2009-09-28T00:14:14 *** mck9 has left #openils-evergreen 2009-09-28T01:07:11 *** dbs has quit IRC 2009-09-28T05:00:07 *** eguest309 has joined #openils-evergreen 2009-09-28T05:31:19 helo 2009-09-28T05:31:20 good moring 2009-09-28T05:31:40 hows everybady :) 2009-09-28T05:31:44 hope cool 2009-09-28T05:31:47 questin 2009-09-28T05:33:31 how can i modify the circulation scripts so that if someone owes they cant borrow more books or renew their books 2009-09-28T05:34:13 i dont kno whether the fine overdue field in the group _dration script can cater for this? 2009-09-28T07:00:40 *** eguest309 has quit IRC 2009-09-28T07:47:06 eguest309 - give me an email, i will send you a copy of Indiana's circ scripts to use as examples 2009-09-28T07:47:14 we have conditions for many scenarios that may help 2009-09-28T07:53:34 *** mck9 has joined #openils-evergreen 2009-09-28T08:14:39 *** wlayton has joined #openils-evergreen 2009-09-28T08:16:51 *** lisppaste6 has joined #openils-evergreen 2009-09-28T08:40:34 *** diogenes_ has joined #openils-evergreen 2009-09-28T08:43:58 *** senator has joined #OpenILS-Evergreen 2009-09-28T08:59:22 *** Ryanon has joined #openils-evergreen 2009-09-28T09:20:57 *** jenny has joined #OpenILS-Evergreen 2009-09-28T09:25:43 *** diogenes_ has quit IRC 2009-09-28T09:28:30 *** atz has joined #openils-evergreen 2009-09-28T10:26:27 *** natschil has joined #openils-evergreen 2009-09-28T10:29:25 *** jenny1 has joined #OpenILS-Evergreen 2009-09-28T10:36:26 *** jenny2 has joined #OpenILS-Evergreen 2009-09-28T10:36:42 *** dbs has joined #openils-evergreen 2009-09-28T10:37:04 updated the 1.6 install instructions a bit, like not telling people to install xulrunner 1.8 :) 2009-09-28T10:37:12 * jeff grins 2009-09-28T10:46:23 that would be a trick 2009-09-28T10:46:23 *** jenny has quit IRC 2009-09-28T10:46:57 *** cincy has joined #openils-evergreen 2009-09-28T10:49:40 would like to figure out exactly what broke zotero integration with evergreen in the 1.4 -> 1.6 phase 2009-09-28T10:50:01 i think we had similar breakage in Koha recently too 2009-09-28T10:50:12 http://forums.zotero.org/discussion/8118/error-capturing-record-in-zotero-1010-via-unapi-mods-and-evergreen-16/ suggested MODS namespace issue 2009-09-28T10:50:39 but I'm pretty sure I tried the suggested change to no avail :/ 2009-09-28T10:51:14 how's the zotero codebase... static or are they still making new releases? 2009-09-28T10:51:35 1.0.10 came out a while ago, and 2.0 is still in beta 2009-09-28T10:54:26 *** jenny1 has quit IRC 2009-09-28T10:55:54 *** dbs has quit IRC 2009-09-28T10:56:33 *** dbs has joined #openils-evergreen 2009-09-28T11:21:31 I noticed in the 1.6.0.0 install instructions that "cp dojo-1.3.2/* /openils/var/js/dojo/." isn't included as a step. Is it still necessary? (I've been blindly copying it over out of habit when I install). 2009-09-28T11:21:56 wlayton: I think it is, but I don't know for sure. 2009-09-28T11:22:13 pretty sure it is 2009-09-28T11:24:45 * phasefx_ thinks it is too 2009-09-28T11:25:09 or at least, I was told to do it for trunk 2009-09-28T11:26:06 wlayton: it isn't 2009-09-28T11:26:20 because in 1.6 tarball, you'll get the dojo layer (presumably) 2009-09-28T11:26:30 ah, doh 2009-09-28T11:26:45 oops - let me just revert that change I just made to the install instructions... 2009-09-28T11:26:54 but for anyone building outside of that, yep - I went back and forth on including the Dojo instructions earlier this monring 2009-09-28T11:26:58 hah! 2009-09-28T11:28:04 also, if you are doing any kind of local customizations that rely on dojo, your dojo bits may not be in the stock eg-provided dojo layer. 2009-09-28T11:28:23 yay, semi-fix0red zotero; looks like a fix to O:W:SuperCat:Feed is needed to set the namespace properly; and perhaps we need to rename "mods" to "mods2" and "mods32" to "mods" in O:A:SuperCat 2009-09-28T11:29:16 jeff: right, that was one of the reasons I went back and forth. trying to keep things as simple as possible on one hand, while trying to help the more advanced on the other. 2009-09-28T11:29:43 would be nice if dokuwiki supported twisties or something to show/hide extended discussions of topics. or something. 2009-09-28T11:31:34 yeah. We could hack something up with html tags :) 2009-09-28T11:32:11
