Open Source Integrated Library System

Evergreen on IRC

#evergreen Logs for Monday, June 20th, 2011

< Sunday, June 19th, 2011Raw Log FileTuesday, June 21st, 2011 >
#TimeNickMessage
#03:37:39atz__ has joined #evergreen
#03:40:37atz_ has quit IRC
#05:48:34natschil has joined #evergreen
#07:26:15artunit_ has joined #evergreen
#07:28:03artunit has quit IRC
#07:28:08artunit_ is now known as artunit
#07:34:05sfortin has joined #evergreen
#08:09:06kmlussier has joined #evergreen
#08:28:40AaronZ-PLS has joined #evergreen
#08:33:11adbowling-isl has joined #evergreen
#08:44:31adbowling-isl has quit IRC
#08:48:45drdata_esi has joined #evergreen
#08:52:32jenny has joined #evergreen
#09:04:28Meliss has joined #evergreen
#09:05:25collum has joined #evergreen
#09:11:26kmlussier has quit IRC
#09:16:19sfortin has quit IRC
#09:16:32sfortin has joined #evergreen
#09:16:34Dyrcona has joined #evergreen
#09:40:47dbs has joined #evergreen
#09:40:47dbs has joined #evergreen
#09:48:32csharp has joined #evergreen
#09:56:28csharp_GPLS has joined #evergreen
#10:03:24dbsungh. what's the SSH port on open-ils.org now?
#10:04:12dbs moves that port 22 be re-added to the list of SSH ports that open-ils.org listens to
#10:05:49phasefx shared. security_through_obscurity+-
#10:06:19phasefxjust be glad there's no port knocking :)
#10:06:25dbsI imagine a port scan would turn up the same information, I'm not sure the inconvenience is worth it
#10:06:39dbs@praise a lack of port knocking
#10:06:39pinesol_greendbs: Error: There are no praises in my database for #.
#10:06:51phasefx wishes scp and ssh used the same params for port
#10:07:39dbs believes in the power of rsync over SSH
#10:08:31phasefx throws random crap into his public_html
#10:10:20dbsdev meeting tomorrow!
#10:32:32kmlussier has joined #evergreen
#10:41:33dbsberick / shadowspar / others interested in template toolkit opac: I'm uploading a VirtualBox VDI to ~/denials/tt_opac-2011-06-19.vdi.bz2 - based on current template-toolkit-opac branch + master
#10:42:49dbsfor phasefx / myself in future: rsync-over-ssh approach = 'rsync --partial --progress -e "ssh -p $OBSCURE_PORT_NUM" /path/to/tt_opac-2011-06-19.vdi.bz2 denials@open-ils.org:/home/denials/public_html/.
#10:43:11dbs(--partial means that if the 500MB upload is interrupted, you can resume the upload later)
#10:43:20dbsthat's the big win IMO over scp/sftp
#10:43:56phasefxsweet
#10:44:08Dyrconarsync++
#10:45:01eeevildbs: doesn't ~/.ssh/config with Host open-ils.org / Port 3399 handle everything for you?
#10:45:16dbseeevil: yeah, if I set that up on every machine that I use
#10:45:24eeevilgotcha
#10:45:48dbsoh noes, you gave away the OBSCURE_PORT_NUM!
#10:45:55phasefx waves his hand in an obscuring manner, "this not the port number you're looking for
#10:45:55eeeviland I just busted the obscurity ... DOH!
#10:46:02dbslulzsec will be at the door IN NO TIME
#10:46:36eeevilwell, I'm agreed in doubting the actual utility of that mechanism too ... so I don't feel too bad
#10:46:39Dyrconajust disable password authentication and require keys.
#10:46:59eeevilDyrcona: that I'd favor
#10:47:41phasefxI guess it might help against drive-by hits against port 22, but not a targeted effort to hack the site
#10:48:54dbspubkey-only-auth++
#10:49:24dbsNow that we've documented the obscure port number, we could change it back to 22 to confuse crackerz
#10:49:51gmcharlteeevil: I've tested and pushed the updated 2.0.6-2.0.7
#10:51:17dbsone note on the uri reingest function - i ended up adding a join to bre with a test for well-formed-xml on our system, because we have some invalid XML in bre.marc that pooched the reingest otherwise
#10:52:09dbsI'm not sure whether we want to accommodate bad bre.marc in the uri reingest upgrade function, but if we do it's two lines
#10:53:30gmcharltdbs: I suggest adding that join; since those records are malformed anyway and thereby useless until fixed, no need to hold up the rest of the upgrade
#10:53:57gmcharltI coudl see forcing the issue, but some of the larger consortia could potentially have tons of non-well-formed records
#10:54:11dbs puts hand up for his consortia :)
#10:54:21dbsokay, I'll push a quick branch out with the added bit
#10:55:06bshumTidbit for the version bump for 2.1-rc1 commit. There was a change to the windowssetup.nsi file from beta to rc1 (but it should be beta) else staff client building is wonky. This should be changed back sometime.
