Open Source Integrated Library System

Evergreen on IRC

#evergreen Logs for Monday, August 23rd, 2010

< Sunday, August 22nd, 2010Raw Log FileTuesday, August 24th, 2010 >
#TimeNickMessage
#01:20:23atz_ has joined #evergreen
#01:23:40atz has quit IRC
#03:25:30magnusenger_ has joined #evergreen
#03:49:35leed has quit IRC
#03:49:35_bott_ has quit IRC
#03:49:35berick has quit IRC
#03:59:55leed has joined #evergreen
#03:59:55_bott_ has joined #evergreen
#03:59:55berick has joined #evergreen
#07:32:08sfortin has joined #evergreen
#07:33:10rickd has quit IRC
#07:34:05rickd has joined #evergreen
#07:35:30artunit has quit IRC
#08:05:56granitize has joined #evergreen
#08:18:52cerpy has joined #evergreen
#08:50:53sfortin has quit IRC
#08:53:09Callender has quit IRC
#08:55:45dbs has joined #evergreen
#08:57:55Callender has joined #evergreen
#09:03:04natschil has joined #evergreen
#09:08:28natschil has quit IRC
#09:08:52berickdbs: trying the authority ui.. Submit/<Enter> aren't sending any searches to the server. tried in staff client and firefox (as admin). no JS errors. curuious if you have any ideas
#09:09:38dbsberick: weird, I'll try on a clean Debian install to ensure there's no local interference
#09:10:04natschil has joined #evergreen
#09:10:31Dyrcona has joined #evergreen
#09:11:04berickdbs: thanks. this is latest trunk w/ dojo 1.3.2. /me updates to 1.3.3 to rule that out
#09:11:25dbsokay, yeah I've been using dojo 1.3.3 so you never know
#09:14:04bshum has joined #evergreen
#09:14:09bshum has joined #evergreen
#09:15:17natschil has quit IRC
#09:16:35berick1.3.3 does not appear to affect it
#09:17:23berickoh..hm, actually, i should be checking the apache logs, not osrfsys.log..
#09:17:29berickfor the initial call
#09:18:03dbsor use firebug to see if the POST request is made
#09:19:24berickok, 404's, so probably apache config issue
#09:19:51dbs*phew* -ish
#09:20:03natschil has joined #evergreen
#09:20:37dbsoils_web.xml?
#09:21:23dbsupdated entities or dojo i18n file?
#09:22:24dbs(actually, I haven't dojo-i18nized those files yet, so lang.dtd would be a likely culprit)
#09:22:44dbs(actually, that was only for the staff client menu IIRC, so ignore me)
#09:24:37berickno, berick not following the commits closely and debugging from the wrong direction.. eg_vhost.conf
#09:25:20dbsoh, /browse /startwith
#09:25:38dbsman, that was ages ago :)
#09:26:17dbsstill, firebug should have shown the POST. hrm
#09:26:27berickwasn't using firebug
#09:26:31berickwas watching activity.log ;)
#09:27:30natschil has quit IRC
#09:29:08bericksorry for the alarm, dbs
#09:29:27dbsNo problem, given how many times I've pestered y'all its only fair!
#09:30:12dbsshould we move the templates from /openils/var/web/templates to /openils/var/templates before release
#09:31:05dbsDoesn't seem to be a huge threat, but we've had that comment in oils_web.xml for a long time now
#09:31:49yboston has joined #evergreen
#09:32:25berickit would be nice to move them
#09:32:48berickmaybe after the alpha1?
#09:33:16dbssure
#09:46:39dbsOpenSRF 1.6.0-alpha is now in the downloads directory (http://evergreen-ils.org/downloads/OpenSRF-1_6_0-alpha.tar.gz) with md5sum (http://evergreen-ils.org/downloads/OpenSRF-1_6_0-alpha.tar.gz.md5) and changelog (http://evergreen-ils.org/downloads/OpenSRF-ChangeLog-1.4.0-1.6.0-alpha)
#09:46:50jeffdbs++
#09:47:19dbsstill needs some release notes, if anyone feels ambitious, and certainly could use some sniff testing
#09:47:48jeffis there a handy reference page for pairings of opensrf/evergreen versions?
#09:48:35dbsOther than the downloads page?
#09:49:39dbsYou mean something like "Evergreen 1.6 should work with OpenSRF 1.2, 1.4, and 1.6, but almost everybody is running 1.2.2"?
#09:52:09dbsI think jamesrf was maintaining a wiki page somewhere
#09:52:10jeffand something with more history and more "we think 2.0's going to use"
#09:52:30dbsyeah, http://evergreen-ils.org/dokuwiki/doku.php?id=versioning
#09:53:36dbs bumps the 2.0 reqs
#09:54:18moodaepodbs++
#09:55:05moodaepomiker_++
#09:55:51miker_building a tarball now
#09:58:14jeffdbs: aha! that's exactly what I was looking for. thanks.