normal wiki markup
2009-09-28T11:51:08 *** brendan_ga has quit IRC 2009-09-28T11:51:59 *** brendan_ga has joined #OpenILS-Evergreen 2009-09-28T12:10:07 ooh... perl5 moved to git 2009-09-28T12:10:10 http://perl5.git.perl.org/perl.git 2009-09-28T12:27:47 *** brendan_ga has joined #OpenILS-Evergreen 2009-09-28T12:28:42 *** r123__ is now known as r123 2009-09-28T12:43:25 in what schema.table do the barcodes live? 2009-09-28T12:44:23 actor.card 2009-09-28T12:44:33 if you meant patrons 2009-09-28T12:44:53 otherwise select barcode from asset.copy; 2009-09-28T12:47:35 *** anoopa has quit IRC 2009-09-28T12:48:01 *** moodaepo has joined #openils-evergreen 2009-09-28T12:48:52 jeff: that's the one i was thinking of. thx. 2009-09-28T12:49:18 ats: http://open-ils.org/documentation/evergreen_1.1.3_erd.html this always helps me 2009-09-28T12:51:10 * dbs throws hate at google apps for domains for refusing to send him his own emails, so he can't carry on a thread with himself 2009-09-28T12:54:38 i feel like every tiime i try to update, the breakage is more and more serious. 2009-09-28T12:55:06 currently can't view item holdings in OPAC or staff... can't circulate 2009-09-28T12:55:19 method=open-ils.circ.checkout.full 2009-09-28T12:55:19 params=["557df1666a65e7f115d9d3da8dfc3a34",{"barcode":"777491473045","patron":2},0] 2009-09-28T12:55:19 THROWN: 2009-09-28T12:55:20 {"payload":[],"debug":"osrfMethodException : *** Call to [open-ils.circ.checkout.full] failed for session [1254156819.915439.125415681926390], thread trace [1]:\nCan't use an undefined value as an ARRAY reference at /openils/lib/perl5/OpenILS/Application/Circ/Circulate.pm line 528.\n\n","status":500} 2009-09-28T12:55:20 STATUS: 2009-09-28T12:56:13 line 528-529 is: 2009-09-28T12:56:13 $copy = $e->search_asset_copy( 2009-09-28T12:56:14 [{barcode => $self->copy_barcode, deleted => 'f'}, $flesh ])->[0]; 2009-09-28T13:00:39 what os / what evergreen version? 2009-09-28T13:01:00 i had something like this when using an unsupported os (ubuntu 9.10 i think it was?) 2009-09-28T13:02:56 debian lenny, w/ trunk 2009-09-28T13:03:24 atz: so perl's saying that $self at that point is undefined, and therefore can't be used as an array reference 2009-09-28T13:04:41 (just as much for my benefit as yours ;) 2009-09-28T13:06:03 no, i think it is the return from search_asset_copy 2009-09-28T13:06:08 *** phase_bb has joined #openils-evergreen 2009-09-28T13:09:50 since $self is invoked in similar ways earlier in the sub, and that type of reference would be hashref (method) instead of arrayref 2009-09-28T13:09:52 in theory, it could also be $e that's undefined. of the three, yours seems most likely. 2009-09-28T13:10:04 ah! right again! 2009-09-28T13:10:12 (see, more for my benefit...) 2009-09-28T13:11:56 $e and $self are hashref, the result of the search_asset_copy call is an arrayref. 2009-09-28T13:12:01 (and now i'm just talking aloud) 2009-09-28T13:12:50 yeah, should be anyway.... though it makes sense to me for a search that gets no hits to return undef 2009-09-28T13:13:45 can't find where that method is defined though 2009-09-28T13:18:17 atz: methods of that form in the CStoreEditor class are generated rather than explicitly defined. see the runmethod sub in CStoreEditor.pm 2009-09-28T13:18:31 not sure how much that helps track down the bigger problem though 2009-09-28T13:23:10 and from the comments there: 2009-09-28T13:23:10 # not-found events are generated on retrieve and serach methods. 2009-09-28T13:23:10 # action=search methods will return [] (==true) if no data is found. If the 2009-09-28T13:23:14 # caller is interested in the not found event, they can do: 2009-09-28T13:23:16 # return $e->event unless @$results; 2009-09-28T13:27:26 *** atz has quit IRC 2009-09-28T13:27:32 *** [1]atz has joined #openils-evergreen 2009-09-28T13:27:38 *** [1]atz is now known as atz 2009-09-28T13:32:34 umm... we do $method .