#10:57:49gmcharltbshum: bshum regarding parallel hold targetting and fine generation, I suggest just bumping up to two parallel processes to start and working up
#10:58:13gmcharltI don't remember - do you have a separate utility server?
#10:58:36bshumgmcharlt: We do, but it couldn't handle doing all the work of every cron, so I ended up breaking it out to different bricks temporarily.
#10:58:52bshumgmcharlt: Right now hold_targeter.pl runs on brick2
#10:59:30bshumIt seems that somehow we ended up with ridiculous number of holds being processed within the same hourly block. Like 4.5k between 4-5pm and 6k between 5-6pm
#10:59:45bshumI think that's why our hold_targeter script is choking and taking too long.
#10:59:58gmcharltbshum: which is also used for normal OPAC and staff use? in that case, yeah, just start with two parallel processes
#11:00:04gmcharltand how frequently do you run the targeter?
#11:00:18bshumgmcharlt: Right now, it's set for every 15 (which I think was the default suggested in the example)
#11:00:23bshum*15 minutes
#11:00:25tsberebshum: Did you migrate between, say, 4-6 pm?
#11:00:36tsbereOr at least migrate *holds* during that time?
#11:00:56bshumtsbere: Our migration didn't come with holds. I think we screwed up by not running hold_targeter.pl properly at first.
#11:01:21dbsgmcharlt / eeevil: see user/dbs/rel_2_0_upgrade_well-formed-XML
#11:01:52bshumSo the prev_check_time didn't update properly till we successfully ran through the hold_targeter.pl script
#11:02:41bshumOur utility server died in the morning when we failed to raise the max_children and it became overwhelmed. Presumably running the processes later in the day moved all the time for hold checking into those hours.
#11:03:07natschil has quit IRC
#11:03:27natschil has joined #evergreen
#11:03:32bshumgmcharlt: I'll start with two parallel, thanks. I actually removed that brick from active rotation in case we needed its full attention for hold processing.
#11:04:55berickdbs++ image
#11:08:06gmcharltdbs: any reason why biblio.reingest_uris shouldn't be in 030.schema.metabib.sql?
#11:08:30gmcharltif you don't think it's worth keeping as a permanent part of the schema, upgrade script should drop it after use, I suppose
#11:11:11dbsgmcharlt: I didn't expect reingest_uris() to be required regularly, so a DROP would make sense - assuming that the reingest_uris() call succeeds
#11:12:39gmcharltdbs: well, if it doesn't succeed, the transaction rolls back, so tacking a DROP before the commit should have the desired effect
#11:12:51dbsgmcharlt: right, sure
#11:13:07dbsI'll add that to the branch if you would like
#11:13:57gmcharltplease
#11:14:03natschil has quit IRC
#11:14:29natschil has joined #evergreen
#11:17:16dbspushed
#11:22:56gmcharltdbs: tested; pushing; I'll also cherry-pick the change to 0559 for master and rel_2_1
#11:23:12dbsgmcharlt++
#11:24:02dbs hopes to be able to try out bjwebb's DESTDIR branches for OpenSRF and Evergreen this afternoon on a clean Squeeze image
#11:26:11gmcharltand done
#11:26:35dbsgotta run to a meeting, but thanks to moodaepo we have minutes of the last dev meeting to review at http://evergreen-ils.org/dokuwiki/doku.php?id=dev:meetings:2011-06-07 before tomorrow's dev meeting
#11:32:29frank_ has joined #evergreen
#11:45:49asimon_ has joined #evergreen
#11:45:50eeevilgmcharlt: could I get you to move my errant rel_2_1_rc1 tag-ish branch to tags/rel_2_1_rc1?
#11:46:06eeevilshould'a pushed it right the first time... :)
#11:46:10tsbereeeevil: If gmcharlt isn't available to I can
#11:46:18eeeviltsbere: rock, thanks
#11:47:22tsbereeeevil: What about this rel_1_6_2_0_rc1 branch?
#11:47:58tsbere(outside of tags)
#11:48:24asimon_Would someone verify that the 2.0 Register Patron single screen uses user_edit.xhtml and user_edit.js, and that ue_config.js is fully depracated? Thanks.
