| # | Time | Nick | Message |
|---|
| # | 00:15:30 | bshum has quit IRC |
| # | 00:15:59 | bshum has joined #evergreen |
| # | 00:38:25 | bshum has quit IRC |
| # | 00:38:25 | bshum has joined #evergreen |
| # | 00:44:54 | mfaizch | hi i have an installation problem can anybody help |
| # | 00:48:41 | mfaizch | ./autogen.sh -c /data/openils/conf/opensrf_core.xml -u |
| # | 00:48:58 | mfaizch | this command is giving error |
| # | 00:51:03 | mfaizch | ./autogen.sh -c /openils/conf/opensrf_core.xml -u |
| # | 00:51:34 | mfaizch | error == Updating Evergreen organization tree and IDL using '/openils/conf/opensrf_core.xml' |
| # | 00:51:53 | mfaizch | Updating fieldmapper |
| # | 00:52:42 | phasefx | mfaizch: I'm responding to your email as we speak |
| # | 00:52:47 | mfaizch | Exception: OpenSRF::EX::Session 2011-08-18T00:50:46 OpenSRF::Utils::SettingsClint /usr/local/share/perl/5.10.0/OpenSRF/Utils/SettingsClient.pm:103 Session Error: router@private.localhost/opensrf.settings IS NOT CONNECTED TO THE NETWORK!!! |
| # | 00:53:22 | mfaizch | please guide me out of this as i have stoped believing it will ever work |
| # | 00:53:47 | mfaizch | i am following the document as it is but no use |
| # | 00:56:18 | phasefx | mfaizch: if you do this, do you get output? grep \\[ERR /openils/var/log/*.log |
| # | 00:57:06 | phasefx | if you do, use http://paste.lisp.org/new/evergreen to share |
| # | 00:57:43 | phasefx | (and paste the resulting URL back into the channel, since the paste bot appears to be not on channel today) |
| # | 01:06:21 | mfaizch | it repeats the same message again and again |
| # | 01:06:34 | mfaizch | which i pasted |
| # | 01:06:57 | mfaizch | Exception: OpenSRF::EX::Session 2011-08-18T00:50:46 OpenSRF::Utils::SettingsClint /usr/local/share/perl/5.10.0/OpenSRF/Utils/SettingsClient.pm:103 Session Error: router@private.localhost/opensrf.settings IS NOT CONNECTED TO THE NETWORK!!! |
| # | 01:09:01 | phasefx | so we want the very error message from /openils/var/log/*.log, preferably after stopping everything, clearing the logs, and then starting things up slowly, one piece at a time |
| # | 01:09:13 | phasefx | the very first error message, I mean |
| # | 01:11:30 | mfaizch | it scrolls away |
| # | 01:11:53 | mfaizch | how do i clear logs |
| # | 01:12:18 | phasefx | if you've already done stop_all, then you can do rm /openils/var/log/*.log |
| # | 01:14:59 | Faiz_ | ok got it all errors cleared |
| # | 01:15:09 | Faiz_ | now what do i do |
| # | 01:15:29 | phasefx | ps -aef | grep opensrf |
| # | 01:15:44 | phasefx | do you see lines like OpenSRF Drone ? |
| # | 01:17:17 | Faiz_ | not at |
| # | 01:17:21 | tsbere has quit IRC |
| # | 01:17:24 | tsbere_ has joined #evergreen |
| # | 01:18:07 | phasefx | Faiz_: alright, so just to be safe, let's be sure services are stopped: /openils/bin/osrf_ctl.sh -l -a stop_all |
| # | 01:19:08 | Faiz_ | confirmed |
| # | 01:19:22 | Faiz_ | it says not running |
| # | 01:19:28 | phasefx | now we start the router like so: /openils/bin/osrf_ctl.sh -l -a start_router |
| # | 01:20:26 | Faiz_ | OpenSRF Router not started, but PID file exists, removing file and starting |
| # | 01:21:17 | phasefx | ooh, I don't like that, means things haven't been stopped cleanly in the past |
| # | 01:21:52 | phasefx | osrf_ctl.sh is smart enough these days to handle stale PID's, but let's be extra careful and do /openils/bin/osrf_ctl.sh -l -a stop_all again |
| # | 01:22:05 | phasefx | then do rm /openils/var/run/* |
| # | 01:22:34 | Faiz_ | i used this command and it started without error |
| # | 01:22:51 | phasefx | and delete the logs again rm /openils/var/log/*.log |
| # | 01:23:13 | Faiz_ | /openils/bin/osrf_ctl.sh -c /openils/conf/opensrf_core.xml -l -a stop_all |
| # | 01:23:32 | Faiz_ | ok i will follow your commands |
| # | 01:26:03 | Faiz_ | done |
| # | 01:26:40 | phasefx | alright, so if everything is stopped and the logs cleared and /openils/var/run is emptied, then let's start the router again: /openils/bin/osrf_ctl.sh -l -a start_router |
| # | 01:27:08 | Faiz_ | yyes it says |
| # | 01:27:20 | Faiz_ | Starting OpenSRF Router |
| # | 01:27:29 | Faiz_ | without error |
| # | 01:27:29 | phasefx | now let's grep the logs for errors: grep \\[ERR /openils/var/log/*.log |
| # | 01:28:07 | Faiz_ | grep: /openils/var/log/*.log: No such file or directory |
| # | 01:28:46 | phasefx | hrmm.. do you have opensrf_core.xml configured to use syslog? or to log to some other directory? |
| # | 01:29:16 | Faiz_ | no i have followed the document no persnal settings |
| # | 01:30:21 | Faiz_ | in srfsh the log path was changed as /tmp/srfsh.log but not other than that |
| # | 01:30:55 | Faiz_ | should i chk this file |
| # | 01:31:00 | Faiz_ | right away |
| # | 01:32:14 | phasefx | in /opensrf/conf/opensrf_core.xml, there should be <logfile> stanzas that have paths including /openils/var/log/ The ones mentioning syslog should be commented out (with <!-- and -->) |
| # | 01:32:47 | phasefx | if you do ps -aef | grep Router do you see any OpenSRF Router's? |
| # | 01:33:55 | Faiz_ | yes 2 of them |
| # | 01:34:07 | Faiz_ | opensrf 5558 1 0 01:26 ? 00:00:00 OpenSRF Router |
| # | 01:34:18 | Faiz_ | opensrf 5557 1 0 01:26 ? 00:00:00 OpenSRF Router |
