27T12:55:57 phasefx: yeah, they tried that too, ran into problems running it on VMWare. They needed to convert it from ovf tol .vmx, but couldn't. 2010-04-27T12:56:09 * davidWalker no expert 2010-04-27T12:56:11 davidWalker: ughers 2010-04-27T12:56:43 * dbs should probably stop trusting VirtualBox's "Export Virtual Appliance" option, which is supposed to support VMWare and VirtualBox (but apparently just causes trouble for everyone) 2010-04-27T12:56:51 phasefx: they've really given it the old college try. Now we're looking at other options. 2010-04-27T12:57:32 davidWalker: hmm. Red Hat is a pretty fringe OS here in Evergreen-land, although there were patches a few months ago to add support for it 2010-04-27T12:57:47 the DIG folks have a 1.6 server, though I'm not sure how open it is 2010-04-27T12:57:54 dbs: the one problem i encountered was that the 1.6.0.1 image had existing udev rules for the MACs that you had, but the import to virtualbox gave it new macs, so instead of eth0/eth1 i had eth2/eth3. fixing it is a matter of removing the udev rules before making the image available, though. 2010-04-27T12:59:04 dbs: yeah, we figured as much. They found and tried some Red Hat instructions that were on the web, and actually got it running, but it would crash and burn after a short time. Unfortunately, we only have Red Hat available to us at the moment for this. 2010-04-27T12:59:24 and the VMWare, but that didn't work either 2010-04-27T13:00:15 jeff: ah, man... I used to do that (remove the udev rules) way back in my early images. Must have lost that rule from my "prep image for virtual export" script. Sigh 2010-04-27T13:01:44 <_bott_> heh just had to explain that udev rule thing to someone yesterday 2010-04-27T13:03:26 you can hardcode MACs in the ovf, but that will only cause other problems later when someone tries to start two on the same layer 2... 2010-04-27T13:04:44 we need to support the davidWalkers + associated organizations of the world 2010-04-27T13:06:51 is running demo.gapines.org and dev.gapines.org (with the mission of something like "keep the latest 2 stable versions of Evergreen running on demo and dev respectively") another task that we can delegate to the community? 2010-04-27T13:07:03 I'm willing to upgrade dev barring any reasons not to. There was some reason or another last time I itched to do it.. I think the hardware is due to be replaced 2010-04-27T13:07:38 and acq.open-ils.org is sort-of trunk, although I'm not sure what state it is in these days 2010-04-27T13:08:03 would go hand-in-hand with the testing.esilibrary.com stuff, I think - just spread out across multiple branches 2010-04-27T13:20:29 i accidentally ordered an ipad 2010-04-27T13:20:37 "oops" 2010-04-27T13:21:08 * dbs has hammers and cement floors readily available for wjr 2010-04-27T13:21:56 I think i'm going to repurpose several square feet of bubble wrap as a case... 2010-04-27T13:22:30 I'm also going to carry around a few pairs of rubber gloves, in case anyone wants to use it. 2010-04-27T13:22:36 oh wait... :P 2010-04-27T13:24:16 wow that just sounded wrong 2010-04-27T13:24:17 *** bshum has quit IRC 2010-04-27T13:24:22 *** bshum has joined #evergreen 2010-04-27T13:24:53 wjr++ 2010-04-27T13:26:20 -= THIS MESSAGE NOT LOGGED =- 2010-04-27T13:26:39 -= THIS MESSAGE NOT LOGGED =- 2010-04-27T13:28:38 jeff: heh - if I could convince them to exchange it, then it would still be free. but the google, they're not very big on the communication. or apparently, on the marketing brains in this particular move 2010-04-27T13:29:42 They went to a lot of expense to ship me a kick-butt smartphone that I can only use as a wifi device. Not exactly the best way to get me to show it off to people... 