2011-09-28T00:01:45 *** dbs has quit IRC 2011-09-28T02:40:45 *** bwicksall has quit IRC 2011-09-28T02:53:17 *** bwicksall has joined #evergreen 2011-09-28T05:28:37 *** lisah_ has quit IRC 2011-09-28T06:01:25 *** lisah_ has joined #evergreen 2011-09-28T07:51:09 *** collum has joined #evergreen 2011-09-28T08:00:20 *** artunit has quit IRC 2011-09-28T08:00:29 *** artunit_ has joined #evergreen 2011-09-28T08:00:29 *** artunit_ is now known as artunit 2011-09-28T08:25:50 *** Shae has joined #evergreen 2011-09-28T08:32:19 *** Dyrcona has joined #evergreen 2011-09-28T08:36:22 *** collum has quit IRC 2011-09-28T08:39:51 *** AaronZ-PLS has joined #evergreen 2011-09-28T08:40:28 *** dbs has joined #evergreen 2011-09-28T08:40:59 *** dbs has quit IRC 2011-09-28T08:40:59 *** dbs has joined #evergreen 2011-09-28T08:52:24 *** sal_ has joined #evergreen 2011-09-28T09:01:36 *** kmlussier has joined #evergreen 2011-09-28T09:03:18 is there a recommended way to delete rows from money.materialized_billable_xact_summary? 2011-09-28T09:03:58 You trying to get rid of bad/corrupt ones, or old ones that no longer exist? 2011-09-28T09:05:00 * Dyrcona is feeling wicked and will resist the urge to give bad advice. 2011-09-28T09:05:02 * tsbere notes that the table has no rules preventing "DELETE FROM money.materialized_billable_xact_summary WHERE x" 2011-09-28T09:05:03 I've deleted the billings and payments from before 1/1/2011 so I guess I'd like to remove any from the same timeframe 2011-09-28T09:05:43 tsbere: yeah - I was going to just do it, but I didn't want to risk breakage without asking first ;-) 2011-09-28T09:06:14 Dunno how fast it would be, but one option is likely "DELETE FROM money.materialized_billable_xact_summary WHERE id NOT IN (select id FROM money.billable_xact)" 2011-09-28T09:06:17 * Dyrcona breaks stuff all the time. That's what development/test servers are for. 2011-09-28T09:07:04 Dyrcona: yeah - it wouldn't be the end of the world, though these deletions are taking hours and hours - and I don't want to have to re-do anything 2011-09-28T09:08:28 I have learned so far that huge deletions tend to consume too much memory/resources, so I'm doing them in year-sized batches 2011-09-28T09:09:01 * csharp pulls the trigger to see what happens 2011-09-28T09:10:09 *** lisah_ has quit IRC 2011-09-28T09:11:37 *** akilsdonk has joined #evergreen 2011-09-28T09:16:07 *** Meliss has joined #evergreen 2011-09-28T09:19:57 hey dbs, i was going to test lp856567_reporter_html_utf8_declaration for you but realized I may not have anything to report on that would have Non-ASCII characters. Do you have a bib record or something you could export that I could test with? Or know of one I could try to pull with z39? 2011-09-28T09:23:37 csharp: once your deletes to the billing and payment tables are complete, you'll notice that money.materialized_billable_xact_summary is already cleaned up 2011-09-28T09:28:07 *** dbs has quit IRC 2011-09-28T09:28:20 Oh, hey, triggers! 2011-09-28T09:28:25 * tsbere likes triggers that do cleanup for him 2011-09-28T09:29:37 dbs: I can at least confirm the forcing of utf-8 works : https://master.evergreen.lib.in.us/reporter/1/1/1/report-data.html (admin/openils) 2011-09-28T09:29:52 *** sal_ has quit IRC 2011-09-28T09:29:58 eeevil: yeah - I saw the triggers, but didn't notice a decrease in disk space 2011-09-28T09:30:10 Oh 2011-09-28T09:30:14 That is due to how postgres works 2011-09-28T09:30:23 You need to vacuum the DB ;) 2011-09-28T09:30:30 tsbere: ah - thank you! 2011-09-28T09:30:50 internally postgres just flags the rows as deleted until vacuum comes along 2011-09-28T09:31:05 autovac can help, but if you do a lot of deletes running a manual may be a good idea 