= '.atomic'; # in all 3 parts of the conditional there 2009-09-28T13:36:49 [2009-09-28 17:07:48] open-ils.circ [ERR :18554:EX.pm:66:1254148165260994] Exception: OpenSRF::DomainObject::oilsMethodException 2009-09-28T17:07:48 OpenILS:: 2009-09-28T13:36:49 Utils::CStoreEditor /openils/lib/perl5/OpenILS/Utils/CStoreEditor.pm:384 <500> Severe query error -- see error log for more details 2009-09-28T13:37:13 of course... that error is IN the error log. 2009-09-28T13:37:30 atz: i was told once before that those messages refer to postgres' own error log 2009-09-28T13:37:41 if it's not logging at the query level, it can be configured to 2009-09-28T13:37:53 interesting... 2009-09-28T13:38:05 [2009-09-28 17:28:16] /openils/bin/opensrf-perl.pl [INFO:21875:Inbound.pm:96:] loading router info ARRAY(0x1024278) 2009-09-28T13:38:22 ^^ also some stuff suggesting bogus xml 2009-09-28T13:38:56 *** jamesrf has joined #openils-evergreen 2009-09-28T13:40:02 also, on the severe query error, i believe the opensrf logs (osrfsys.log or wherever you've re-pointed them) at will include the query (but maybe only if you're at DEBUG) 2009-09-28T13:41:02 jeff: yeah, it's in there 2009-09-28T13:41:41 [2009-09-28 17:07:48] open-ils.circ [ERR :18554:CStoreEditor.pm:104:1254148165260994] editor[1|1] request error open-ils.cstore.direct.asset.copy.search.atomi 2009-09-28T13:41:41 c : [{"barcode":"777491473045","deleted":"f"},{"flesh_fields":{"acn":["record"],"acp":["call_number"]},"flesh":2}] : Exception: OpenSRF::DomainObject::oilsMet 2009-09-28T13:41:41 hodException 2009-09-28T17:07:48 OpenILS::Utils::CStoreEditor /openils/lib/perl5/OpenILS/Utils/CStoreEditor.pm:371 <500> Severe query error -- see error log 2009-09-28T13:41:41 for more details 2009-09-28T13:43:28 open-ils.cstore: Error retrieving asset::copy with query [SELECT "acp".age_protect, "acp".alert_message, "acp".barcode, "acp".call_number, "acp".circ_as_type, "acp".circ_lib, "acp".circ_modifier, "acp".circulate, "acp".copy_number, "acp".create_date, "acp".creator, "acp".deleted, "acp".dummy_isbn, "acp".deposit, "acp".deposit_amount, "acp".dummy_author, "acp".dummy_title, "acp".edit_date, "acp".ed 2009-09-28T13:43:28 itor, "acp".fine_level, "acp".holdable, "acp".id, "acp".loan_duration, "acp".location, "acp".opac_visible, "acp".price, "acp".ref, "acp".status FROM asset.copy AS "acp" WHERE "acp".barcode = '777491473045' AND "acp".deleted = 'f';] 2009-09-28T13:45:26 so it comes down to: 2009-09-28T13:45:27 ERROR: column acp.dummy_isbn does not exist 2009-09-28T13:45:30 the dummy_ fields... 2009-09-28T13:45:35 yeah. i was just saying. :) 2009-09-28T13:45:47 those are new. they probably exist in your fm_IDL.xml and not in your db schema. 2009-09-28T13:47:24 we really need an admin interface to "reality check" the IDL against the DB 2009-09-28T13:47:49 great automated testcase. :) 2009-09-28T13:52:40 thinking of similar tools in other projects, Drupal has a module that adds that and other related features: ``Schema comparison: the module compares the live database structure with the schema structure declared by all enabled modules, reporting on any missing or incorrect tables.'' 2009-09-28T14:00:35 *** natschil has quit IRC 2009-09-28T14:08:35 dear network: you suck today. 2009-09-28T14:09:07 (if it weren't for screen, i'd be wrecked) 2009-09-28T14:18:12 apparently the connection is getting hammered by some IP in amsterdam... welcome to the blocklist, bro 2009-09-28T14:19:58 hmm, wonder if it's repke and friends 2009-09-28T14:24:47 *** r123 has quit IRC 2009-09-28T14:34:24 ahh, worked out the simple fix for zotero 2009-09-28T14:42:21 *** r123 has joined #openils-evergreen 2009-09-28T14:46:02 *** phase_bb has quit IRC 2009-09-28T14:49:16 *** jenny2 has quit IRC 2009-09-28T14:57:43 *** atz has quit IRC 2009-09-28T14:58:26 *** atz has joined #openils-evergreen 2009-09-28T15:06:38 *** jenny1 has joined #OpenILS-Evergreen 2009-09-28T15:36:09 atz: thanks for the CC suggestion; I'll try that 2009-09-28T15:40:03 dbs: with cc/bcc or "reply from sent items" approaches, keep in mind that your subject line will lack the subject tag, so your Re: original subject will "fall out of" google's "conversation" pseudo-threading unless you stick the mailing list subject tag in there. 2009-09-28T15:40:21 annoyance * annoyance == frustration! 2009-09-28T15:40:31 *** jenny has joined #OpenILS-Evergreen 2009-09-28T15:41:56 miker_: any harm in me putting money.billing and money.payment under the sway of permacrud so we can edit existing notes on them? Or would you rather wrapper functions that ensure just the notes can be edited? 