2010-04-27T13:32:42 *** Meliss has joined #evergreen 2010-04-27T13:34:11 Or maybe they're geniuses, and this way they force recipients like me to sell it at a discounted price to someone else (benefiting the person who would otherwise be paying full price + taxes + shipping) so that we can then purchase our own full-price N1. One gift seeds two adopters? 2010-04-27T13:34:21 Man. We should give away Evergreen! 2010-04-27T13:36:51 i have a demo on a floppy 2010-04-27T13:42:10 dbs - I setup a dedicated SIP server as we discussed doing last week, it works perfectly now. Thanks for talking to me about this issue 2010-04-27T13:42:30 atheos++ # would be good to document, too! 2010-04-27T13:43:15 I'll make a note on the Wiki 2010-04-27T13:49:21 *** jenny1 has joined #evergreen 2010-04-27T13:49:30 *** jenny1 has left #evergreen 2010-04-27T13:51:54 *** jenny has quit IRC 2010-04-27T13:59:00 phasefx: not to press the issue or anything, but if you do plan on upgrading the dev server to 1.6, by all means let me know. Otherwise, I think we're on to Plan C here. 2010-04-27T14:46:41 *** sfortin has quit IRC 2010-04-27T14:47:28 *** mrpeters-isl has quit IRC 2010-04-27T14:47:34 *** mrpeters-isl has joined #evergreen 2010-04-27T14:49:05 equinox++ # for the shared report templates 2010-04-27T14:51:46 *** pmplett has joined #evergreen 2010-04-27T14:54:50 * mrpeters-isl agrees 2010-04-27T14:54:55 equinox++ 2010-04-27T14:55:00 nice job shae, steve and others 2010-04-27T15:05:55 davidWalker: dbs: I'm going to try putting trunk on dev. Should be compatible with 1.6 clients, but will likely roll a trunk client (with a timestamp build id) as well 2010-04-27T15:06:11 phasefx: sounds good to me 2010-04-27T15:06:38 dev > acq 2010-04-27T15:07:04 hrmm 2010-04-27T15:07:08 phasefx: awesome. Know when you might get that done? No rush, just so I might tell folks here. 2010-04-27T15:07:17 but I think dev and demo may share the same postgres instance 2010-04-27T15:07:34 trunk would likely need an updated postgres, but I don't think demo could survive one 2010-04-27T15:07:37 phasefx: maybe tackle acq.open-ils.org then? it looks untouched for months 2010-04-27T15:12:05 davidWalker: I think end of day tomorrow is a safe bet, for whatever I end up doing 2010-04-27T15:12:28 phasefx: wow, that's quick. thanks! 2010-04-27T15:15:09 *** Dyrcona has joined #evergreen 2010-04-27T15:19:55 ISBN 10/13 discussion! 2010-04-27T15:22:11 jeff: ooh, looking to normalize? 2010-04-27T15:23:07 that's been on my backburner (on my 40-element stove) for quite some time. along with LC call number normalization. 2010-04-27T15:23:25 Business::ISBN 2010-04-27T15:26:38 normalize as having both? 2010-04-27T15:28:40 * eby checks his email 2010-04-27T15:29:06 eby: right 2010-04-27T15:29:48 isbn's here are a wasteland 2010-04-27T15:30:09 ones that do have one recorded are invalid 2010-04-27T15:30:20 good chunk i mean 2010-04-27T15:30:20 i would like to suggest normalization at ingest, but miker may beat me up. 2010-04-27T15:30:30 normalization at ingest would make sense 2010-04-27T15:30:53 another option would be to normalize all searches to 13 digit 2010-04-27T15:31:18 then your general keyword searches wouldn't necessarily match 2010-04-27T15:31:23 that could be done quickly in the js -- probably quicker than an ingest change. both may have advantages. 2010-04-27T15:31:26 or identifier|isbn 2010-04-27T15:31:38 (or keyword|identifier that we currently have) 2010-04-27T15:31:41 * jeff nods 2010-04-27T15:32:07 for a keyword search, can't always assume that 10 digits + passes checksum == "isbn, make it 13!" 