2011-09-28T09:31:13 *** sal_ has joined #evergreen 2011-09-28T09:31:25 * csharp will run a manual vacuum 2011-09-28T09:31:35 well, if the deletes are just one statement, and not finished yet, then the rows aren't gone 2011-09-28T09:31:42 That too 2011-09-28T09:31:43 thanks for pointing that out tsbere - it would have eventually sunk in 2011-09-28T09:31:55 eeevil: right - I've been running them singly 2011-09-28T09:42:25 *** jenny has joined #evergreen 2011-09-28T09:45:20 *** dbs has joined #evergreen 2011-09-28T09:48:03 *** lisah_ has joined #evergreen 2011-09-28T09:51:31 hmm - the OS is still not seeing the freed space after a VACUUM 2011-09-28T09:51:48 I'm reading about VACUUM FULL, but what I'm reading does not recommend it 2011-09-28T09:56:00 I think that is what you want, though. What isn't recommending it? 2011-09-28T09:56:26 the postgresql wiki: http://wiki.postgresql.org/wiki/VACUUM_FULL 2011-09-28T09:56:45 (using 8.4 DB currently, btw) 2011-09-28T09:58:05 of course, this is not a production DB, so maybe VACUUM FULL is what I want 2011-09-28T10:00:47 * csharp shrugs and runs VACUUM FULL 2011-09-28T10:01:36 don't know why I'm still timid about using postgres - I think eeevil put the fear of god into me in my early PINES sys admin days ;-) 2011-09-28T10:10:23 csharp: That advice on the Pg wiki is generally correct. 2011-09-28T10:11:08 csharp: you tryin to shrink your db disk size? 2011-09-28T10:11:20 csharp: That said, vacuum full is the only vacuum command option that will result in more free disk space on your server. All others will keep the allocated space so that Pg can use it without needing to allocate more space. 2011-09-28T10:11:48 csharp: In the above sense, vacuum full can make your database run more slowly. 2011-09-28T10:12:22 * Dyrcona runs autovacuum and lets the server worry about it. Also, doesn't use the Pg server for anything but Pg, so doesn't miss the disk space. 2011-09-28T10:13:21 mrpeters-isl: yes - trying to end up with a more portable, manageable (but realistic) database 2011-09-28T10:13:39 Dyrcona: thanks for the advice 2011-09-28T10:13:43 cool 2011-09-28T10:14:38 my plan is: shave down the DB to something that would comfortably fit on an 80 GB or so HDD, then pg_dump it out so it could be imported into any PG version 2011-09-28T10:15:05 csharp: sounds like you have a good reason to run vacuum full, then. you probably won't notice much of a speed hit, really. 2011-09-28T10:15:07 if anyone has advice (or warnings) that would be appreciated 2011-09-28T10:15:13 Dyrcona: good 2011-09-28T10:16:23 *** matt_carlson has joined #evergreen 2011-09-28T10:17:05 Not sure you need to vacuum before pg_dump to shrink the dump 2011-09-28T10:17:40 tsbere: hmm - I'm sure you're right 2011-09-28T10:18:09 *** agJohn has quit IRC 2011-09-28T10:18:10 I guess I'm trying to see how much space I've got so I'll know if I need to lop more data off 2011-09-28T10:21:25 *** dbs has quit IRC 2011-09-28T10:31:08 trying to fix a little bug in clark-kent.pl but getting hung up on some Perl logic. http://pastie.org/2605814 -- down where the debug file is written, I've tried to correct an incorrect reference to "$file" with the correct path "$output_dir" but it seems I am unable to use that variable. I'm not quite understanding why. 2011-09-28T10:31:10 csharp: re space, that mat-view is small, so you're not going to save much space by shrinking it. focus on bibs (and related data) if you really want to shrink the size 2011-09-28T10:31:36 Global symbol "$output_dir" requires explicit package name at /openils/bin/clark-kent.pl line 481, which I know is due to an undefined variable. 