#11:48:51tsbereasimon_: Dunno about user_edit.js.....the 2.0 register patron uses a register.js somewhere though.
#11:49:12tsbereeeevil: did fix the rel_2_1_rc1 one though
#11:51:21asimon_tsbere: Would that be /openils/var/web/js/ui/default/actor/user/register.js?
#11:51:29tsbereasimon_: Looks right.
#11:52:24eeeviltsbere: meh on the 1.6.2.0-rc1
#11:53:47asimon_tsbere: Thanks.
#12:04:54frank_hello everybody, i´d want to know if someone could tell me what are the step i have to keep to could configure the email notificacion, such as the password reset
#12:32:12asimon_ has quit IRC
#12:39:39kmlussier has quit IRC
#12:40:14kmlussier has joined #evergreen
#12:44:34joseph_ has joined #evergreen
#12:46:10joseph_Is buildbot working with git?
#12:51:22dbsjoseph_: it should be, yes
#12:51:31dbsjoseph_: what is your real question? :)
#12:52:27joseph_dbs: My real question is, how can I set it up to check my personal repo so I can see why it isn't building?
#12:52:41joseph_dbs: (:
#12:53:38dbsjoseph_: ah, so you want to have a particular branch of another repo added to the buildbot at testing.evergreen-ils.org ? Or you want to create your own local buildbot instance?
#12:55:11joseph_dbs: either would work, the repo I'm working out of is in the main evergreen git, building by hand is just getting old (I could very well be doing it wrong though).
#12:55:49asimon_ has joined #evergreen
#12:55:53dbsjoseph_: ah, so buildbot only builds the branches that it's told to build
#12:56:53dbsyou could tailor OpenSRF/examples/buildbot.cfg to point at the working repo + your branch for a local buildbot instance
#12:57:11asimon_I'd like to be able to change some specific Register Patron options, but can't figure out how to do so with 2.0. How can I set the County field to be required? How can I change the default entry for the Internet Access Level?
#12:57:25joseph_dbs: thanks, I'll check in to that :)
#12:57:55dbsright now, the general EG buildbot just builds core branches on a few different platforms; branches from the working repo aren't part of that
#12:58:33joseph_I can certainly see why.
#12:59:06dbsit would be awesome to have somebody else on the project who can make buildbot jump through hoops!
#13:00:16joseph_ imagines robots jumping through hoops...
#13:02:51joseph_asimon_: I believe I have what you want, if you open the staff client, then go to the admin menu, local administration, then library settings editor, I can give you the names of the settings to change.
#13:04:09gmcharltjoseph_++ # training bots to jump hoops ;)
#13:05:15joseph_the two I think you want are: "Default level of patrons' internet access" and "GUI: Require county field on patron registration" dbs or gmcharlt can correct me if I'm wrong though.
#13:07:48tsbereasimon_: Default inet access level is an org unit setting, I think.
#13:08:19tsbereasimon_: For default things to required you need manual edits in 2.0, and org unit settings in 2.1. Which you using?
#13:08:53gmcharltjoseph_: you have it
#13:09:43tsbere is under the impression that that GUI option doesn't exist in 2.0, just 2.1+, because he doesn't recall that patch of his being backported (due to being a new feature)
#13:10:29asimon_joseph_: Default level of patron
#13:11:00asimon_s internet access is a text field. Does it have to match one of the three default statuses?
#13:11:53tsbereasimon_: I seem to recall having to set that to an integer
#13:11:54joseph_asimon_: That I do not know, sorry, if someone answers, I can look in to filing a bug report.
#13:12:17asimon_joseph_: I don't have "GUI: Require county field on patron registration" in my 2.0.6 system.
#13:12:38tsbereasimon_: As I said, that is a 2.1+ feature. You have to manually edit a file in 2.0.
#13:12:52asimon_tsbere: Where can I find the Internet Access Entries in the database?
#13:13:03gmcharltasimon_: config.net_access_level
#13:13:25asimon_gmcharlt: Thanks.
#13:14:10dbsjeff++ # multiple book detection was the culprit with our self-check, it seems
#13:14:18jeffhooray!
#13:14:21jeffglad to hear.
#13:15:11lisppastephasefx pasted "another take at long-line highlighting in vim, this time with a toggle" at http://paste.lisp.org/display/122764
#13:16:10asimon_tsbere: Which file contains the required/optional toggle for the County field?
#13:16:59tsbereasimon_: You have to edit register_table.tt2, I think it was, to add the required="required" (or change required="show" to required="required") entry
#13:19:22phasefxincidentally, you can edit a copy of register_table.tt2 in another directory to as a local template if you add that directory to oils_web.xml
#13:21:29dbstemplates++
#13:22:14atz__ has quit IRC
#13:25:39jenny has quit IRC
#13:32:10asimon_tsbere: Thanks.