2010-04-27T15:32:40 jeff: normalization at ingest is fine! (in trunk!) just propose a normalizer that takes ISBN lookin' strings and turns them into both ISBN10 and ISBN13 in one string 2010-04-27T15:33:56 I would also be in favour of augmenting the bib itself, although you would want a switch for that 2010-04-27T15:34:14 dbs: agreed, although there are specific rules 2010-04-27T15:34:26 i.e., if incoming bib has an 020 with an ISBN-10 2010-04-27T15:34:31 can replace it with two 020s 2010-04-27T15:34:38 first having ISBN-13, second having the origin ISBN-10 2010-04-27T15:34:53 and any trailing text in the $a and other subfields (e.g., $c) should be preserved 2010-04-27T15:35:04 gmcharlt: sure 2010-04-27T15:35:30 isbn_trailing_text-- 2010-04-27T15:36:56 davidWalker: http://demo.evergreencatalog.com/ I don't get credit for this, been in the works at Equinox and it just now became ready. Still need to go through some things before we advertise it more widely, but you're welcome to sign in and play around. admin/open-ils. Looks like you should grab the 1.6.0.3 client for it 2010-04-27T15:37:36 phasefx++ # I'll give you credit nonetheless 2010-04-27T15:38:15 it's a smallish box, not meant for stress testing 2010-04-27T15:38:40 * eby has plenty of $2.95 transposed into isbn 2010-04-27T15:40:03 phasefx, do you have a patron test account for the demo catalog? I'd like to see the my account section, and it's prompting me to change the admin password. (don't worry, I won't!) 2010-04-27T15:40:35 atheos: I don't think there are any other accounts yet, but feel free to register one :) yeah, admin isn't good for testing the opac 2010-04-27T15:40:45 will do, thanks! 2010-04-27T15:40:49 register one or as many as needed 2010-04-27T15:41:54 I think an account named after each stock profile/perm group and belonging to namesake would be dandy 2010-04-27T15:42:09 I missed some of the information at the conference last Friday, but did briefly hear that the OPAC will be accepting credit cards. Does this demo have that feature? 2010-04-27T15:43:15 I'm eagerly anticipating the day I can remove this from our website and direct patrons to the OPAC https://www.zionsville.lib.in.us/hmmpl/page/bills 2010-04-27T15:43:20 atheos: it's only 1.6.0.3, so I don't think so 2010-04-27T15:43:50 atheos: i have a link from the Fines tab in our opac that takes the patron to our credit card payment interface without needing to log in a second time. 2010-04-27T15:43:55 phasefx: I'm in, this is great, thanks so much! 2010-04-27T15:44:07 equinox++ 2010-04-27T15:44:55 *** StephenGWills_ has joined #evergreen 2010-04-27T15:47:08 jeff - we are the only library in our system taking CC's at the moment, so I didn't go a similar route. In July, we have new legislation that will allow our libraries to accept payments on behalf of other libraries. At that time, I'd like to use a single credit card processing account and make this feature system wide. I'm guessing this feature is on the list of things King County requires before their launch, so I'll ride on their coattails if it's avai 2010-04-27T15:48:05 atheos: even if you're all on one skin, you should be able to make the display of the button in the opac depend on the patron home library, or similar. 2010-04-27T15:48:13 atheos: your last truncated at "coattails if it's avai" 2010-04-27T15:48:35 able by July :) 2010-04-27T15:48:40 atheos: aha :) 2010-04-27T15:48:54 *** StephenGWills has quit IRC 2010-04-27T15:48:54 *** StephenGWills_ is now known as StephenGWills 2010-04-27T15:48:56 atheos: i've been talking with catherine regarding the payments-for-other-libs thing. 2010-04-27T15:48:58 fwiw, we'll be interested in that too - albeit with a Canuck flavour and I've suggested to our peeps that summer 2011 is a realistic time frame :) 2010-04-27T15:49:07 atheos: will you go to 1.6.1 before july? 2010-04-27T15:50:27 we currently accept payments for any library in our system and we will be sending monies for item replacement charges to the proper library. 