2009-09-28T15:41:59 jeff: I use evolution 90% of the time anyway 2009-09-28T15:42:26 proper_threading++ 2009-09-28T15:43:09 we should all move to nntp 2009-09-28T15:43:25 * phasefx_ would love a web forum with an email/nntp gateway 2009-09-28T15:43:34 or vice-versa 2009-09-28T15:43:36 i've seen 'em. 2009-09-28T15:43:54 * phasefx_ tried to get one going way back when; couldn't handle it 2009-09-28T15:44:23 the typical failures are due to the different expectations of users of their preferred interface. 2009-09-28T15:45:03 (once you get past the annoying technical hurdles) 2009-09-28T15:45:19 * phasefx_ thinks wrapper function is better, though pcrud is easier. Tries the former 2009-09-28T15:45:41 i've also seen both mailing lists and web forums go completely unused 2009-09-28T15:45:52 *** jenny1 has quit IRC 2009-09-28T15:54:34 phasefx: fudforum did a reasonable job of providing integrated mailing list/web forum 2009-09-28T16:01:11 I'll have to play with that, see what my chaos powers do 2009-09-28T16:05:40 in this table insert value, is the 193 in oils_i18n_gettext harmless given that 193 is already used and the number should have been 203 if it were following the pattern? (203, 'group_application.user.staff.acq_admin', oils_i18n_gettext(193, 'Allows a user to add/remove/edit users in the "Acquisitions Administrators" group', 'ppl', 'description')) 2009-09-28T16:06:37 miker_ said oils_i18n_gettext is a dummy function, but I don't know how it gets used 2009-09-28T16:07:38 phasefx: it gets parsed out by one of the i18n scripts (db-seed-i18n.py) to build a set of PO that can be translated and then generate corresponding translated inserts, so the 193 should be a 203 2009-09-28T16:08:19 hrmm, so if I correct that, should I also make an upgrade script to correct any instance of it they may be lurking around? 2009-09-28T16:08:40 what are the last two values? 2009-09-28T16:09:19 phasefx: slide 23 of http://coffeecode.net/uploads/talks/2009/Globalization1.odp 2009-09-28T16:09:25 dbs++ 2009-09-28T16:09:40 oils_i18n_gettext(keyval INT|TEXT, string TEXT, class_hint TEXT, property TEXT ) 2009-09-28T16:10:10 dbs-- # for not clearly documenting this in the wiki 2009-09-28T16:10:41 or the source 2009-09-28T16:10:48 ah 2009-09-28T16:10:49 it is in the source 2009-09-28T16:10:54 -- Parameters are: primary key, string, class hint, property 2009-09-28T16:11:08 but \df+ doesn't show those 2009-09-28T16:11:09 so many lies in http://open-ils.org/dokuwiki/doku.php?id=backend-devel:i18n 2009-09-28T16:18:31 *** jenny has quit IRC 2009-09-28T16:21:05 hrmm.. I see a possible problem with adding perms via upgrade scripts. I'm adding a perm to the seed data, but I see that permission.perm_list_id_seq is set to max(id).. so we're not reserving a range of perm id's for development. Folks can and will create perms for use as application/edit perms for permission groups. I guess my solution would be to not worry about pinning the perm to a specific id? that breaks the pattern being used for 2009-09-28T16:21:48 truncated at "breaks the pattern being used for" 2009-09-28T16:22:29 for all prior perms 2009-09-28T16:22:43 *** jenny has joined #OpenILS-Evergreen 2009-09-28T16:22:44 yeah, perms are tricky. we should probably reserve the first 1000 to be safe 2009-09-28T16:22:47 it also makes me unsure about what to use for the keyvalue argument of oils_i18n_gettext 2009-09-28T16:23:30 it has to be the actual, literal primary key for the row 2009-09-28T16:23:30 I made something up the last time I used oils_i18n_gettext :D 2009-09-28T16:24:00 so we can't leave it to fate, at least not easily 2009-09-28T16:24:04 nope 2009-09-28T16:24:12 insert and then edit with knowledge of lastval 2009-09-28T16:24:54 would that be transparent to translators, or would it throw them off the id could potentially change? 