2011-09-28T10:32:02 mrpeters-isl: It doesn't live in that scope 2011-09-28T10:32:57 *** matt_carlson has quit IRC 2011-09-28T10:33:34 ah, so it only applies to the first scope, but would need to be redefined in the subs? 2011-09-28T10:34:57 Or passed into the subs. 2011-09-28T10:35:26 ok. i'll do some reading. 2011-09-28T10:35:36 Adding another parameter to the subs and doing a my $output_dir = shift; after all other = shift statements might work. 2011-09-28T10:35:42 * tsbere hasn't looked closely though 2011-09-28T10:36:36 mrpeters-isl: Actually....$file, in the context in question, *is* based on $output_dir. Specifically, "$output_dir/report-data.html" 2011-09-28T10:36:39 eeevil: thanks for the tip... when using gmcharlt's disk-space-by-table query ( http://galencharlton.com/blog/2010/05/database-server-disk-space-usage-in-evergreen/ ) the mat-view table was maybd 54G? 2011-09-28T10:36:53 it was larger than I thought it would be 2011-09-28T10:36:59 mrpeters-isl: the method you're fixing get $r passed in, right? 2011-09-28T10:37:00 csharp: whoa ... indeed 2011-09-28T10:37:22 yeah, it does get $r with a shift 2011-09-28T10:37:56 csharp: replace pg_total_relation_size with pg_relation_size in your query 2011-09-28T10:38:08 you add $r->{output_dir} = $ouput_dir at the top, where $r is setup and after $output_dir is defined. 2011-09-28T10:38:14 csharp: could just be index bloat 2011-09-28T10:38:26 then, use $r->{output_dir} in the mehtod. 2011-09-28T10:38:36 cant' ypte odayt 2011-09-28T10:38:46 :) 2011-09-28T10:39:30 * Dyrcona curses his dyslexic fingers. 2011-09-28T10:39:34 eeevil: money.materialized_billable_xact_summary | 29 GB 2011-09-28T10:39:47 so not as large as I thought, but still big 2011-09-28T10:39:57 billing table was the largest when I started 2011-09-28T10:40:44 csharp: I'd suggest REINDEX TABLE blah; to start 2011-09-28T10:40:56 eeevil: cool - will do 2011-09-28T10:43:12 Dyrcona: when you say "the top" you mean the top of the build_html sub? 2011-09-28T10:43:48 no. about line 248 . 2011-09-28T10:44:21 assuming you've not modified anything above there 2011-09-28T10:44:36 ah ok i see what you mean now 2011-09-28T10:45:29 mrpeters-isl: I'm not sure that the others would approve of this, since $r is the data of and about the report, so it's arguable if output_dir should be a member, but I don't see why not, really. 2011-09-28T10:46:48 The issue is that the output dir is, itself, a *filesystem* path. 2011-09-28T10:46:50 that seems to have really broken clark... 2011-09-28T10:47:20 is it? i was reading it as reporter/template/report/ and so on... 2011-09-28T10:47:34 https://bugs.launchpad.net/evergreen/+bug/861415 is the bug 2011-09-28T10:48:05 output_dir is built off of output_base 2011-09-28T10:48:15 output_base is defined in the config files. 2011-09-28T10:48:30 ah right /openils/var/web/repoter... 2011-09-28T10:48:48 *** bgoble has joined #evergreen 2011-09-28T10:50:30 mrpeters-isl: What I think you want is to change to 2011-09-28T10:50:32 mrpeters-isl: maybe try Back to output index (or the correct spell)? 2011-09-28T10:51:00 as the filename "report-data.html" is pretty much hardcoded anyway (even the debug link assumes it) 2011-09-28T10:51:08 so maybe using the $s2/$s3 variables to construct the href='reporter/$s2/$s3 ? 2011-09-28T10:51:40 eeevil: hmm, i wish i thought to make things that simple. i guess that would work fine 2011-09-28T10:52:18 mrpeters-isl: Everything in the build_html assumes report-data.html is the file name. Just change the one $file reference for the "back to output index" to report-data.html instead of the variable. 