| # | 01:34:32 | phasefx | that's good, though it worries me there's nothing in /openils/var/log/, but maybe it's my logging level (which is non-standard) that makes them show up right away for me |
| # | 01:35:02 | phasefx | let's try to the perl services next: /openils/bin/osrf_ctl.sh -l -a start_perl |
| # | 01:35:27 | phasefx | if you're familiar with top, run top after doing that and wait until the CPU usage dies down, then quit top |
| # | 01:38:21 | Faiz_ | pearl says everthing started and am not not familiar with top i ran it and quit as things settle |
| # | 01:38:43 | phasefx | now let's test the logs again: grep \\[ERR /openils/var/log/*.log |
| # | 01:39:24 | Faiz_ | blank nothing |
| # | 01:39:35 | phasefx | cool |
| # | 01:39:49 | phasefx | alright, now for the c services: /openils/bin/osrf_ctl.sh -l -a start_c |
| # | 01:40:17 | Faiz_ | Starting OpenSRF C (host=localhost) |
| # | 01:40:23 | Faiz_ | it has also started |
| # | 01:40:30 | phasefx | and the logs again: grep \\[ERR /openils/var/log/*.log |
| # | 01:40:47 | Faiz_ | nothing again |
| # | 01:41:56 | phasefx | good |
| # | 01:42:11 | phasefx | now the moment of truth: /openils/bin/autogen.sh -u |
| # | 01:43:06 | Faiz_ | Updating fieldmapper |
| # | 01:43:22 | Faiz_ | Updating web_fieldmapper |
| # | 01:43:38 | Faiz_ | this is the error line |
| # | 01:43:42 | Faiz_ | Can't call method "opac_visible" on an undefined value at org_tree_html_options.pl line 54. |
| # | 01:43:51 | Faiz_ | and more over |
| # | 01:44:24 | Faiz_ | grep \\[ERR /openils/var/log/*.log returns nothing again |
| # | 01:45:58 | phasefx | okay.. are you using the opensrf.xml/opensrf_core.xml based on the .example files from the Evergreen distribution or from OpenSRF? |
| # | 01:46:36 | Faiz_ | yes |
| # | 01:46:47 | phasefx | one good test would be: grep open-ils.actor /openils/conf/opensrf_core.xml |
| # | 01:46:51 | phasefx | should get two lines |
| # | 01:47:18 | Faiz_ | yes |
| # | 01:47:23 | Faiz_ | 2 lines indeed |
| # | 01:47:39 | Faiz_ | <service>open-ils.actor</service> |
| # | 01:50:48 | phasefx | let's try this, start srfsh |
| # | 01:51:27 | Faiz_ | "textcode":"LOGIN_FAILED", |
| # | 01:51:39 | Faiz_ | "desc":"User login failed", |
| # | 01:52:06 | phasefx | no, instead of trying to login, try request open-ils.actor open-ils.actor.org_types.retrieve |
| # | 01:52:09 | Faiz_ | "ilsevent":1000, |
| # | 01:52:20 | phasefx | in srfsh |
| # | 01:52:52 | phasefx | if you get something like [ { "__c":"aout", ... "Consortium".. etc., that's good |
| # | 01:53:30 | Faiz_ | thats what i get verticallly :) |
| # | 01:53:46 | phasefx | now try the same thing, but org_tree instead of org_types |
| # | 01:54:08 | Faiz_ | that would be \ |
| # | 01:54:11 | phasefx | should get some nested "__c":"aou" 's |
| # | 01:54:26 | Faiz_ | request open-ils.actor open-ils.actor.org_tree.retrieve |
| # | 01:54:31 | phasefx | right |
| # | 01:55:06 | Faiz_ | it says command not found |
| # | 01:55:37 | phasefx | double-check spelling |
| # | 01:56:04 | Faiz_ | request open-ils.actor open-ils.actor.org_tree.retrieve i paste this command in ssh window |
| # | 01:56:41 | phasefx | hrmm, try using the up arrow until you get to the org_types command, and then change it directly, instead of cut/paste |
| # | 01:57:27 | Faiz_ | ok my mistake |
| # | 01:57:31 | Faiz_ | it says |
| # | 01:57:34 | Faiz_ | Received no data from server |
| # | 01:57:47 | Faiz_ | and Request Completed Successfully |
| # | 01:58:00 | Faiz_ | Request Time in seconds: 0.108489 |
| # | 01:58:06 | phasefx | now that's very weird, but would explain the error with org_tree_html_options.pl |
| # | 01:58:35 | phasefx | have you modified the org tree at all (say, in psql?) |
| # | 01:59:02 | Faiz_ | no not at all |
| # | 01:59:10 | Faiz_ | where would i do that |
| # | 01:59:11 | phasefx | and, any errors in /openils/var/log/ ? |
| # | 01:59:32 | phasefx | it would be a database table, actor.org_unit, but if you left it alone, we'll ignore that avenue |
| # | 02:00:06 | Faiz_ | no errors still |
| # | 02:00:30 | Faiz_ | yes i did left it alone |
| # | 02:00:57 | phasefx | alright, you should update the mailing list with your new error, and the fact that open-ils.actor.org_tree.retrieve is returning no data from the server in srfsh |
| # | 02:01:06 | phasefx has to get some sleep, alas |
| # | 02:04:37 | Faiz_ | ok willl do that |
| # | 02:05:39 | phasefx | if you would, try replying to my last email, so that things will be threaded in folks' mail readers |
| # | 02:06:29 | phasefx = Jason Etheridge on the list |
| # | 02:06:52 | mfaizch | oh its you |
| # | 02:07:04 | mfaizch | was looking for you since day before yesterday |
| # | 02:07:11 | mfaizch | wait for a minute |
| # | 02:07:20 | mfaizch | i will post it as you say |
| # | 02:08:19 | Faiz_ | as in my email i have first error on memcached |
| # | 02:09:34 | Faiz_ | and i follow the document as it it is why is my memcached not running it shows disabled and than i have to enable it in # /etc/default/memcached |
| # | 02:10:41 | Faiz_ | was thinking will get the error if we start from start |