2010-04-27T15:50:49 jeff - if credit card acceptance is available in the OPAC for 1.6.1, that will be one of the deciding factors on us upgrading. I don't make those decisions though, I just make recommendations. 2010-04-27T15:51:21 * dbs curses his slippery ethernet cable 2010-04-27T15:51:22 jeff - we would, but our current law restricts us from doing so. 2010-04-27T15:52:56 we're going to be having all payments routed to the state library, and reconsiliation reports ran monthly for each library 2010-04-27T15:54:16 mrpeters-isl - I'm assuming that since I haven't heard anything about it, you guys are bidding out the credit card processing instead of setting up an account directly with authorize.net ? 2010-04-27T15:55:19 * mrpeters-isl doesnt know yet 2010-04-27T15:58:03 *** Meliss has quit IRC 2010-04-27T15:58:25 I am trying to find the code that runs the first time that a patron logs in via the OPAC. 2010-04-27T15:59:02 I *think* that I have found bits and pieces of the process, but my puzzle is missing a few pieces. 2010-04-27T16:02:32 *** clem_munns has joined #evergreen 2010-04-27T16:02:45 hello 2010-04-27T16:03:13 Any pointers on where to look? 2010-04-27T16:03:24 Dyrcona: What pieces are you looking for in particular? 2010-04-27T16:03:29 *** jamesrf has quit IRC 2010-04-27T16:03:29 Out of curiosity :) 2010-04-27T16:08:29 I want to know what code executes when a patron logs in for the first time and the OPAC sees this, and requests that the patron change their password. 2010-04-27T16:09:26 That's javascript isn't it? (just musing, I'm not fully familiar with it either) 2010-04-27T16:09:41 I also want to know how the last_xact_id that is then put in actor.usr is generated. It looks like it might be somehow tied to the OpenSRF session id, but not sure. 2010-04-27T16:10:34 Well, I can't find anything in the JavaScript, nor in the perlmods. Perlmods actually looked a bit more promising, 'cause I can some things going on in Storage that look like they actually update the patron record with the current session id. 2010-04-27T16:13:10 What I'm not finding is the code that determines that this is the usr's first login. 2010-04-27T16:13:21 That's an interesting thought. 2010-04-27T16:13:39 Trying to see if you can somehow set it in the database or something so users don't *have* to change it right away? 2010-04-27T16:13:42 (post migration) 2010-04-27T16:14:35 Yes, that may be the ultimate goal of this. It depends on what others here decide. 2010-04-27T16:15:06 Dyrcona: that's not the opac seeing that the patron is logging in for the first time 2010-04-27T16:15:06 that's just a regex that says "this password isn't secure enough, you've got to change it" 2010-04-27T16:15:27 Dyrcona: by "first login" do you mean "this workstation needs to be registered"? or what dbs said? 2010-04-27T16:15:53 what dbs, said, but the message indicates it is the first time the usr logged in. 2010-04-27T16:16:27 Dyrcona: is this a customized Evergreen install, or an out-of-the-box install? 2010-04-27T16:16:37 I do remember seeing from regex looking at passwords. 2010-04-27T16:17:04 Currently it is an out of the box install of r16318 from trunk. 2010-04-27T16:17:22 or there abouts... 2010-04-27T16:17:26 sorry to interrupt, but I was wondering if I might get a quick bit of advice? 2010-04-27T16:18:29 clem_munns - don't ask to ask, just throw it out there 2010-04-27T16:19:24 alright, getting this error: Unable to communicate with service open-ils.auth 2010-04-27T16:19:37 when trying the test admin login after install 2010-04-27T16:20:13 ran settings-tester.pl and got an OK on all fronts... 2010-04-27T16:22:02 clem_munns can you login via srfsh ? 2010-04-27T16:22:31 yes, that's where I'm getting the error, after #srfsh 2010-04-27T16:22:52 also, when I first installed it, all was great, but I shut it 2010-04-27T16:23:10 down until I had time to come back and now no auth 2010-04-27T16:23:51 did you try osrf_ctl.sh -a start_all ? 