2009-09-28T16:25:06 it wouldn't work 2009-09-28T16:25:08 throw off the scripts.. oy, it would 2009-09-28T16:25:19 has to be pinned, okay 2009-09-28T16:25:50 man do I ever get sidetracked from what I want to do :) 2009-09-28T16:26:54 so how does this sound.. as one changeset, I change the SETVAL for the perm_list sequence so that we have some reserved id's, and as part of a corresponding upgrade script, I renumber any existing perms higher than the last seeded value by an offset 2009-09-28T16:27:21 and then change the sequence again 2009-09-28T16:27:52 so seed data, changes the sequence to 1000 or so.. and upgrade script, renumbers perms and changes the sequence to max(id) + 1 2009-09-28T16:31:25 it sounds reasonable 2009-09-28T16:33:39 as for my previous misuse of oils_i18n_gettext, I'll fix in the scripts, but not try to fix any instances of it (it's all trunk) 2009-09-28T16:34:58 awww.. crap 2009-09-28T16:35:54 ignore me 2009-09-28T16:36:33 *** jamesrf has quit IRC 2009-09-28T16:36:51 * dbs ignores phasefx :) 2009-09-28T16:37:55 atz++ # CCing worked. nice, simple solution 2009-09-28T16:38:14 *** wlayton has quit IRC 2009-09-28T16:44:00 dbs: cool 2009-09-28T16:49:10 * phasefx_ fears this perm_list renumbering idea 2009-09-28T16:50:27 people have added missing perms manually before, too 2009-09-28T16:51:40 how much clean-up effort should we put into it? delete all perms with known codes, renumber anything that remains, and then re-seed the perms? 2009-09-28T16:58:18 * phasefx_ will try that tact 2009-09-28T16:58:19 phasefx: that would screw up existing translations 2009-09-28T16:58:33 hrmm 2009-09-28T16:58:41 *** jamesrf has joined #openils-evergreen 2009-09-28T16:58:41 although gettext might do a reasonable fuzzy match 2009-09-28T16:58:43 gotta go 2009-09-28T16:59:30 do upgrade scripts work at all with translations? 2009-09-28T17:03:00 ah man, we have perms in seed with no pinned id and no descriptions, i18n or otherwise 2009-09-28T17:04:32 also, you can't just "delete all perms with known codes", because of those being referenced by id in the various permission.*_perm_map tables 2009-09-28T17:05:06 jeff: I was hoping that I could get around that by virtue of being in a transaction, but I think it depends on how exactly the foreign keys are created 2009-09-28T17:05:17 whether they are deferred or not 2009-09-28T17:06:01 * phasefx_ could just tackle it the other way.. renumber everything that is _not_ in the known code list 2009-09-28T17:06:03 they're both FOREIGN KEY (perm) REFERENCES permission.perm_list(id) ON DELETE CASCADE DEFERRABLE INITIALLY DEFERRED 2009-09-28T17:06:31 so, probably not an issue (at least, not THAT issue) 2009-09-28T17:06:36 you know, I wasn't even thinking about renumbering the references too.. oy 2009-09-28T17:06:44 I wonder if it automatic 2009-09-28T17:06:53 * phasefx_ doubts it is 2009-09-28T17:07:22 if I renumber say perm 300 to 1300, I'll likely have to go find every reference to 300 and change it too 2009-09-28T17:07:28 probably not automatic on "delete all these and then load all these" ;) 2009-09-28T17:07:34 haha 2009-09-28T17:07:51 that's why I'm voicing all this outloud and hoping for miker_ to poke his head in 2009-09-28T17:08:03 too big and new for me to just do it 2009-09-28T17:08:18 but it's in my way 2009-09-28T17:08:34 * phasefx_ just wanted to add an UPDATE_BILL_NOTE perm :D 2009-09-28T17:31:28 *** eguest039 has joined #OpenILS-Evergreen 2009-09-28T17:36:12 whats up guys 2009-09-28T17:38:35 *** jenny has quit IRC 2009-09-28T17:41:25 I wonder how multiple libraries maintain their circ rules on a tiered eg setup...lets say they want to change something do they tell the admin/devs to make the change? 2009-09-28T17:42:22 I should check how the pines folks have the local admin setup 2009-09-28T17:43:27 *** cincy has quit IRC 2009-09-28T17:44:05 yeah.one change one library makes could have an effect on the whole system :) 2009-09-28T17:44:21 kind of like a chain reaction 2009-09-28T17:54:35 eguest039 yea but I want to make sure libraries/sub-libraries don't have to contact a techie necessarily to make changes... 2009-09-28T17:58:04 so you kind of want to isolate the different libraries in terms of administration although they are all connected? 