#13:40:16asimon_Can someone verify that the permission.grp_tree.perm_interval entry for 1 year should be "1 year" and not "1 years"?
#13:40:33dbseither one should work
#13:41:24eeevilsomeone with wiki admin creds mind poking the miker account? it's not sending me a pw reset email, and I can't log in :(
#13:41:43asimon_dbs: Thanks.
#13:41:51dbseeevil: will do
#13:41:58eeevildbs: thankee sai
#13:43:16dbsin theory, you should have an email at your gmail.com address real soon now
#13:44:23eeevildbs: thanks ... any idea from what addr this might be coming? perhaps it's getting mis-filed somewhere
#13:45:01dbseeevil: lemme check
#13:45:04eeevilyep ... it got sent to the spam folder
#13:45:07dbs(also, check your spam)
#13:45:10dbsthere ya go
#13:47:12asimon_Has anyone created a County drop-down box for the Register Patron screen?
#13:50:03bshumasimon_: I think that could be assigned using the zip code generator.
#13:51:07asimon_bshum: Thanks.
#13:51:38lisppastefrank_ pasted "untitled" at http://paste.lisp.org/display/122766
#13:52:23frank_i did those steps, but i doesnt get the email, what else do i have to do?
#13:52:44bshumasimon_: For some clues to that: http://dmagick.wordpress.com/2011/01/06/notes-on-installing-zipcode-information/
#13:54:43atz has joined #evergreen
#13:56:13tsbereQuestion: is class->search_where searching basically a json search on that specific class?
#14:03:52jenny has joined #evergreen
#14:08:04senatortsbere: sorta like that. see class::dbi docs
#14:20:19DmagickOff the top of my head, I think bshum is correct. If counties have been added into the file, then it should populate.
#14:20:40bshumDmagick: They do, ours did, even though we don't actively track counties.
#14:23:21lisppastetsbere pasted "closed date query" at http://paste.lisp.org/display/122767
#14:23:31tsberesenator: Any thoughts on that paste? ^^
#14:27:54senatortsbere: eeevil may be able to help you better, he knows/uses class::dbi more than i do. it looks reasonable, but i'd only know for sure by running it
#14:29:32asimon_ has quit IRC
#14:31:47eeeviltsbere: search_where is added by Class::DBI::AbstractSearch -> http://search.cpan.org/~miyagawa/Class-DBI-AbstractSearch-0.07/lib/Class/DBI/AbstractSearch.pm
#14:32:25eeevilIIRC, it returns an iterator in scalar context
#14:33:02eeevilbut ... what in particular is your where-clause trying to do? find closings that span "today"?
#14:36:00Dyrconaberick: poke.
#14:36:55tsbereeeevil: Looking for all-day or multi-day closings that include the current termination date for a circulation + grace period. Working on making grace period extend through closed days of the "day of week" and "explicit entry" varieties.
#14:38:10tsbere should probably add a "limit => 1" in there somewhere as well
#14:39:28eeeviltsbere: there's logic ... somewhere in there ... to take a timestamptz and return a start/end pair (that might both be the original timestamptz) that describe the bounds of any overlapping closed dates and hours-of-operations closings
#14:40:40eeevil(not that that's an answer to your question...)
#14:41:24eeevilbut, the answer is, I believe, "cast to a perl array if you want to test found-ness, since an iterator will be defined regarless of size"
#14:43:37tsbereHuh. I have a ) where I want a } in the paste. <_<
#14:45:22phasefx has quit IRC
#14:47:00lisppastetsbere annotated #122767 "While Clause <_<" at http://paste.lisp.org/display/122767#1
#14:47:15tsbere hasn't even checked if that is valid perl yet
#14:48:59Dyrconaperl -c sez?
#14:50:09eeeviltsbere: you might want to look at OpenILS::Application::Storage::Publisher::actor::new_org_closed_overlap (which, currently, is hidden via api_level=>0, but probably is doing what you want if you supply a direction of 1)
#14:50:41phasefx has joined #evergreen
#14:51:25eeeviltsbere: the adjacent org_closed_overlap sub is what's in production atm
#14:52:11eeeviland, currently, works fine. the new one is an attempt to use DateTime sets to abstract away some of the complexity
#14:54:05tsbereeeevil: Given that I am in Application/Storage/Publisher/action.pm I am not sure I have much in the way of worries on api level.