| # | 02:10:56 | phasefx | incidentally, just searching the mailing list, other folks have had the method "opac_visible" error. There are multiple potential causes |
| # | 02:11:18 | Faiz_ | ok am listning |
| # | 02:12:18 | phasefx | I don't see any that might apply to you just skimming quickly. You're not changing language settings in the config files, for example |
| # | 02:12:33 | phasefx | older versions of EG too |
| # | 02:13:09 | Faiz_ | no not at all no change in language settings |
| # | 02:13:25 | phasefx | alright, must really go now. Hopefully someone will have an "ah ha" moment, and respond to your email |
| # | 02:14:14 | Faiz_ | ok will post on the mailing list |
| # | 02:14:36 | Faiz_ | am trying to install it from las 12 days |
| # | 02:14:45 | Faiz_ | and stuck with it |
| # | 02:15:59 | phasefx | probably something silly we're overlooking. g'night (from my perspective:) |
| # | 02:17:11 | Faiz_ | yes good noght have nice dreams wit Faiz running Evergreen :) |
| # | 03:04:22 | Faiz_ has quit IRC |
| # | 03:09:49 | mfaizch | is therer any one who can help |
| # | 03:52:27 | mfaizch | i have an installation problem can any body help |
| # | 04:14:18 | hughbris has joined #evergreen |
| # | 04:44:51 | edoceo has quit IRC |
| # | 05:00:23 | bjwebb has joined #evergreen |
| # | 05:00:23 | bjwebb has joined #evergreen |
| # | 06:34:00 | natschil has joined #evergreen |
| # | 07:31:47 | dbs has joined #evergreen |
| # | 07:33:28 | kmlussier has joined #evergreen |
| # | 08:26:04 | tsbere_ is now known as tsbere |
| # | 08:27:03 | sfortin has joined #evergreen |
| # | 08:29:11 | collum has joined #evergreen |
| # | 08:29:11 | atheos has quit IRC |
| # | 08:29:58 | atheos has joined #evergreen |
| # | 08:31:03 | bjwebb has quit IRC |
| # | 08:51:31 | mrpeters-isl has joined #evergreen |
| # | 09:14:04 | collsk12 has joined #evergreen |
| # | 09:15:46 | collsk12 | Good morning all! I have a quick (hopefully) question for the room: Any idea why I would be getting a "Can't call method "opac_visible" on an undefined value at org_tree_html_options.pl line 54." error, when running autogen.sh for the first time? I tried Googling to no avail. |
| # | 09:17:24 | akilsdonk has joined #evergreen |
| # | 09:19:27 | tsbere can never remember how to link to a mailing list message. Or rather where to link to them from |
| # | 09:20:15 | dbs generally goes to markmail, then finds the original archive entry on list.georgialibraries.org by date |
| # | 09:20:29 | tsbere | collsk12: http://libmail.georgialibraries.org/pipermail/open-ils-dev/2011-August/007479.html ? |
| # | 09:21:44 | collsk12 | tsbere: You rock! Thank you so very much!! |
| # | 09:22:55 | dbs | so what's our strategy when we know the current stable release is broken out of the box? |
| # | 09:23:34 | phasefx | assign blame, drive our developers before us, and hear the lamentations of their spouses |
| # | 09:24:28 | tsbere blames postgres |
| # | 09:24:30 | tsbere | :P |
| # | 09:24:47 | gmcharlt | invest in the stock of brown bag manufacturers? |
| # | 09:26:50 | gmcharlt | though seriously, is it just time to do a cut of 2.0.9, or am I missing something? |
| # | 09:26:56 | dbs | Well - do we pull the 2.0.8 release and point back to 2.0.7 until we confirm that https://bugs.launchpad.net/evergreen/+bug/826844 resolves the problem and push out 2.0.9? |
| # | 09:27:43 | dbs | and do we do any reflection like "Geez, what steps can we take to prevent this particular problem from happening again in future releases?" |
| # | 09:27:48 | StephenGWills has joined #evergreen |
| # | 09:30:03 | yboston has joined #evergreen |
| # | 09:30:54 | dbs | In which case a possibility would be to add to the testing checklist: "Add a new library to the org tree; run autogen.sh; confirm that you can add users / items to that library" |
| # | 09:31:49 | tsbere | dbs: Amazingly enough, that might not have caught this issue |
| # | 09:31:50 | collsk12 | The fix worked. My suggestion might be to update the wiki with a note referencing the bug. |
| # | 09:32:16 | dbs | tsbere: I'm saying add a test that would catch the issue, obviously |
| # | 09:33:16 | dbs | although if reading 826844 says that postgresql's behaviour is variable, then we're doomed |
| # | 09:33:17 | tsbere | dbs: The code as is works fine for me in 2.1/master on postgres 9, so if I was running the test (and I did run that kind of test) then I wouldn't have had it fail on me. |
| # | 09:34:08 | dbs | tsbere: so if it's an 8.4 difference, then that suggests that one person running a test in one environment - or a bunch of people running a test but all in the same environment - isn't enough |
| # | 09:34:34 | dbs | and if it's a 2.0 vs. 2.1 difference, did anyone test it in 2.0? |
| # | 09:34:35 | tsbere admits to not having a 2.0 testing setup with postgres 8.4 |
| # | 09:34:44 | tsbere | or a 2.0 testing setup at all |
| # | 09:34:56 | collsk12 | Is there any talk about moving away from postgres? Maybe towards mysql? |
| # | 09:35:02 | mrpeters-isl has quit IRC |
| # | 09:35:11 | mrpeters-isl has joined #evergreen |
| # | 09:35:19 | dbs looks at http://testing.evergreen-ils.org still running but way below it's potential |
| # | 09:35:22 | dbs | s/it's/its |
| # | 09:35:57 | phasefx | collsk12: no, we depend too heavily on postgres specific stuff |