2009-09-28T17:58:09 policy is administered by a Unix admin if using script based circ, though in-db circ behavior (in EG 1.6) can be farmed out to specific users via permissions and a GUI 2009-09-28T17:59:36 cool 2009-09-28T17:59:59 phasefx_ that's what I have been suspecting..I did setup our demo sys librarian with torisesvn so she could make changes to the script based rules 2009-09-28T18:00:48 also, depending on how do your policies, normal staff could have lots of knobs and dials for adjusting circ behavior (loan duration and fine level fields on items, circ modifiers, etc.) 2009-09-28T18:01:20 then I'd sync the rules to the system say nightly or whatever....I'll have to look at in-db circ behaviour once I move to 1.6. we want to keep our current demo at 1.4 2009-09-28T18:01:40 and docs yet for in-db rules? 2009-09-28T18:02:05 wait I'll do a search...how lazy am I? : ) 2009-09-28T18:02:21 I don't think so, and I think the GUI's are still being polished too, and may not be in 1.6.0.0 2009-09-28T18:02:41 phasefx_ yea but the knobs would have to be thought out and put in place by the unix admin 2009-09-28T18:02:54 thought by somebody else hopefully : ) 2009-09-28T18:03:08 I believe there was/is some in-db stuff in 1.4, but I don't think there's any GUI at all there, and I"m not sure how usable it is 2009-09-28T18:03:09 is there any though of making the circ behaviours have a gui interface? 2009-09-28T18:03:23 with the scripts and all. 2009-09-28T18:04:33 eguest039: the thought is that the scripts will be replaced for the most part by in-db rules that can be manipulated with GUI. The scripts will linger on as an escape hatch for any exceptional behavior. By definition, a circ "matrix" is limited, but the scripts are turing complete and very powerful 2009-09-28T18:05:09 you could even implement a matrix within the scripts if you want to :) 2009-09-28T18:05:18 phasefx_ I wonder if eguest039 is using scripts to mean bootstrap kinda scripts 2009-09-28T18:06:00 eguest039: ah, if you're talking about config.cgi, that set of interfaces has been deprecated. Admin -> Server Settings in the staff client is one way to access the new ones 2009-09-28T18:06:50 i se 2009-09-28T18:07:15 i am using evergreen 1.2 so i dont know much about hte new features in the more advanced versions 2009-09-28T18:08:23 for 1.2, config.cgi is the way to go 2009-09-28T18:08:33 eguest039 that makes sense...you are still a few years behind 2009-09-28T18:09:20 :) 2009-09-28T18:10:10 how far behind exactly? you make me sound like some old man 2009-09-28T18:10:16 :( 2009-09-28T18:11:05 Well dbs has the Canadian using 1.6 already I think and Indiana is planning to go there too 2009-09-28T18:13:01 quick question.if i am to change the status of a user from one to another how would it affect the circualtion of books hte usr has out 2009-09-28T18:13:23 *** moodaepo is now known as anoopa\ 2009-09-28T18:13:31 *** anoopa\ is now known as anoopa 2009-09-28T18:13:32 or it would just pick up the new circualion behavior atthaced to the new user whenever there is an action to be performed? 2009-09-28T18:13:52 circ rules are locked in at the time of check out. If the item gets renewed, new behavior can slip in 2009-09-28T18:16:51 oaky 2009-09-28T18:18:07 how about if i want to delete a partiular library from the system.if supposing there are no users attached to that library .would i be able to do this? 2009-09-28T18:18:46 you have to dig into the database to do it 2009-09-28T18:19:05 at least in 1.2 2009-09-28T18:19:38 I believe the newer interfaces have a deletion option for org units, but I think it's still effectively blocked if anything is associated with the library 2009-09-28T18:20:26 ok 2009-09-28T18:23:31 * phasefx_ waves, "has to jet 2009-09-28T18:30:58 i recently tried importing some student data into the system 2009-09-28T18:31:06 the name were in unicode 2009-09-28T18:31:42 the import worked but when yu enter the staff clientand try to edit any info you get an error say ing the names are of the wrong type 2009-09-28T18:32:45 any ideas/ 2009-09-28T18:32:48 ? 2009-09-28T18:37:20 eguest039: there are regexes in the staff client that need to be relaxed 2009-09-28T18:37:41 yeah 2009-09-28T18:38:02 i get it 2009-09-28T18:38:03 eguest039: http://svn.open-ils.org/trac/ILS-Contrib/browser/conifer/trunk/xul/server/patron might be useful to you 2009-09-28T18:38:24 see http://svn.open-ils.org/trac/ILS-Contrib/browser/conifer/trunk/xul/server/patron/ue_config.js for example 2009-09-28T18:39:04 sam prblem goes for my phone number field as well i think the zip code field 2009-09-28T18:39:46 yeah - see the laxRegex in the file that I pointed to for what we use 2009-09-28T18:40:10 okay.thanks 2009-09-28T18:40:23 arguably, it should be more relaxed out of the box and sites would implement local restrictions to meet local policy... 