#14:54:32tsbere is in the fine generator, specifically
#14:55:42eeeviltsbere: method_lookup to the rescue!
#14:56:15tsbereeeevil: My problem with generic searching of "in an interval the library is closed during" is "I don't want to count the fact they are closed because it is midnight"
#14:58:47eeeviltsbere: if I'm understanding correctly, open-ils.storage.actor.org_unit.closed_date.overlap doesn't do what you don't want
#14:59:06eeevilit only cares about HOO values where the start and end are both 00:00:00
#15:09:22tsbereeeevil: I am looking for "hoo says they are closed all day OR there is an explicitly all day closed date" specifically, unless you think grace period auto-extend should apply to any time the grace period falls in a closed date period of any kind. I am currently running on 'the grace period has to exist and be > 1 day before we consider extending it' to start with.
#15:10:00tsbereWell, >= 1 day
#15:32:06Dyrcona@later tell berick Please look at https://bugs.launchpad.net/evergreen/+bug/799864 when you have a chance.
#15:32:06pinesol_greenDyrcona: The operation succeeded.
#15:35:01eeeviltsbere: I think, without looking at the context of the code, I'm not following. that code I pointed at will return either undef (date you passed does not overlap anything) or start and end of the overlapping closure (regardless of the number or type of adjacent/overlapping closures it takes to calculate that start/end pair)
#15:37:37tsbereeeevil: With context: http://git.mvlcstaff.org/?p=tsbere/ILS.git;a=blob;f=Open-ILS/src/perlmods/lib/OpenILS/Application/Storage/Publisher/action.pm;h=73f4d32731beb9546f59c2e0dafcda314076ae8b;hb=refs/heads/grace_auto_extend#l901
#15:38:33tsbereeeevil: My goal is to extend grace period if and only if the library was closed for the entire day the grace period ended on. Partial-day closings and such would not count.
#15:39:01tsbere isn't done, but did a commit+push to provide context
#15:40:00sfortin has quit IRC
#15:56:06eeevilok, I see. in that case, calling open-ils.storage.actor.org_unit.closed_date.overlap with due_date+grace_interval will give you a fast-path and a case to test. the fast-path is a return value of undef (no overlap) and the test would then be something like "grace_start = (due_date + grace_interval) % day; grace_end = grace_start + 1 day; if (overlap_start <= grace_start AND overlap_end >= grace_end) then extend grace"
#15:56:39eeevilor similar
#15:59:36Meliss has quit IRC
#16:00:05collum has quit IRC
#16:03:35dbs has quit IRC
#16:10:01berickDyrcona: I grabbed the ticket. cursory glance looks good. I'll review further and probably merge, probably this week.
#16:10:16berick needs to review copy active_date first
#16:10:17Dyrconaberick: thank you muchly.
#16:10:43berickDyrcona: thank you!
#16:12:17collum has joined #evergreen
#16:58:21tsbereeeevil: Not 100% sure what you are getting at there. Got very distracted with "our system decided to stop working".
#17:09:16Dyrconayes, but hands have been slapped and the guilty party will be up for reeducation starting tomorrow.
#17:29:24jeff\r\n instead of \n in print_list_templates -- odd.
#17:31:27Dyrcona has quit IRC
#17:31:48pmplett has joined #evergreen
#17:41:40drdata_esi_ has joined #evergreen
#17:41:55Callender_ has joined #evergreen
#17:42:29kmlussier_ has joined #evergreen
#17:45:41joseph_ has quit IRC
#17:47:00jenny has left #evergreen
#17:50:07gmcharlt` has joined #evergreen
#17:50:22kmlussier has quit IRC
#17:50:23natschil has quit IRC
#17:50:23Callender has quit IRC
#17:50:23drdata_esi has quit IRC
#17:50:23atheos has quit IRC
#17:50:23gmcharlt has quit IRC
#17:50:24atheos_ has joined #evergreen
#17:56:15frank_ has quit IRC
#18:15:00kmlussier_ has quit IRC
#19:16:46gmcharlt` is now known as gmcharlt
#19:16:47gmcharlt has joined #evergreen
#21:39:25pmplett has quit IRC
#21:47:22pmplett has joined #evergreen
#22:02:03Dmagick_ has joined #evergreen
#22:08:05pmplett has quit IRC
#22:08:54Callender_ has quit IRC
#22:08:54Dmagick has quit IRC
#22:08:55lisppaste has quit IRC
#22:08:55pinesol has quit IRC
#22:17:08pmplett has joined #evergreen
#22:18:52pmplett has quit IRC
< Sunday, June 19th, 2011Raw Log FileTuesday, June 21st, 2011 >