2009-12-26T05:03:29 *** natschil has joined #evergreen 2009-12-26T05:43:11 *** natschil has quit IRC 2009-12-26T06:34:56 *** natschil has joined #evergreen 2009-12-26T07:31:23 *** natschil has quit IRC 2009-12-26T09:30:45 *** mck9 has joined #evergreen 2009-12-26T11:37:19 *** rsinger has quit IRC 2009-12-26T13:20:56 *** brendan_ga has joined #evergreen 2009-12-26T15:07:51 *** natschil has joined #evergreen 2009-12-26T15:37:57 *** natschil has quit IRC 2009-12-26T15:42:31 *** phase_bb has quit IRC 2009-12-26T16:34:15 *** dbs has joined #evergreen 2009-12-26T16:35:13 Completely untested guess about our problems with non-ASCII characters in SIP: the code in http://svn.open-ils.org/trac/ILS/changeset/11975/trunk/Open-ILS/src/perlmods/OpenILS/SIP/Item.pm assumes NFD data 2009-12-26T16:43:27 a quick test suggests that that could very well be the case (given that our data is NFC in the database) 2009-12-26T18:35:25 (also assuming that a brand-new 3M self-check unit is a "bad SIP client" in terms of handling Unicode, and that the problem mentioned in the fix would actually fix the problem I'm seeing. lots of assumptions) 2009-12-26T18:37:11 *** moodaepo has quit IRC 2009-12-26T18:38:24 finally, if the problem is that checksums get screwed up with unicode chars, it's not just titles that would need the special treatment aka mangling of NFD + s/\pM+//; author names, patron names, addresses... everything would probably need it 2009-12-26T19:19:17 *** dbs has quit IRC 2009-12-26T19:57:10 *** frzosima has joined #evergreen 2009-12-26T19:57:19 *** dbs has joined #evergreen 2009-12-26T20:00:47 *** pmplett has quit IRC 2009-12-26T20:14:00 *** frzosima has quit IRC 2009-12-26T20:14:16 *** frzosima has joined #evergreen 2009-12-26T20:19:15 *** frzosima has quit IRC 2009-12-26T21:10:24 *** frzosima has joined #evergreen 2009-12-26T21:15:00 *** sboyette has quit IRC 2009-12-26T21:15:00 *** AbizzalsX has quit IRC 2009-12-26T21:15:00 *** jeff has quit IRC 2009-12-26T21:17:54 *** sboyette has joined #evergreen 2009-12-26T21:17:54 *** AbizzalsX has joined #evergreen 2009-12-26T21:17:54 *** jeff has joined #evergreen 2009-12-26T22:15:15 dbs: just poking my head in for a second, but the problem is that the SIP2 spec, IIRC, specifies a windows codepage for non-ascii chars. so, in reality, it can't handle unicode at all, per spec. thus the non-spacing mark stripping (though we'd have to strip all non-ascii to reall "fix" things). also IIRC, that's from corespondance with a 3m dev