2011-09-28T10:52:52 Or maybe all the $file references. How many are there? <_< 2011-09-28T10:53:06 looks like 7 2011-09-28T10:53:16 on the debug page 2011-09-28T10:53:19 * tsbere isn't a fan of using javascript for that kind of thing 2011-09-28T10:53:43 If you are feeling generous, add a "back to the main page" link for the tabular HTML output too. Bugs me that there isn't a way to go back in the staff client. 2011-09-28T10:54:21 yep that worked like a charm 2011-09-28T10:54:35 and good call on the go back for tabluar 2011-09-28T10:56:07 *** lisah_ has quit IRC 2011-09-28T10:57:43 got that added too 2011-09-28T10:57:47 cool! i'll push a branch 2011-09-28T11:06:09 http://git.evergreen-ils.org/?p=working/Evergreen.git;a=commitdiff;h=0e8c022c72d098e518f7cebf4d95819ad8f9ccd8 2011-09-28T11:30:05 *** heath has joined #evergreen 2011-09-28T11:33:39 *** dbs has joined #evergreen 2011-09-28T11:35:39 I am having an issue with some packet corruption. I have been using the GPLS Helpdesk ticketing service but we seem to have come to a standstill. This was the last response I received. I am not sure what I would be capturing with Firebug as I have only very limited experience with it. The "line noise" is very likely the return packets for the request, but gzipped by Apache, which makes it non-human readable. Is there any trouble ha 2011-09-28T11:47:10 *** lisah_ has joined #evergreen 2011-09-28T11:54:11 heath: your message was truncated - it ended with "Is there any trouble ha" - FYI 2011-09-28T11:55:31 Thanks Chris. Here is the rest of the message: Is there any trouble happening within the OPAC? That may be easier for us to troubleshoot using Firefox and Firebug." 2011-09-28T11:58:49 *** kfrg56 has joined #evergreen 2011-09-28T12:02:37 Currently stuck on Evergreen Install Step 2.II. Keep receiving error of "Open-ILS/src/extras/Makefile.install: No such file or directory." Any suggestions towards finding the Makefile? 2011-09-28T12:02:55 kfrg56: what does 'pwd' tell you? 2011-09-28T12:03:04 find / -name Makefile.install 2011-09-28T12:03:29 * Dyrcona is just full of bad advice today. 2011-09-28T12:03:33 heh 2011-09-28T12:05:09 was getting "make *** No rule to make target 'Open-ILS/src/extras/Makefile.install' . Stop." 2011-09-28T12:05:31 kfrg56: what is the exact command you're typing? 2011-09-28T12:05:33 kfrg56: run pwd and tell us what it says. also ls . 2011-09-28T12:06:13 kfrg56: the second error sounds like you missed the -f 2011-09-28T12:06:31 make -f Open-ILS/src/extras/Makefile.install 2011-09-28T12:06:31 the command I've been imputting is: make -f Open-ILS/src/extras Makefile.install install_pgsql_server_debs_84 2011-09-28T12:07:00 kfrg56: you're missing a / between extras and Makefile 2011-09-28T12:08:23 *** dmoses has joined #evergreen 2011-09-28T12:08:23 I noticed that as I typed it in this chat 2011-09-28T12:08:43 typed it in the terminal correctly, still producing same error tho 2011-09-28T12:08:57 kfrg56: can you type 'pwd' and paste in the output? 2011-09-28T12:09:06 eeevil: What (when making a release) is needed for configure.ac changes? Or rather, what would be going into those two points? 2011-09-28T12:10:31 pwd output was "/root" 2011-09-28T12:11:13 kfrg56: ah - you'll need to be in /home/opensrf/Evergreen-ILS-/ for that to workd 2011-09-28T12:11:16 work 2011-09-28T12:11:49 (replace with "2.0.9" or whatever the actual version number is - no brackets) 2011-09-28T12:12:52 tsbere: the new version number ... 2.1.0, say 2011-09-28T12:14:39 *** lisah_ has quit IRC 2011-09-28T12:17:02 ok, thanks a ton 2011-09-28T12:20:31 tsbere: specifically, that replaces 'trunk' 2011-09-28T12:20:46 eeevil: Yea. Working on a slightly more generic variant. 