2010-04-27T16:23:57 clem_munns - did you have any errors when you started services via osrf_ctl.sh ? 2010-04-27T16:25:04 yes to start command, no to errors, right now I'm looking at the find errors wiki page, let me run those commands now 2010-04-27T16:25:14 brb 2010-04-27T16:26:54 back to my question: I basically see this message because we're using last 4 digits of a patron's phone number for their password? 2010-04-27T16:28:00 Because of the message content and the change in the last_xact_id, I thought it had something to do with the last_xact_id. 2010-04-27T16:28:14 *** brendan_bywater has quit IRC 2010-04-27T16:29:11 exit 2010-04-27T16:29:26 I have found code in OpenILS/Application/Actor.pm that does passwd checking, but that looks like it runs when the passwd is reset. Is that the same code that runs at login in the OPAC? 2010-04-27T16:30:22 Dyrcona: the "this is the first time you've logged in, please change your password" is actually "your password is weak, please change it" 2010-04-27T16:30:26 and it is located... 2010-04-27T16:31:11 That sounds like it would be some sort of locale wording. 2010-04-27T16:31:18 skin/default/js/sidebar.js -- sub strongPassword 2010-04-27T16:31:36 skin????? didn't think to look under skin. :) 2010-04-27T16:31:39 thanks. 2010-04-27T16:32:22 function strongPassword, even 2010-04-27T16:32:42 k, I get a NameVirtualHost*:80 has no virtual hosts warning on the restart of apache2 2010-04-27T16:32:45 there's also function loginDance, which is where it gets called from and shows the change_pw_table if it's weak. 2010-04-27T16:33:01 clem_munns: harmless in most configurations. 2010-04-27T16:33:12 that's what I wuz thinkin 2010-04-27T16:33:25 clem_munns: we get that all the time, 'cause we're lazy. :) 2010-04-27T16:33:44 :-D i know lazy 2010-04-27T16:33:55 jeff: just saw loginDance() and laughed. 2010-04-27T16:34:10 Dyrcona: the offending message is found in Open-ILS/web/opac/locale/en-US/opac.dtd, if you wanted to change it to something like "your password sucks, you should make it harder to crack" 2010-04-27T16:34:10 jeff: at the name of the function, not at its content. 2010-04-27T16:34:22 dbs: thanks. 2010-04-27T16:34:33 bshum: there's an entity string for it, &login.first.time and &common.password_criteria 2010-04-27T16:34:42 clem_munns: does "ps wax | grep open-ils.auth" show any services running? 2010-04-27T16:34:53 jeff: That's what I thought. Thanks :) 2010-04-27T16:35:13 You will need to change your password."> 2010-04-27T16:35:42 get this: 3164 pts/0 R<+ 0:00 grep open-ils.auth 2010-04-27T16:35:51 so, open-ils.auth isn't running 2010-04-27T16:35:53 *** jamesrf has joined #evergreen 2010-04-27T16:36:23 at the start command I note that it isn't one of the mods loaded 2010-04-27T16:36:30 which means that your opensrf.xml config file might not have a hostname matching 2010-04-27T16:36:54 open-ils.auth is a C service, and those aren't enumerated at start-up time 2010-04-27T16:37:20 one dum q: do I have to redo the config xml files everytime I start up and shutdown? 2010-04-27T16:37:21 perhaps you meant to specify "localhost" as the hostname using the "-l" flag for osrf_ctl.sh? 2010-04-27T16:37:48 I did use the -l flag, yes 2010-04-27T16:38:00 nope. but if your hostname changes for some reason, and you're not forcing OpenSRF to treat the hostname as "localhost" using -l, then that would be a problem 2010-04-27T16:38:29 other possibilities are running out of memory, so the C services got killed off... 2010-04-27T16:38:42 it's test system and it'll always stay localhost 2010-04-27T16:39:28 you could try "osrf_ctl.sh -l -a restart_c" to see if they get started up this time 2010-04-27T16:39:35 *** brendan_bywater has joined #evergreen 2010-04-27T16:40:28 jeff: Dyrcona: that function you mentioned strongPassword, it seems to check for a setting first. This is likely tied to the library settings value for password format I'm guessing? Otherwise, it seems to default down to 7 characters, and some other regular expression stuff. 