2009-09-28T18:40:36 yeah .exactly 2009-09-28T18:42:03 was wondering how i could also chane the sytem wid currency which is used .we dont use dollars down here but it seems the whole system is calibrated around that 2009-09-28T18:42:30 with the calcuations as well as the display of currencies around money elements 2009-09-28T18:42:43 or this has more to do with localization? 2009-09-28T18:50:55 Yeah, I don't think the currency displays have been localized yet 2009-09-28T18:51:15 I had been hoping to convert those all to dojo-based i18n displays way back, but ... time 2009-09-28T18:52:15 and... xul appears to have currency boxes, so that would probably be better (https://wiki.mozilla.org/XUL:Specs:NumberBox#Currency_Box) 2009-09-28T18:54:00 * dbs looks at https://wiki.mozilla.org/Firefox/Feature_Brainstorming:Internationalization - maybe those are just specs, hmm 2009-09-28T19:08:54 they seem to have a lot of ideas on paper 2009-09-28T19:20:41 has 1.6 been tested on ubuntu jaunty? 2009-09-28T19:24:27 no idea 2009-09-28T19:28:18 twirlip: I think some people have given it a spin there, but I haven't personally 2009-09-28T19:28:48 ubuntu hardy or debian lenny are the quasi-recommended distros 2009-09-28T19:29:56 ah, ok 2009-09-28T19:32:49 * dbs starts installing karmic-alpha6 in a virtual machine 2009-09-28T19:33:56 dbs: now i feel guilty for giving you more work to do. :) 2009-09-28T19:34:41 twirlip: my invoice is in the mail ! 2009-09-28T19:35:14 * twirlip is glad he's using a pseudonym on this channel. 2009-09-28T20:11:17 *** eguest039 has left #OpenILS-Evergreen 2009-09-28T20:12:02 bah, karmic-alpha6 hasn't packaged subversion yet. sheesh 2009-09-28T20:14:37 or... huh, needed to manually "aptitude update" first. guess I expected them to do that automatically 2009-09-28T20:19:44 karmic-wha? oh, ubuntu. 2009-09-28T20:22:40 phasefx_: re earlier, the perm pkey stuff, we can make the fkey cascade and then it will, indeed, magically renumber referencing columns when you update the perm id 2009-09-28T20:23:22 we'd need to find the refs, drop the constraint, then readd the constraint with ON UPDATE CASCADE 2009-09-28T20:23:54 I'm packing for my trip, but it's not tough, I think. feel free to attack with an upgrade sql file 2009-09-28T20:24:10 and in the baseline schema, of course 2009-09-28T20:37:50 *** r123__ has joined #openils-evergreen 2009-09-28T21:10:46 *** brendan_ga has quit IRC 2009-09-28T21:25:39 *** r123___ has joined #openils-evergreen 2009-09-28T21:31:08 *** brendan_ga has joined #OpenILS-Evergreen 2009-09-28T21:31:42 *** jamesrf has quit IRC 2009-09-28T21:44:05 *** r123__ has quit IRC 2009-09-28T22:06:35 opensrf trunk works on karmic alpha 6. yay 2009-09-28T22:59:01 NOTE TO DBS: need to enable memcached in /etc/default/memcached in karmic 2009-09-28T23:00:10 "Child router process 9267 has no transport configuration" is a warning from starting opensrf router, but seems harmless for now 2009-09-28T23:00:48 w00t, srfsh login in karmic works 2009-09-28T23:01:09 postgresql 8.4 coming at you CRAZY 2009-09-28T23:04:59 dbs: I think I'm going to commit that perm_list patch. The execution is different than the cascading keys miker_ suggested, but I assume he was okay with the original premise. You and mck9 are the only other real database-muckers, so feel free to scream at me :) 2009-09-28T23:06:11 worse thing that happens is that we fix it up before cutting a new branch off of trunk or backporting 2009-09-28T23:06:14 phasefx: worst comes to worst, we can revert 2009-09-28T23:06:21 right, zactly 2009-09-28T23:06:26 dbs++ # reassurance 2009-09-28T23:11:33 * phasefx_ is going to do some other perm mangling too (give everything id's and a oils_i18n_gettext description) 2009-09-28T23:18:03 phasefx_++ 2009-09-28T23:18:37 *** r123___ has quit IRC 2009-09-28T23:18:50 dbs: I don't think "Child router process 9267 has no transport configuration" is harmless 2009-09-28T23:19:29 now mind you, these perms that are missing descriptions now, I'm just going to cheat and and do something like oils_i18n_gettext(204,'FIXME: Need description for ASSIGN_GROUP_PERM', 'ppl', 'description') ;) 2009-09-28T23:19:30 dbs: if it's what I think it is, it's a message I put in earlier today. Are you working with fresh code from trunki? 