2011-09-28T12:23:00 *** lisah_ has joined #evergreen 2011-09-28T12:24:47 heath: your issue was discussed a bit on 9/19 (see at around 14:33: " a lib with opac troubles...": http://evergreen-ils.org/irc_logs/evergreen/2011-09/%23evergreen.19-Mon-2011.log 2011-09-28T12:25:02 that might spawn new questions - hope it 2011-09-28T12:25:05 's helpful 2011-09-28T12:32:39 *** dmoses has quit IRC 2011-09-28T12:36:52 *** kfrg56 has quit IRC 2011-09-28T12:39:48 *** heath has quit IRC 2011-09-28T12:40:30 eeevil: So far I have a script that does everything (including committing to the current branch) up to "make install_all_locales" (which I am testing now) :D 2011-09-28T12:40:41 No pushing, though. I don't like automated pushes. 2011-09-28T12:41:43 *** dmoses has joined #evergreen 2011-09-28T12:43:20 ahh... ok, so, everything including using git-archive to get a clean tree, etc 2011-09-28T12:43:30 Yep. 2011-09-28T12:43:54 * tsbere intends for this to end up being "run command, it does it all, you upload a few files and push a branch" 2011-09-28T12:44:32 cool ... that would be great. allows creating a test tarball, too 2011-09-28T12:44:41 tsbere++ 2011-09-28T12:44:47 that's awesome 2011-09-28T12:44:56 did you name the script "miker.sh"? 2011-09-28T12:45:08 nope. 2011-09-28T12:45:13 * tsbere should consider doing that, though 2011-09-28T12:45:19 Currently named "test.sh". <_< 2011-09-28T12:45:25 * dbs hopes it will live in build/tools/ 2011-09-28T12:45:47 *** matt_carlson has joined #evergreen 2011-09-28T12:46:09 *** dmoses has quit IRC 2011-09-28T12:48:06 I have one major issue. Lack of obvious place to get dojo. 2011-09-28T12:49:37 Probably should be in version control somewhere, given our customizations to it. No guarantee that there won't be more customizations before we get current again 2011-09-28T12:50:34 No one is actually using the O::A::Actor::Friends stuff are they? 2011-09-28T12:53:35 * dbs is not 2011-09-28T12:53:51 *** akilsdonk has quit IRC 2011-09-28T12:54:29 I am looking at it so as to add an interface in tpac so a patron can manage friends. 2011-09-28T12:55:20 Breaking some of what it does now might make that and further expansion of it easier. 2011-09-28T12:55:38 Not really sure if breakage will be necessary, though. 2011-09-28T12:59:58 *** collum has joined #evergreen 2011-09-28T13:00:19 So, a list of "friends permissions" should go where? 2011-09-28T13:00:36 permission.friends_perms or config.friends_perms? 2011-09-28T13:00:53 Just a list of available permissions, not who has what permission for whom. 2011-09-28T13:06:58 Another thought is to move friends from buckets to its own schema. 2011-09-28T13:08:15 It would probably help to articulate what "friends" means in this context. "people to whom I've delegated some power over my account" I guess? 2011-09-28T13:08:25 dbs: exactly. 2011-09-28T13:08:50 tsbere just suggested also a list of friends with the possbility of some having no permissions. 2011-09-28T13:08:59 * dbs doesn't like "Friends" very much for that, it's misleading, but it's kinda-done 2011-09-28T13:09:14 We can call it anything we want for the patron side, as there is no interface for it yet 2011-09-28T13:09:20 delegates? surrogates? peons? 2011-09-28T13:09:44 Names matter at all layers 2011-09-28T13:09:48 I think "friends" is comfy cozy for the patrons. 2011-09-28T13:10:12 Friends makes it sound personal and nice. 2011-09-28T13:10:20 "Friends and Family" might be another way of looking at labeling for the patrons. 2011-09-28T13:10:56 Dyrcona: bucket types (with a particular type format) determine the perms 2011-09-28T13:11:08 eeevil: yep. I've been looking at that. 2011-09-28T13:11:11 ok, 2011-09-28T13:11:27 jsut re "a list of "friends permissions" should go where?" 