2010-04-27T16:41:01 bshum: that is correct. 2010-04-27T16:41:39 if you set the password regex library setting to "z", then it will require that you password contain a "z" 2010-04-27T16:41:50 Aha 2010-04-27T16:41:59 still no auth mod after restart, but a bunch of "kill ... no such process" msgs 2010-04-27T16:42:02 all passwords should require at least one z 2010-04-27T16:42:28 atheos: and it should be pronounced "zed" 2010-04-27T16:42:45 zeds dead baby 2010-04-27T16:42:54 clem_munns: do you have an open-ils.circ process? 2010-04-27T16:43:10 yes 2010-04-27T16:43:41 sounds like the C services are dying viciously for some reason then 2010-04-27T16:44:27 one thing, when I do the start command it says they are already running even though I've stopped them prior to now start command 2010-04-27T16:44:48 clem_munns you might need to wipe your pid files 2010-04-27T16:45:01 kewl, HOWTO? 2010-04-27T16:45:01 *** dkyle-mac has quit IRC 2010-04-27T16:45:03 So I guess in that case, there's no magic behind password changing at the start, other than the fact that the defaults that are loaded aren't fitting to the regex requirements for strong passwords. 2010-04-27T16:45:20 rm /openils/var/run/*.pid 2010-04-27T16:45:29 clem_munns check in /var/run/evergreen/pid - remove whatever's there 2010-04-27T16:45:42 as root or opesrf? 2010-04-27T16:45:43 or what atz says. you know, wherever they are :) 2010-04-27T16:45:46 But if we chose, we could try implementing an org specific setting that lowered the password strength (not that we should, but it is possible) 2010-04-27T16:46:00 clem_munns opensrf 2010-04-27T16:46:06 lol 2010-04-27T16:46:17 yeah - I suppose if open-ils.auth previously had a PID that is now attached to a process that opensrf user can't kill (like X.org or something like that), then that would explain the problem 2010-04-27T16:46:25 bshum: yep 2010-04-27T16:46:29 bshum: jeff: dbs: we currently use last 4 digits of a patron's phone number. not saying that won't change when we go to Evergreen, but I will need to give the committee options. :) 2010-04-27T16:46:34 password_regex = "" 2010-04-27T16:46:34 *** grantjohnson has left #evergreen 2010-04-27T16:46:46 okay, now that i've backported fifo/next-not-nearest to 1.4 and 1.6 from trunk, i should finalize the org unit setting name/etc :P 2010-04-27T16:47:03 Dyrcona: We are also using the last 4 digits (or rather the last four characters, our patron imports have done some quirky things with extensions "x123" for example 2010-04-27T16:47:04 Dyrcona: password_regex="\d{4}" 2010-04-27T16:47:12 Dyrcona: if you change that org unit setting, you can allow patrons to continue to have four digit passwords with no forced change. 2010-04-27T16:47:16 gotta go get a kid from school, will report back with outcome, thanks 4 help 2010-04-27T16:47:18 what dbs said. 2010-04-27T16:47:42 * dbs looks askance at predictable passwords 2010-04-27T16:47:49 dbs: I actually want to force staff to have tougher passwords than patrons. Maybe that'll be a nice patch to send in later when I have this stuff figured out. 2010-04-27T16:48:01 Dyrcona: please do! 2010-04-27T16:48:11 Dyrcona: or you could got the jeff route and prevent staff from changing their passwords at all :) 2010-04-27T16:48:16 hey! 2010-04-27T16:48:28 we don't prevent from changing, we just prefer that they not reset via e-mail. 2010-04-27T16:48:34 dbs: certain accounts, I just may. 2010-04-27T16:48:40 mainly because we do not restrict what staff account email addresses can be 2010-04-27T16:48:53 where dbs does restrict to internal mail system only :) 2010-04-27T16:49:15 my members are going to insist on generic accounts for circulation... so I won't want them changing passwords on those. 