2009-09-28T23:19:46 double-w00t, xulrunner + opac running happily on karmic (record imported successfully) 2009-09-28T23:19:58 mck9: yep, fresh checkouts as of about two hours ago 2009-09-28T23:20:34 mck9: maybe reversed logic? all seems to be working well 2009-09-28T23:21:25 dbs Is the process still running? 2009-09-28T23:21:41 i.e. 9267 2009-09-28T23:22:56 mck9: I'll restart clean 2009-09-28T23:23:56 dbs: What the message means is that the newly spawned child process can't find some critical config info, so it throws up its hands and dies. 2009-09-28T23:24:21 The message goes to stderr because it doesn't know where the log file should be. 2009-09-28T23:24:22 mck9: ah, you're right. the first two Router processes immediately die 2009-09-28T23:24:51 but the next two Router processes happily keep running and serving as one would expect 2009-09-28T23:25:50 dbs: Hmm that's odd; they should all be looking at the same config, right? 2009-09-28T23:26:06 dbs pasted "mck9 - router without transport config" at http://paste.lisp.org/display/87877 2009-09-28T23:26:20 mck9: indeed 2009-09-28T23:27:20 dbs: actually each child process gets a different subset of the loaded config file, so some can work while the others don't 2009-09-28T23:28:27 The ones that are dying are looking for a section under "transport" and not finding it 2009-09-28T23:28:51 If you can read C you can look at osrf_router_main.c 2009-09-28T23:28:58 mck9: sorry, thought you meant "same config file" 2009-09-28T23:29:34 yep, I can read (and very poorly write) C - just popped that file open 2009-09-28T23:29:45 Are you running one router process that spawns four children, or two that spawn two each? 2009-09-28T23:30:13 Until today that message wasn't there. The process would just die silently. 2009-09-28T23:30:34 I suspect they've always died silently 2009-09-28T23:32:40 ah, because the context passed by osrf_ctl.sh is "routers" 2009-09-28T23:33:37 and there's 2 /config/opensrf/routers/router entries which have no transport subsection 2009-09-28T23:34:37 Yeah, that sounds about right -- I'm a bit hazy about what the config file should look like 2009-09-28T23:34:38 versus the 2 /config/routers/router entries which do; ergo, I suspect berick / miker_ expected the "return 0;" to just return silently 2009-09-28T23:35:16 kind of messy to spawn a process that you know is just going to die... 2009-09-28T23:35:25 Maybe. But that seems perverse. 2009-09-28T23:36:29 Well if that's what they intended, then it's working, I guess, but the message I put in is, um, misleading? 2009-09-28T23:37:39 well... can we move the test for "transport" prior to the fork()? 2009-09-28T23:38:14 uh oh baby 2009-09-28T23:41:22 dbs: It probably makes more sense to parse the config file better 2009-09-28T23:41:53 There's a routers section under opensrf, and another routers section under the root. 2009-09-28T23:42:22 We're grabbing both of them and then throwing away the one we don't want. I think. 2009-09-28T23:43:36 Also I'm not sure why I didn't see that message in my testing (admittedly somewhat cursory). 2009-09-28T23:44:54 Oh, wait, I did get that message. I just didn't notice. 2009-09-28T23:45:51 I'll comment out that message for now, but there's got to be a better way. 2009-09-28T23:47:36 mck9: baby is back asleep (yay) - yep, better parsing would make sense 2009-09-28T23:48:30 phasefx_: you didn't add the config.upgrade_log insert to your upgrade sql file 2009-09-28T23:49:20 miker_: thanks 2009-09-28T23:50:42 always bums me when I can't get everything right in a single changeset :) 2009-09-28T23:53:37 phasefx: join the club :) 2009-09-28T23:54:45 my worry is that I'll go to backport a changeset later and forget something (not sure what the ideal workflow is for that). I'm sure some of us can share how git makes it easy :) 2009-09-28T23:58:52 miker_ gets to meet Cory Doctorow, the bastage :)