2011-09-28T13:11:43 by which I mean, it shouldn't :) 2011-09-28T13:11:44 Friends, some of whom have proxy perms 2011-09-28T13:11:45 * tsbere is taking a lunch break from his release-building scripting, but still has to figure out the dojo chunk 2011-09-28T13:12:07 tsbere: grab tarball, untar, copy? 2011-09-28T13:12:08 :) 2011-09-28T13:12:30 eeevil: What, have it go find an older release tarball? 2011-09-28T13:12:30 eeevil: That means you have to update code in many places if you add new permissions. A user interface would have to be updated along with the backend. 2011-09-28T13:12:47 Dyrcona: no, just add bucket types 2011-09-28T13:13:12 tsbere: or point at one that we designate as "provider of the dojo" 2011-09-28T13:13:33 Or we just upload "the dojo" in a tarball somewhere. 2011-09-28T13:13:43 tsbere: sure 2011-09-28T13:13:45 eeevil: Seems to me that is overloading bucket types. 2011-09-28T13:14:01 Dyrcona: meh ... it's there and works 2011-09-28T13:14:02 eeevil: Then we can update the README to use that location instead of the official dojo site 2011-09-28T13:14:21 tsbere: no arguments from me 2011-09-28T13:14:36 eeevil: Well, no. It doesn't work yet. There's no interface for it. 2011-09-28T13:14:44 * tsbere goes to get his lunch 2011-09-28T13:15:11 *** matt_carlson has quit IRC 2011-09-28T13:15:12 Dyrcona: the code that's there works 2011-09-28T13:16:06 Dyrcona: and, regarding the UI, I'd request a sketch of what you're thinking ;) 2011-09-28T13:16:23 not a drawing sketch ... you know what I mean 2011-09-28T13:17:32 eeevil: yep. 2011-09-28T13:18:29 only because there are thoughts floating around for use cases already ... so we can avoid making those harder without slowing your momentum too much 2011-09-28T13:19:59 eeevil: I'm not opposed to the buckets. I was just thinking out loud that making a schema might prove more flexible in the long run. I'm reevaluating that thought in light of the bucket types. 2011-09-28T13:20:25 eeevil: I was basing most of my thoughts on the code and there are very few places that use the Friends code at the moment. 2011-09-28T13:20:53 Using the list of usr_bucket_types could work for granting permissions in an interface, though. 2011-09-28T13:22:07 right, that's the thought we had back in the before times ... generate the interface from that 2011-09-28T13:30:39 *** Green_ has joined #evergreen 2011-09-28T13:31:46 Hi 2011-09-28T13:31:54 agJohn-- 2011-09-28T13:32:25 * dbs lols 2011-09-28T13:33:20 Hi tsbere 2011-09-28T13:33:30 hi 2011-09-28T13:34:00 i have still strucked in that problem 2011-09-28T13:34:13 * Dyrcona like the permission to let other modify public bookbags could be useful for group book bags. 2011-09-28T13:34:39 * Dyrcona disappears in a puff of typos. 2011-09-28T13:36:16 can i acess yesterdays chat session 2011-09-28T13:36:41 Yep. I think someone left you an @later with that, see if you have a pinesol_green tab/window 2011-09-28T13:36:46 http://evergreen-ils.org/irc_logs/evergreen/2011-09/ 2011-09-28T13:47:25 all right SIP folks, I've got this question from a library filling out a form for Overdrive access: "Does the authentication response sent to Overdrive include a field for home library, and if so, what is that field name? (This is for the user activity reports so we can break it down by branch.)" 2011-09-28T13:48:02 does anyone know if home library is included? (assuming it is...) 2011-09-28T13:48:19 today is my day to hate ejabberd. same Debian Squeeze VM as always, apt-get update, apt-get install ejabberd... but today it won't start. dpkg --purge ejabberd and reinstall ejabberd and it still won't start 2011-09-28T13:48:32 * dbs loves ejabberd when it works 2011-09-28T13:49:12 probably an erl process hanging around, will --purge again and reboot 2011-09-28T13:49:27 Oh, fun, 3.6.22 is already gone. <_< 2011-09-28T13:49:56 csharp: Interesting question. I'm going to go see what exactly it is we've done. 2011-09-28T13:50:27 bshum: thanks 2011-09-28T13:50:39 We might not be paying attention to that though. 