2010-04-27T16:49:18 jeff: not yet! but in the delicious future 2010-04-27T16:49:24 there is the issue where staff can't edit their own details in the staff client -- that surprises people sometimes. 2010-04-27T16:49:33 * dbs looks even more askance at generic circ accounts 2010-04-27T16:49:38 is there any problem with the same usr being logged in multiple times simultaneously? 2010-04-27T16:49:50 Dyrcona: there's no restriction that I'm aware of 2010-04-27T16:50:08 Dyrcona: yeah, best thing to do there is to set a "strong" password (where your "stronger for staff" count useful!), and if someone changes it... deal with it then. no technical solution required. 2010-04-27T16:50:27 dbs: I don't like it, but the users don't want to have to login and out every time they move during the day. 2010-04-27T16:50:33 Dyrcona: no issues that i am aware of. we do make use of role accounts in some situations, where multiple users are logged in from multiple workstations. 2010-04-27T16:51:59 Dyrcona: I hear you. We don't allow generic circ accounts, but our circ transactions are rather suspiciously weighted towards just a few of our staff. Surprising, as they're accountable for the cash drawer balance. 2010-04-27T16:52:51 Have a nice day folks, enjoy. 2010-04-27T16:53:01 bshum: enjoy! 2010-04-27T16:53:11 l8r, ben. 2010-04-27T16:53:23 *** bshum has quit IRC 2010-04-27T16:59:36 Whom should I email about getting a dokuwiki account? 2010-04-27T17:00:27 Dyrcona: you can just pm me 2010-04-27T17:00:42 I'll get you sorted - Name, Email address, and wiki username 2010-04-27T17:02:23 dbs: i think i just sent you the info. 2010-04-27T17:03:46 you've got mail (or whenever the SMTP stuff finishes you will) 2010-04-27T17:04:46 dbs: tyvm. 2010-04-27T17:06:30 * Dyrcona waits on the gray list processing..... 2010-04-27T17:06:41 * jeff grins 2010-04-27T17:07:55 *SPAM SPAM SPAMMITY SPAM* 2010-04-27T17:13:51 clem_munns pasted "untitled" at http://paste.lisp.org/display/98463 2010-04-27T17:14:34 did the grep ejabberd and got some errors 2010-04-27T17:18:17 clem_munns: looks okay to me 2010-04-27T17:18:32 false bit means nadda 2010-04-27T17:20:01 clem_munns: matches what I have on my test server 2010-04-27T17:20:28 dbs annotated #98463 "same command on my server" at http://paste.lisp.org/display/98463#1 2010-04-27T17:20:51 so tiz 2010-04-27T17:21:39 so, to start it up from a rebooted box, what first? 2010-04-27T17:24:38 if you didn't shut down cleanly, i would do: 2010-04-27T17:25:10 (as opensrf): rm -f /openils/var/run/* ; osrf_ctl.sh -l -a start_all 2010-04-27T17:25:22 (as root): /etc/init.d/apache2 reload 2010-04-27T17:26:56 *** Dyrcona has quit IRC 2010-04-27T17:26:56 is "shutdown -h now" considered clean? 2010-04-27T17:28:11 btw: after running those commands, the list of mods started still does not include the open-ils.auth service 2010-04-27T17:29:49 it worked! 2010-04-27T17:30:13 clem_munns: might be a good time to run "free" to see how much memory you have left 2010-04-27T17:30:24 from srfsh? 2010-04-27T17:30:38 clem_munns: no, just from the Linux command line 2010-04-27T17:31:01 dbs annotated #98463 "My free memory..." at http://paste.lisp.org/display/98463#2 2010-04-27T17:31:04 3 gigs 2010-04-27T17:31:11 free, 1 used 2010-04-27T17:31:21 huh. okay, that should certainly be plenty 2010-04-27T17:31:37 * dbs doesn't know why your C services wouldn't start before, but oh well :) 2010-04-27T17:31:50 evergreen is the only thing running on the box 2010-04-27T17:32:33 other than SSH 2010-04-27T17:33:39 *** davidWalker has left #evergreen 2010-04-27T17:33:52 now to install the client 2010-04-27T17:43:39 jeff: do you have slides for all that stuff that you skimmed over in your lightning talk? 2010-04-27T18:01:52 * dbs closes eyes tightly to avoid seeing all of the inline CSS in craftsman 2010-04-27T18:02:35 dbs: you’re going to need a screenreader.... 