#09:58:37magnusenger_ has left #evergreen
#10:06:15brendan_bywater has joined #evergreen
#10:06:51miker_not that I'm concerned for alpha1, but I'm getting i18n errors for accesskeys: warnings.warn("Could not find accesskey for %s" % entity)
#10:07:07atz_:\
#10:08:36cerpy has quit IRC
#10:10:33moodaepo has quit IRC
#10:13:04dbsmiker_: good, the script is doing it's job then?
#10:14:29miker_still running (lots of new locales), but no death-causing errors that I can see in the output as yet
#10:14:30dbs(essentially, if an entity is defined with a name m/^(.*?)\.label$/ it expects to find $1.accesskey)
#10:14:32moodaepo has joined #evergreen
#10:15:03dbsYeah. I haven't synced up the translations from launchpad in quite some time. Have I mentioned we need a real i18n maintainer?
#10:16:44miker_biab to tar it up ... meeting real quick-like
#10:34:15youdonotexist has joined #evergreen
#10:35:15dbsI'm exporting a virtual image of OpenSRF trunk & Evergreen trunk built on Debian Squeeze atm. Just remembered that the XUL staff client directory is just a timestamp though. :/
#10:36:04phasefxtimestamps, ptooey
#10:38:39dbsah well, I'll want to create a new image based on the Official OpenSRF 1.6.0 and Evergreen 2.0 alphas anyway
#10:41:35lisppasteberick pasted "marc:::record mfhd version error (2.0.2 installed)" at http://paste.lisp.org/display/113778
#10:41:57bericki've seen this kind of perl wonkiness before...
#10:43:11dbsyuck
#10:43:12atz_dbs: actually i tweaked the update.sh script to always stick the build in a timestamp directory, but it will symlink whatever STAFF_CLIENT_BUILD_ID you might set in ENV
#10:43:32dbsatz_: I don't use the update.sh script
#10:43:41gmcharltberick: a 5.8ism?
#10:44:05atz_previously if you tried setting the build id to "current" it would still try to symlink "current" to the build_id-created dir... which would obviously not work.
#10:44:07dbsIs the param supposed to be quoted, perhaps?
#10:44:09berickgmcharlt: it is 5.8.8, so perhaps
#10:44:30atz_i.e., current -> current
#10:44:39berick tries quoting
#10:44:56berickdbs: *ding* *ding*
#10:45:23berick is also not heartened by Useless use of a constant in void context at /openils/lib/perl5/OpenILS/Application/Search/Biblio.pm line 263.
#10:45:51dbsEvergreen 2.0 requires Perl 5.10? :)
#10:45:54atz_at least that error is potentially harmless....
#10:46:33dbsthat line looks wonky
#10:46:44dbstwo =~ operators in a single line?
#10:46:45gmcharltI suspect the first =~ should be =
#10:47:01dbsright
#10:47:11atz_doesn't make sense otherwise
#10:47:26atz_and should mirror the line above it
#10:58:36jeff has quit IRC
#10:58:57sfortin has joined #evergreen
#10:59:53jeff has joined #evergreen
#11:03:34phasefxI think I've tried something like this before but can't it to work, ideas? cd ~/git/ILS/trunk ; git checkout some_branch ; git diff master > /tmp/patch ; cd ~/svn/ILS/rel_2_0 ; patch -p0 < /tmp/patch
#11:04:12gmcharltphasefx: in what way does it break for you?
#11:04:18berickphasefx: -p1 might be needed
#11:04:39atz_git diff --noprefix master
#11:04:53phasefxpatch complaining about the wrong -p or --strip option
#11:04:58phasefx will try those, thanks!
#11:05:14atz_phasefx: sorry, --no-prefix
#11:05:33berick makes note of --no-prefix
#11:05:59phasefxperfect, atz++
#11:06:42atz_np
#11:11:13Dyrcona has quit IRC
#11:12:09atz_ok, my current effort is to build the human-readable presentation of an EDI message. in part, this is what Business::EDI is built for. but that would involve a lot of traditional CGI code that the staff client otherwise doesn't have.
#11:12:46atz_berick: you mentioned adding another apache module... any other possibilities?
#11:14:35atz_that is, another v.server
#11:15:57berickright, i was thinking a new openils::www::* mod_perl handler.
#11:16:14berickcould also turn on perl exec in template toolkit and use a combo of TT/raw perl
#11:16:59atz_yeah, i was thinking about that... handler is probably preferable
#11:21:13miker_Evergreen-ILS-2.0-alpha1.tar.gz[.md5] uploaded
#11:21:28bradldun dun DUUUUN
#11:21:50berickatz_: there's precedent for the mod_perl approach. and this is not really the type of thing that people will need to style.