| # | 09:36:02 | tsbere | How difficult would it be to rig stuff up to allow for someone to literally run multiple postgres server instances on a test box *just* to make sure the base schema appears to load? |
| # | 09:36:42 | collsk12 | phasefx: Worth a shot ;) Thanks. |
| # | 09:37:23 | dbs | collsk12: I would argue that mysql is worth a shot - of the lead variety |
| # | 09:37:35 | natschil has quit IRC |
| # | 09:38:30 | dbs | tsbere: postgresql 8.4 / 9.0 happily co-exist on the same box |
| # | 09:38:38 | berick | and that was worth a shot.. of the rim variety *rimshot* |
| # | 09:38:43 | collsk12 | dbs: Oooh...not a fan, I presume. |
| # | 09:38:48 | tsbere | dbs: I know that. I was thinking from the buildbot side more than anything |
| # | 09:38:53 | dbs | collsk12: you presume correct :) |
| # | 09:39:37 | eeevil | collsk12: fair warning, dbs has been heavily steeped in DB2 ;) |
| # | 09:39:39 | dbs | tsbere: the difficulty probably lies in getting the credentials right |
| # | 09:39:59 | dbs | berick++ |
| # | 09:40:20 | dbs | berick: we're finished our "escape output" pass over the T-PAC? |
| # | 09:40:22 | tsbere | dbs: can buildbot be told "this test is to run this file you don't control"? Because I am willing to set up a VM to do nothing but sit there and let buildbot tell it to test the base schemas. |
| # | 09:40:23 | eeevil | (fairer warning -- /I've/ lost so much time and data to mysql that I feel reparations are in order) |
| # | 09:40:26 | phasefx | collsk12: my favorite is how mysql will "helpfully" truncate data, say if you overflow a numeric :) |
| # | 09:41:02 | collsk12 | phasefx: LOL! |
| # | 09:41:19 | berick | dbs: as far as I can tell, we slayed the beast |
| # | 09:41:22 | collsk12 | eeevil: DB2, eh? |
| # | 09:41:26 | berick does another quick scan |
| # | 09:41:49 | berick | dbs: you looked at parts/advanced? |
| # | 09:42:03 | eby | phasefx: yeah ran into that with a primary key once |
| # | 09:43:01 | dbs | tsbere: buildbot can be told to do just about anything, so running a particular script is fair game; generally it's considered a best practice to try and build as much into the Makefile as possible |
| # | 09:43:13 | phasefx | eby: oy |
| # | 09:44:21 | dbs | so if there was a "make test_schema" target that did nothing in the absence of a file containing database connection parms, that would be a possibility |
| # | 09:45:46 | dbs | berick: I would still like to factor out the "<big><strong>" type markup and inline style attributes and such back into CSS, but that can be done in master |
| # | 09:46:00 | berick | dbs: you'll find no argument here |
| # | 09:46:18 | dbs | berick: Oh, come on, you love big strong HTML |
| # | 09:46:29 | berick | it makes me feel safe and warm |
| # | 09:46:49 | berick | and surrounded by markup |
| # | 09:48:12 | phasefx | marc, 21, slim, not |
| # | 09:48:33 | bjwebb has joined #evergreen |
| # | 09:48:33 | bjwebb has joined #evergreen |
| # | 10:01:26 | natschil has joined #evergreen |
| # | 10:07:18 | collsk12 has quit IRC |
| # | 10:09:22 | tsbere | dbs: I suspect that make test_schema should, in part, drop the entire database and make it from scratch |
| # | 10:09:29 | dbs | so muhammad faiz's problem is likely the 2.0.8 problem? |
| # | 10:10:16 | dbs | tsbere: if we're testing clean installs, for sure; that means ensuring that the user has perms to createdb, createlang, etc, so more sysadmin work to setup the env correctly |
| # | 10:11:00 | tsbere | dbs: Easy solution? Allow passwordless sudo to a make target that does all of that for the postgres user. ;) |
| # | 10:11:30 | tsbere | Even better, if written correctly that make target can be used for actual installs |
| # | 10:12:31 | tsbere | dbs: Alternatively, use the remote methods of doing everything, actually |
| # | 10:16:43 | dbs | anything that makes actual installs easier _and_ increases testing capability has to be good |
| # | 10:17:09 | tsbere | dbs: I will work on a shell script that the makefile could call |
| # | 10:17:32 | tsbere | Or rather, a sql file that can be called with psql and a -v option |
| # | 10:18:08 | dbs | tsbere: cool - will the makefile then call eg_db_config.pl to actually create the schema? |
| # | 10:18:24 | tsbere | dbs: That would be my view |
| # | 10:18:32 | dbs | sounds great |
| # | 10:19:02 | dbs | phasefx: thanks for the autogen.sh pull |
| # | 10:19:35 | tsbere wishes the phone would stop ringing, though |
| # | 10:20:06 | phasefx | dbs: 2.1 is still an option (if such changes seem okay for 2.1), but I don't have a setup for testing at the moment |
| # | 10:20:41 | dbs | phasefx: we kind of have to get it into 2.1, as otherwise people will be calling cache-generator.sh for one release |
| # | 10:20:56 | phasefx nods |
| # | 10:22:46 | sal_ has joined #evergreen |
| # | 10:24:46 | sal_ | Windows client (from trunk) compiled/installed/runs fine with xulrunner 3.6.20. If there was a change to application.ini I was supposed to make, I didn't (and it seems to work fine.) |
| # | 10:25:40 | bshum | sal_++ for testing that |
| # | 10:28:14 | sal_ | :-) |
| # | 10:28:15 | dbs | phasefx: Once the T-PAC is merged, I think master should become 2.1-RC3 anyways |