2010-04-27T18:05:00 has anyone played with ample? 2010-04-27T18:06:07 * dbs knows not what ample is 2010-04-27T18:06:19 cross browser xul rendering among other things http://www.amplesdk.com/about/compatibility/ 2010-04-27T18:06:49 jamesrf: i have slides and the slides were left on the laptop used for the presentation, so in theory grpl has 'em. i'll see that they get posted one way or another. 2010-04-27T18:07:08 jamesrf: send me your email address and i'll make sure you get a copy tonight? 2010-04-27T18:07:25 sure thanks jeff 2010-04-27T18:07:37 jeff++ 2010-04-27T18:07:37 a question, do I need to have an opensrf user on my local machine to run the client? 2010-04-27T18:08:02 jeff: you could always post'em on the wiki at http://evergreen-ils.org/dokuwiki/doku.php?id=eg10 2010-04-27T18:08:22 dbs: i'll make sure they get linked there one way or another. 2010-04-27T18:08:26 jeff++ 2010-04-27T18:08:29 * dbs goes home 2010-04-27T18:08:42 *** dbs has quit IRC 2010-04-27T18:43:55 well, I've got the client running but it won't take the admin/open-ils auth pair, anybody got a suggestion? 2010-04-27T18:44:22 meaning: staff client 2010-04-27T18:51:04 nevermind, figured it out...got the 1.6.0.3 client pointing to 1.6.0.4 server version 2010-04-27T18:51:04 clem_munns: did you try logging into srfsh? 2010-04-27T18:51:10 ah 2010-04-27T18:51:29 do I just need to rename that directory? 2010-04-27T18:51:46 or new symlink? 2010-04-27T18:52:03 symlink should be ok 2010-04-27T18:52:16 where do I find that file? 2010-04-27T18:52:30 /openils/var/web/xul 2010-04-27T18:52:32 is it an apache conf file 2010-04-27T18:52:49 you probably have /openils/var/web/xul/rel_1_6_0_4 2010-04-27T18:53:04 you need a symlink pointing to that at /openils/var/web/xul/rel_1_6_0_3 2010-04-27T18:53:27 although might be easier just to download a 1.6.0.4 client 2010-04-27T18:53:45 indeed, where do I get one? 2010-04-27T18:53:50 open-ils.org 2010-04-27T18:54:02 do'h! 2010-04-27T18:54:51 I'm on a Linux machine, though 2010-04-27T18:56:45 nevermind 2010-04-27T19:56:20 miker_: 'round? 2010-04-27T20:00:19 @later tell miker_ I just got around to testing 1.6.0.4 and get the same error r16263 does not seem to have gone through. I checked Open-ILS/src/sql/Pg/1.6.0.3-1.6.0.4-upgrade-db.sql and I still see the materialized.simple_record NOT reporter.materialized_simple_record. 2010-04-27T20:00:19 moodaepo: The operation succeeded. 2010-04-27T20:45:35 *** pmplett is now known as pmpafk 2010-04-27T21:46:06 *** jamesrf has quit IRC 2010-04-27T21:52:54 *** clem_munns has left #evergreen 2010-04-27T22:18:12 *** pmpafk is now known as pmplett 2010-04-27T22:49:30 * jeff yawns 2010-04-27T23:28:45 *** dbs has joined #evergreen 2010-04-27T23:29:50 * dbs sobs as he reads moodaepo's last message 2010-04-27T23:30:04 and also as he posts his nexus one to ebay 2010-04-27T23:31:18 :( 2010-04-27T23:39:50 I'll push an updated 1.6.0.4 to the downloads directory with that fix 2010-04-27T23:40:30 although I'm tempted to just jump straight to 1.6.0.5 so that we can roll in the in-db circ db upgrade that we missed, too 2010-04-27T23:48:28 moodaepo: upgrade script in 1.6.0.4 as-of-now should work 2010-04-27T23:49:20 @later tell miker_ if you decide to rename my version of 1.6.0.4 to 1.6.0.4.1 or something, the original 1.6.0.4 is in the downloads directory with a .bad extension should you want to restore it 2010-04-27T23:49:20 dbs: The operation succeeded.