#11:22:23berickmiker_++
#11:22:30berickand dbs++ # opensrf
#11:22:38atz_sounds good... where should the backend script itself live?
#11:22:41gmcharltmiker_++
#11:22:46berickoh, speaking of, i can merge my net-server gutting now.. muahahhaa
#11:23:08miker_of course, there are now at least 2 changes that aren't in there... gah
#11:24:22bradlheh, wouldn't be right otherwise
#11:24:38jeffwhat, you mean you bought a new computer and it's already obsolete?
#11:24:43jeff:)
#11:24:44miker_ tags 2.0-alpha1.1
#11:24:50miker_(no, not really)
#11:26:19dbsuploading 2010-08-23 trunk image
#11:26:38gmcharlt... then 2.0-alpha1.2-stop-fixing-bugs-already ;)
#11:27:33berickatz_: the code that loads business::edi and translates an edi message into html? i was assuming that would live in the mod_perl handler.
#11:27:35miker_gmcharlt: egg sack lee
#11:28:12berickatz_: though, i could see part of it living under o::a::acq::edi:: if it would be useful in other contexts. e.g. action/trigger edi message displays for printing or some such
#11:28:40berickbut anything specific to html seems like it should live in the mod_perl handler
#11:28:42atz_ok that makes sense
#11:37:17csharpmiker_: dbs - I see that you'll be on FLOSS weekly - cool!
#11:37:20csharphttp://spreadsheets.google.com/pub?key=pYAJMbVobYCTro_z4LGo3ZQ
#11:39:39miker_csharp: it is true, sir
#11:40:38csharpI was just listening to the episode on VirtualBox and heard Randal Schwartz say your names - pretty wierd ;-)
#11:40:50csharpweird, even
#11:41:24dbsYep, evergreen - the online card catalog!
#11:41:32csharpheh - I caught that
#11:41:56csharp shuffles his cards
#11:43:32dbs does enjoy the occasional poker session...
#11:44:05csharpan Austin pub sounds like a good place for it
#11:45:03atz_floss'n it.
#11:49:54dbs curses rogers.com for throttling sftp transfers
#11:51:06dbs(using my in-law's net connection to upload the virtual image at a pathetic rate)
#11:53:31leedmiker_: eh?
#11:59:17dbs goes to forage for food to burn off some steam whilst rogers impedes the progress of software freedom for libraries...
#12:01:49miker_leed: ?
#12:02:25leedegg sack lee? lol I now know this wasn't directed towards me
#12:02:42miker_ahh
#12:04:45miker_oh, a note about the alpha. I intentionally included the release of dojo 1.3.3, not our layered version
#12:09:00jamesrf has joined #evergreen
#12:20:35jamesrf has quit IRC
#12:20:41jamesrf_ has joined #evergreen
#12:27:57bshumQuick question about setting "grace periods." That would be something set up for the entire system or skipped over? It's not granular at present time to be a per-library setting?
#12:29:18bshumOr is it something that can only be accomplished if we were using scripted circ vs. in-db circ?
#12:40:21gmcharltbshum: global setting, via a command-line parameter to the fine generator cronjob
#12:40:41gmcharltin-DB circ makes no difference
#12:40:43bshumgmcharlt: Thanks for the confirmation, was just poking in markmail and saw that, hmm
#12:40:56bshumIf it's global, then it's an all or nothing deal then, at current.
#12:41:04gmcharltcorrect
#12:42:17bshumThanks gmcharlt
#12:49:31afterl has joined #evergreen
#13:04:36alxp has joined #evergreen
#13:22:18sfortin has quit IRC
#13:28:08afterl has quit IRC
#13:28:35afterl has joined #evergreen
#13:41:21dbsberick: after getting through the various config issues that plagued you this morning, were you amazed by the auth management interface?
#13:51:43berickdbs: i like :)
#13:53:07dbs5 minutes until http://evergreen-ils.org/~denials/Evergreen_trunk_2010_08_23.zip upload finishes
#13:59:59jeffdbs++ / rogers--
#14:00:09dbsDone!
#14:00:13berick likes how grouchy screen gets.. "this IS window 1"
#14:00:19berickthe "doofus" is implied
#14:05:24dbsheh
#14:10:22dbsoff to upei - see you tomorrow, folks
#14:10:25dbs has quit IRC
#14:19:07branflakes has joined #evergreen
#14:49:06jamesrf_ is now known as jamesrf
#15:00:07moodaepoIf I want to un-expire a set of hold_request rows would setting expire_time and cancel_time to null be sufficient? Next time the hold targeter is run from crontab the un-expired hold requests should be targeted. Correct?