| # | 10:29:02 | phasefx | (. O) |
| # | 10:31:25 | berick | heh |
| # | 10:33:16 | sal_ | * exposes ignorance... Would like to patch in the chargeable loans feature (bug 806131) into the build I'm working on (gitted trunk/master originally). And I can't seem to figure out how. Do I need to start from scratch, or is there a way to patch in the changes after the fact? |
| # | 10:33:25 | gmcharlt | dbs: that's one way to stave off the day until we have to actually write release notes ;) |
| # | 10:35:04 | sal_ | (swears she could do this in svn at one point ;-) ) |
| # | 10:36:54 | dbs | gmcharlt: we should probably be writing release notes when features get merged and details are fresh in our minds anyway |
| # | 10:37:51 | natschil has quit IRC |
| # | 10:38:04 | dbs | I'm just frustrated to have features like the openlibrary integration merged a couple of days past the feature cut-off date and not likely to see the light of an actual release until, say, April |
| # | 10:38:24 | gmcharlt | dbs: I agree in principle. to toss out an idea, for major commits or major merge commits that itnroduce a new feature, maybe add decorate to the commit message to indicate that the message is a detailed description suitable for release notes |
| # | 10:38:35 | gmcharlt | might be one way of filtering noise |
| # | 10:39:13 | gmcharlt | even if say a DIGer later rewrites the note, that gives a start point |
| # | 10:39:43 | dbs | "decorate" - so, say we merge the T-PAC today, is that just a convention in the commit message? |
| # | 10:40:27 | gmcharlt | yeah, e.g., something like - "for-release-notes" in the commit message |
| # | 10:41:23 | dbs | sal_: you could do something like "git diff origin/master working/Evergreen <email address hidden>/SIP2_chargeable_loans" and try applying the diff to your build, if your build isn't in git |
| # | 10:41:35 | dbs | (that's probably a horrible way to do things, but I'm a horrible person) |
| # | 10:42:23 | gmcharlt notes that for that purpose, 'git apply' is one of the git subcommands that *doesn't* require that you be inside a git repo |
| # | 10:44:03 | dbs | berick: should we go ahead and wire up the "Basic catalog" link in the JSPAC to point at the TPAC pre-commit? |
| # | 10:44:25 | berick | dbs: let's do it |
| # | 10:44:49 | dbs | I guess we'd need a corresponding "Dynamic catalog" link in the TPAC, although in practice I bet that there are few people who want to go in that direction |
| # | 10:45:05 | dbs | Cool - I was just thinking that would help the for-release-notes commit message :) |
| # | 10:45:13 | dbs will make it so |
| # | 10:45:16 | berick | agreed |
| # | 10:45:58 | bshum | leed++ # I'll try some of those diffs to see how they play with our system, but thanks muchly! |
| # | 10:48:00 | leed | bshum: no problem, the create views at the bottom were added because of the drop cascades i had to do earlier in the diff. Was a 2.0.7 DB prior to the upgrade, I'm hearing I was supposed to start with 2.0.8. Perhaps the script should have its name changed to 208-21-upgrade! Good luck. |
| # | 10:54:57 | tsbere | dbs: http://paste.lisp.org/display/124109 |
| # | 11:02:17 | tsbere | dbs: I wonder if this shouldn't be placed in, say, the src/sql/Pg directory and added as a step in eg_db_config.pl. Say via a "--create-database" option? |
| # | 11:07:03 | dbs | berick: merged the reciprocal T-PAC / JSPAC links |
| # | 11:07:10 | dbs | and pushed |
| # | 11:07:17 | dbs looks at tsbere's link |
| # | 11:08:50 | berick | dbs: cool. I'll give a little more time for nay's and merge, i guess, this afternoon |
| # | 11:09:17 | berick | if anyone has anything in particular they'd like to add to the merge commit message, i'm all ears |
| # | 11:09:45 | dbs | tsbere: sounds like a reasonable plan |
| # | 11:10:20 | tsbere | dbs: I can try and make it happen in a branch for master. Any backporting for 2.0/2.1 will need to be looked at after (and likely have the script adjusted for 2.0) |
| # | 11:10:48 | dbs | tsbere: well, just getting it into master would be a win |
| # | 11:15:17 | natschil has joined #evergreen |
| # | 11:18:53 | natschil has quit IRC |
| # | 11:21:37 | dbs | berick: I'm going to be out of "the office" for a few hours soon, but will try to help out with a merge commit message when I get back |
| # | 11:22:07 | dbs winces at the outdated http://evergreen-ils.org/dokuwiki/doku.php?id=faqs:evergreen_roadmap |
| # | 11:22:11 | berick | dbs++ |
| # | 11:23:36 | natschil has joined #evergreen |
| # | 11:25:50 | bradl | dbs: should put something like teleportation in there just to make sure people are paying attention |
| # | 11:27:28 | bradl | -= THIS MESSAGE NOT LOGGED =- |
| # | 11:36:48 | dbs | teleportation would cut down on transit times; we should get to work on that |
| # | 11:37:30 | gmcharlt | and a partial implementation of teleportation could make it easier to stretch materials budgets |
| # | 11:38:07 | tsbere | teleportation could enable opac-based checkins. And checkouts. And fetching of overdue materials automatically. Etc. |
| # | 11:38:36 | StephenGWills hopes the new feature comes with fly-paper |
| # | 11:40:54 | bradl feels an Apr 1 press release forming |