#15:05:18branflakes has quit IRC
#15:09:15atz_ has quit IRC
#15:16:59jamesrf has quit IRC
#15:23:04miker_moodaepo: subject to the state of prev_check_time, yes
#15:24:45moodaepomiker_: I was thinking of setting it to null also since we have holds targeter checking every 24 hours and they get checked. Do I need to look at unfullfilled_hold_list?
#15:27:16miker_moodaepo: no need to look at unfullfilled_hold_list
#15:28:21miker_moodaepo: IIRC, trunk has cancel_cause too, which I don't believe is in any 1.6, so for a trunk patch you'd want to include that too
#15:29:52moodaepoNice. No 1.6.0.6 has cancel_cause as a field for hold_request and a table hold_request_cancel_cause
#15:30:15miker_ahh, well then, there we go :)
#15:31:28miker_ leaves to prepare for travel
#15:43:50alxp has quit IRC
#16:01:05jamesrf has joined #evergreen
#16:18:44bshum has quit IRC
#16:57:47afterl has quit IRC
#17:03:15yboston has quit IRC
#17:10:04branflakes has joined #evergreen
#17:13:32jamesrfhas anyone ever looked into highlighting search results in the PAC?
#17:18:24atz has joined #evergreen
#17:20:28atz_ has joined #evergreen
#17:24:02atz has quit IRC
#17:46:25moodaepojamesrf: What do you mean by highlighting? Highlight the search terms where they are found in the list/record?
#17:46:44jamesrfyes
#17:47:53moodaepoNot that I have heard of...shouldn't be too hard to do a brute force one since the search criteria is available in the url.
#17:48:17jamesrfyeah i don't imagine that it's possible to do a perfectly accurate one
#17:49:05jamesrfseeing as the index could be in mods or marc
#17:49:15moodaeporight
#17:49:58moodaepo if one were to attempt a hightlight it would be nice to have an "on/off" option
#17:50:31jamesrfi was thinking about just adding it as a supercat metadata format
#17:50:33moodaepoSince I think it sometimes gets in the way of reading a page..sometimes
#17:51:12gmcharltjamesrf: a bit of client-side javascript to do the highlighting might be easier
#17:52:01atz_jamesrf, berick already converted some jaquery highlighting code to dojo for that purpose
#17:52:13atz_*jquery
#17:52:29atz_works pretty well
#17:54:50jamesrfis there somewhere i can look at that?
#17:55:03moodaepoHas anyone gotten reports about Excel formatted files generated by EG reports not work with Office 2010? Getting an error asking to extract the data from the file and then repair it > doing so does not work.
#17:55:44atz_working on some report issues today... nothing 2010 specific though
#17:56:02atz_can always try it w/ open office to see if that thinks it is OK
#17:56:56fgjohnson has joined #evergreen
#17:57:00moodaepoatz_: Will do
#17:57:12fgjohnson has quit IRC
#17:57:44atz_highlighter js: http://svn.open-ils.org/trac/ILS/changeset/17205
#17:58:15jamesrfthx atz++
#19:08:40jamesrf has quit IRC
#19:15:47branflakesmoodaepo: i think sitka has seen something similar with office 2007
#19:16:29branflakesoffice might also act in unpredictable ways if the file you're trying to read as a spreadsheet has more than 65,536 rows
#19:17:11moodaepobranflakes++ # Will check 'morrow
#19:41:29youdonotexist has quit IRC
#20:38:02brendan_ga has joined #evergreen
#20:41:48brendan_bywater has quit IRC
#20:42:46brendan_ga has quit IRC
#21:22:52jeffmoodaepo: yes, i believe i helped someone in here with that issue earlier last month. i'll check.
#21:26:47jeffmoodaepo: are your office 2010 users also using win7?
#21:29:25jeffmoodaepo: there is a bug in OLE::Storage_Lite and/or Spreadsheet::WriteExcel. fixed in OLE::Storage_Lite 0.19 -- mrpeters found that he also needed to upgrade Spreadsheet::WriteExcel, my recommended version there was 2.29 or higher.
#21:31:00jeffmoodaepo: if your are seeing this issue on a version of windows other than win7, this might not be the same issue.
#21:31:44jeffmoodaepo: the perl modules in question would need to be upgraded on the host running clark-kent.pl, and clark-kent.pl restarted. you'd then need to create new report output to test. the old output would remain broken.
#21:34:01jeff@later tell moodaepo check scrollback for a probable lead on your Excel issues
#21:34:01pinesoljeff: The operation succeeded.
#21:36:33brendan_bywater has joined #evergreen
< Sunday, August 22nd, 2010Raw Log FileTuesday, August 24th, 2010 >