| # | 11:58:57 | dbs adds tiny "Development conventions" section to http://evergreen-ils.org/dokuwiki/doku.php?id=dev:opac:template-toolkit |
| # | 11:59:50 | dbs | unfortunately no teleportation section yet |
| # | 12:01:08 | tsbere tests his new --create-database option to eg_db_config.pl |
| # | 12:11:16 | dbs | Updated some of the top-level todos on http://evergreen-ils.org/dokuwiki/doku.php?id=dev:opac:template-toolkit:todos |
| # | 12:17:48 | tsbere | dbs: https://bugs.launchpad.net/evergreen/+bug/828878 |
| # | 12:22:43 | dbs | tsbere: thanks, I'll try it out this afternoon |
| # | 12:22:49 | dbs heaps more onto the TODO pile |
| # | 12:27:58 | tsbere | dbs: Not sure how to do the "make test-database" part at this point. Check for a file that contains paramaters to eg_db_config.pl ? |
| # | 12:31:11 | natschil has quit IRC |
| # | 12:35:50 | bjwebb has quit IRC |
| # | 12:50:16 | sfortin has quit IRC |
| # | 13:04:02 | akilsdonk has quit IRC |
| # | 13:19:40 | tsbere | Less than fun: Sending someone a spreadsheet (xls format), having them print it out, fill in the missing column by hand, and then mailing it back to you for you to process. |
| # | 13:32:30 | eeevil | tsbere: did they take a poloroid picture of it on a wooden background, scan that, print that, and then mail it? because that would circle all the away around not-fun back to awesome |
| # | 13:34:00 | tsbere | eeevil: No. I think this second one passed through the copy machine before and after being filled in by hand, though. |
| # | 13:35:35 | bshum | selfcheck-mdlbry |
| # | 13:35:47 | bshum | Oops, barcode scanner in the wrong window :S |
| # | 13:35:57 | bshum | Sorry about that folks. |
| # | 14:59:50 | StephenGWills has left #evergreen |
| # | 15:02:45 | sfortin has joined #evergreen |
| # | 15:05:22 | akilsdonk has joined #evergreen |
| # | 15:12:02 | mfaizch has quit IRC |
| # | 15:14:13 | dbs | tsbere: "eg_db_config.pl --create-database", and "eg_db_config.pl --create-database --create-schema", appear to do what one would want right out of the box on Fedora 15; nice work with pg_config |
| # | 15:15:04 | tsbere | :D |
| # | 15:15:36 | tsbere | I did not want to hardcode the path for the contrib dir, because I have no clue if it is installed elsewhere on other distros we are supporting. |
| # | 15:15:51 | tsbere | Or might support in the future, for that matter |
| # | 15:16:20 | dbs | I'm a-gonna modify the README to take out the manual database creation bits (with the exception of the superuser account) and replace it with --create-database, then merge it |
| # | 15:16:27 | dbs | yeah, excellent thinking |
| # | 15:17:09 | tsbere | dbs: Should the example "complete set of options" in the help for eg_db_config.pl be updated to include the --create-database? |
| # | 15:17:26 | dbs | probably |
| # | 15:17:34 | akilsdonk_ has joined #evergreen |
| # | 15:18:57 | akilsdonk has quit IRC |
| # | 15:19:25 | akilsdonk has joined #evergreen |
| # | 15:19:56 | tsbere | dbs: Also, may want to include instructions on manually using create_database.sql when, for whatever reason, you don't have the contrib directory anywhere but the actual database server. |
| # | 15:20:00 | dbs | I guess this means that clients are going to have to install pg-contrib packages, which probably means installing pg-server, so a possible painpoint for production setups, but I think that's a reasonable tradeoff for the reduction in manual steps |
| # | 15:20:34 | tsbere | I believe I put some basic "how to use this file alone" info in the top of the create_database.sql file |
| # | 15:21:14 | dbs | Yeah, I think that's a good idea. Possibly for the still-mythical "How to set up a real production Evergreen instance" document |
| # | 15:21:45 | tsbere | The fact that the sql file works standalone is a plus, IMO. :D |
| # | 15:21:53 | tsbere | (provided you know what to tell psql) |
| # | 15:22:28 | akilsdonk_ has quit IRC |
| # | 15:27:42 | akilsdonk has quit IRC |
| # | 15:27:54 | akilsdonk has joined #evergreen |
| # | 15:37:40 | dbs | tsbere: committed |
| # | 15:37:44 | tsbere | dbs++ |
| # | 15:37:52 | tsbere is looking over the readme changes now |
| # | 15:38:54 | dbs | if there's a distro out there without pg_config, I don't wanna support it |
| # | 15:39:15 | dbs | we should have used `pg_config --sharedir`/contrib ages ago in the docs. facepalm |
| # | 15:39:29 | tsbere | Hmmm. Both commits have only one sign-off. Oh well. |
| # | 15:40:16 | dbs | ah, crap. forgot to cherry-pick -s your commit into mine in my haste |
| # | 15:41:17 | tsbere | dbs: I only figured out that pg_config --sharedir could help when I read that contrib files are installed in "SHAREDIR/contrib" |
| # | 15:41:20 | dbs | and I (personally) don't consider README commits to be quite as worrisome, but yeah, I could have just counter-posted a signed-off branch for you to then sign |
| # | 15:42:53 | tsbere wonders if this should now be backported, though |
| # | 15:45:46 | dbs | I'll go on record to say I would like to see it backported to 2.1 |
| # | 15:52:55 | dbs notes that various Conifer branches for T-PAC appear not to have been merged - specifically user/dbs/ttopac-refworks and user/dbs/ttopac-sfx-integration |
| # | 15:53:48 | berick | dbs: IIRC, those branches were already merged into other branches that got merged |
| # | 15:53:51 | dbs | also need to provide an unwrapped script for user/dbs/ttopac-metabib-display-field to get it pullrequesty |
| # | 15:53:59 | dbs | berick: but there's no config.tt2 file? |
| # | 15:54:13 | dbs | oh wait |
| # | 15:54:30 | dbs | I'm still on a checkout of tsbere's branch, which has no t-opac |
| # | 15:54:33 | dbs | dbs-- |
| # | 15:54:38 | tsbere | heh |
| # | 15:56:14 | AaronZ-PLS has joined #evergreen |
| # | 15:57:20 | dbs gets back in sync with the t-pac |
| # | 16:07:40 | rangi | i read tupac when i see that |
| # | 16:07:49 | rangi | and start singing california knows how to party |
| # | 16:07:57 | berick does too |
| # | 16:08:17 | berick sings picture me rollin' ;) |
| # | 16:08:29 | rangi | heh |
| # | 16:19:39 | sfortin has quit IRC |
| # | 16:19:40 | natschil has joined #evergreen |
| # | 16:22:45 | collum has quit IRC |
| # | 17:00:13 | akilsdonk has quit IRC |
| # | 17:20:02 | dbs | phasefx: problems creating org units in 2.0.8 smells strongly of the same evil bug |
| # | 17:20:53 | phasefx | did we do the === !== treatment in 2.0? I didn't think so |
| # | 17:23:01 | phasefx | you know, I'm not even sure if we fixed a refresh issue; I was so focused on just getting the org created |
| # | 17:26:49 | yboston has quit IRC |
| # | 17:36:08 | sal_ has quit IRC |
| # | 17:41:31 | phasefx | yeah, in master, New Child doesn't update the interface correctly.. have to reload to see the placeholder org |
| # | 17:56:25 | kmlussier has quit IRC |
| # | 18:33:01 | dbs has quit IRC |
| # | 18:40:47 | dbs has joined #evergreen |
| # | 18:41:01 | dbs | tabs-- |
| # | 18:41:03 | dbs | copying and pasting SQL that contains tabs sucks |
| # | 18:44:20 | dbs | grabbing 0605 |
| # | 19:07:20 | edoceo has joined #evergreen |
| # | 19:08:16 | dbs notes that rel_2_1 didn't have its config.upgrade_log updated for the last three upgrades, heh |
| # | 19:18:30 | tsbere | dbs: Were those upgrades applies to 2.1, or were they 2.0 and/or master only updates? |
| # | 19:18:37 | tsbere | s/applies/applied |
| # | 19:19:11 | edoceo has quit IRC |
| # | 19:19:20 | dbs | tsbere: you could answer that question yourself |
| # | 19:19:49 | tsbere is lazy and trying to figure out how to best lay out a new table or three |
| # | 19:20:04 | tsbere also isn't sure it is important at this point ;) |
| # | 19:22:14 | edoceo has joined #evergreen |
| # | 19:22:15 | phasefx counts dbs and tsbere as parts of his brain (mu hahaha) |
| # | 19:22:17 | dbs | http://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=e3fd5900b8743458c48876a87304087c1d4c025f |
| # | 19:26:28 | tsbere isn't sure why he bid internally on something that requires him to edit the stat cat editing interface again. |
| # | 19:26:32 | tsbere | I hate editing that interface |
| # | 19:27:54 | dbs | tsbere++ |
| # | 19:27:59 | dbs | masochist |
| # | 19:28:37 | tsbere thinks it is because it will help with things like paying off his new car |
| # | 19:29:03 | dbs thinks dbwells' ou parent-protect fix would be a perfect use of the deprecates / supersedes stuff, but doesn't see any existing uses to date and doesn't want to step into that pile of unexercised worms |
| # | 19:29:15 | dbs | tsbere: what make/model? |
| # | 19:29:28 | tsbere | Honda Insight |
| # | 19:29:33 | dbs | oh nice |
| # | 19:30:46 | edoceo has quit IRC |
| # | 19:31:38 | tsbere | Hmmm |
| # | 19:32:33 | tsbere | Over 4 weeks now, actually, and still running on the original tank of gas from the dealer. |
| # | 19:32:38 | tsbere | :D |
| # | 19:32:43 | dbs | very nice :) |
| # | 19:33:46 | dbwells | Just in case people get anxious during the night and decide to roll 2.0.9, I wish to lobby that this get committed first as well. https://bugs.launchpad.net/evergreen/+bug/827523 It's a simple fix code-wise, and while not quite the problem that the other bug is, it is still much needed. |
| # | 19:34:24 | dbwells | I can also note that we are running it on our production 2.0 instance. |
| # | 19:35:16 | dbwells goes back to eating dinner |
| # | 19:36:38 | tsbere has a second project he has been told to go forward on that was originally a single OU setting and related code, but that he decided to split into two OU settings, and now is debating adding a third. :( |
| # | 19:39:53 | tsbere | Sometimes I think I put too much thought into making things flexible |
| # | 19:42:59 | edoceo has joined #evergreen |
| # | 19:47:16 | dbs is all for not making things flexible if there's a good argument that there is one true way of doing things |
| # | 19:47:29 | dbs | avoiding config hell is good |
| # | 19:48:04 | phasefx | javascript in tables all the way :D |
| # | 19:48:22 | dbs hopes that tsbere / dbwells will offer some t-pac merge votes at some point |
| # | 19:48:42 | natschil has quit IRC |
| # | 19:49:05 | tsbere | dbs: The problem is I have, in looking how I am going to write the code, come up with good arguments AGAINST saying there is one way of doing things |
| # | 19:49:12 | tsbere | <_< |
| # | 19:49:16 | dbs | hah |
| # | 19:49:57 | tsbere | The only argument for "one way of doing things" in this case, for me, is "the basic specs I have say there is only one way". I am not one to stick to the specs if it is going to be too limiting. |
| # | 19:50:10 | dbs | We're going to need an AI to assist people with Evergreen configuration |
| # | 19:50:31 | phasefx has some ideas for configuring the AI |
| # | 19:50:43 | tsbere | As for t-pac merge votes....I haven't had time to look at the merge branch yet, and thus do not feel confident offering my vote either way. |
| # | 19:50:57 | dbs | 0 is a vote |
| # | 19:53:56 | tsbere | dbs: Is there any reason there is a adv_copy_location_filter_row block that is never referred to anywhere else ever again in the tt opac? |
| # | 20:01:47 | edoceo has quit IRC |
| # | 20:07:11 | jeff | good evening |
| # | 20:08:44 | tsbere | Is it good? I hadn't decided yet. |
| # | 20:20:01 | berick | tsbere: you'll find lots of IDs in the t-pac that are never referenced. it was built from a skin of the JS opac and only partially cleaned up thus far |
| # | 20:21:07 | tsbere has done a wonderful job of making his VM choke |
| # | 20:21:54 | tsbere | Copied a 3.2 gig file into it. I think it ran out of physical allocation and has to wait for the host machine to catch up. |
| # | 22:42:08 | Manderson | Can someone tell me, are there socket files suppose to be created somewhere? http://pastebin.com/ezFeWpGR |
| # | 22:42:40 | Manderson | /data/openils/var/sock/unix_sock was called out, but didn't exist on the filesystem. |
| # | 22:46:57 | phasefx | Manderson: I think later versions of opensrf stopped using them |
| # | 22:50:04 | Manderson | ok. |
| # | 22:50:07 | Manderson | thanks |
| # | 22:54:04 | Manderson | still having issues with the autogen.sh script. Keep getting the message "Can't call method "opac_visible" on an undefined value at org_tree_html_options.pl" |
| # | 22:54:07 | Manderson | any ideas? |
| # | 22:54:38 | phasefx | someone else was getting that last night, and I couldn't figure out why |
| # | 22:55:17 | phasefx | in srfsh, try retrieve open-ils.actor open-ils.actor.org_tree.retrieve |
| # | 22:55:17 | Manderson | I think that was me |
| # | 22:55:33 | phasefx | there was Faiz_ and collsk12 |
| # | 22:57:03 | Manderson | phasefx: is that the exact command? 'retrieve open-ils.actor open-ils.actor.org_tree.retrieve' ? |
| # | 22:57:50 | phasefx | at the srfsh# prompt, which you get with /openils/bin/srfsh (but you may have to configure ~/.srfsh.xml) |
| # | 22:58:11 | Manderson | ok...just making sure. |
| # | 22:58:36 | phasefx | if you get back a data structure containing the word Consortium, that's good |
| # | 22:58:46 | phasefx | if you get No response from server, that's bad |
| # | 22:59:05 | tsbere | If you don't, log in to the database with psql and "select * from actor.org_unit;" to see if you get results. |
| # | 23:00:37 | tsbere has a pile of "how to fix X" and "how to implement Y" to deal with tomorrow after some "test to see if you can figure out why Z doesn't work" |
| # | 23:03:08 | Manderson | hmm: srfsh# retrieve open-ils.actor open-ils.actor.org_tree.retrieve |
| # | 23:03:09 | Manderson | ???: retrieve open-ils.actor open-ils.actor.org_tree.retrieve |
| # | 23:03:51 | Manderson | tsbere: I logged into the db and get 0 rows from that select statement. |
| # | 23:04:14 | tsbere | Manderson: Then I assume you fell victim to the "broken trigger prevents inserts to that table" bug |
| # | 23:05:01 | Manderson | oh. |
| # | 23:05:10 | phasefx | Manderson: sorry, use request instead of retrieve, for the first instance of retrieve |
| # | 23:05:19 | phasefx | but |
| # | 23:05:25 | phasefx | psql already told the true tale |
| # | 23:05:48 | Manderson | Is there a fix for this? |
| # | 23:06:43 | Manderson | phasefx: you're right...the statement worked, but still "Received no data from server |
| # | 23:06:46 | Manderson | " |
| # | 23:08:02 | phasefx | Manderson: https://bugs.launchpad.net/evergreen/+bug/826844 |
| # | 23:09:57 | phasefx | assuming a 2.0 instance of EG, I think you can feed this to psql: http://git.evergreen-ils.org/?p=Evergreen.git;a=blob_plain;f=Open-ILS/src/sql/Pg/upgrade/0605.schema.lp826844_org_unit_parent_protect_fix.sql;h=12cf25cb80a6b3c7a34b77d0fa0e923edb65caf3;hb=41f00b25bb95ac45c826304fb8e17844ad0e6fa2 |
| # | 23:23:42 | Manderson | phasefx: ran that and everything looked like it worked OK...that table still has 0 rows though. Is there a subsequent script/scripts that need to be re-run in order to get that populated? |
| # | 23:26:14 | Manderson | I am guessing the steps in the manual where we create the database schema? |
| # | 23:26:28 | phasefx | oh, yeah, pooh, that's not enough |
| # | 23:48:31 | Manderson | alright, I've re-created the entire database, but I'm still returning 0 rows on that table. |
| # | 23:48:34 | Manderson | thoughts? |
| # | 23:56:14 | phasefx | Manderson: we're going to have cut a bugfix release of EG soon; unless you want to install the 2.0 branch from a git checkout, it may be better to wait for that |
| # | 23:57:02 | phasefx | alternately, you could grab whatever files that launchpad ticket refers to and copy them into your source tree prior to install |
| # | 23:59:40 | Manderson | I'll wait...I am in no hurry. Thanks for your help! |
April 10-13