2011-09-28T13:51:03 *** lisah_ has quit IRC 2011-09-28T13:51:39 AQ 2011-09-28T13:51:45 If I were a guessing guy 2011-09-28T13:52:16 could i get the installation help ? 2011-09-28T13:52:53 csharp: Based on the same SIP response in the SIP support wiki page, AQBR1 was part of the string response in patron information. 2011-09-28T13:53:00 Green_: just ask your questions and people will help if they can 2011-09-28T13:53:04 csharp: So I assume AQ precedes the library shortname 2011-09-28T13:53:06 csharp: It is AQ in the 64, and the code comment says it was an extension requested by PINES. 2011-09-28T13:53:17 csharp: library shortname. 2011-09-28T13:53:30 bshum: Dyrcona: excellent - thanks 2011-09-28T13:55:08 *** lisah_ has joined #evergreen 2011-09-28T13:55:59 add "yaz" to the list of antiquated dependencies we have, although in this case it's Net-Z3950-SimpleServer assuming we'll have yaz 4.2+ headers installed, so it only affects SRU/Z39.50 2011-09-28T13:56:26 alternately, as csharp and others have suggested, install Net-Z3950-SimpleServer 1.12 instead of 1.14 2011-09-28T13:56:47 Yaz... *grumble* :( 2011-09-28T14:02:16 http://pastebin.ca/2084040 2011-09-28T14:03:13 Green_: you need to be running /openils/bin/srfsh as the opensrf user 2011-09-28T14:03:15 can i assume the installation of opensrf is properly done and it is working fine 2011-09-28T14:04:04 Green_: I never test OpenSRF by itself, because you have to redo half of the configuration after installing OpenILS/Evergreen. 2011-09-28T14:04:21 Green_: I just skip that part. Install Evergreen and then test. 2011-09-28T14:05:01 *** _bott_ has quit IRC 2011-09-28T14:08:07 But in the document ,if we want run evergreen ,we have to setup its Dependencies first 2011-09-28T14:08:34 and need to test them to run properly 2011-09-28T14:08:51 Green_: When you install Evergreen opensrf_core.xml and/or opensrf.xml are replaced, so you have to reconfigure OpenSRF anyway. 2011-09-28T14:09:04 ok 2011-09-28T14:09:30 Green_: but if you've already configured the files, you should be able to run srfsh as the opensrf user 2011-09-28T14:09:48 Green_: your error was happening because you're trying to run it as root 2011-09-28T14:09:53 no i have not installed evergreen 2011-09-28T14:09:56 csharp is right, if you've put the .srfsh.xml in the proper place. 2011-09-28T14:10:17 i just installed opensrf 2011-09-28T14:10:23 it would work if you copied it to ~root. 2011-09-28T14:10:30 ok 2011-09-28T14:10:44 Green_: please do this: 'su - opensrf' then 'srfsh' 2011-09-28T14:10:52 ok 2011-09-28T14:13:59 http://pastebin.ca/2084047 2011-09-28T14:14:56 nothing happens 2011-09-28T14:15:43 what should i do 2011-09-28T14:16:58 Green_: did you edit (as the opensrf user) /home/opensrf/.srfsh.xml and replace "password" with your actual password? 2011-09-28T14:17:26 no 2011-09-28T14:17:42 Green_: okay - you'll have to do that before the test will work 2011-09-28T14:17:58 shall i need to replace that password with opensrf user password 2011-09-28T14:19:25 Green_: no - with the jabber opensrf user for private.localhost 2011-09-28T14:19:31 *** matt_carlson has joined #evergreen 2011-09-28T14:20:00 Green_: you set those up in step 8. I. 2011-09-28T14:20:49 ok 2011-09-28T14:21:34 But in /home/opensrf there is no .srfsh.xml file to edit 2011-09-28T14:22:23 Green_: ah - well 'cp /openils/conf/srfsh.xml.example /home/opensrf/.srfsh.xml' (as the opensrf user) 2011-09-28T14:22:52 http://pastebin.ca/2084053 2011-09-28T14:23:57 Green_: do the 'cp' command I just suggested, then 'pico /home/opensrf/.srfsh.xml' (spelling really matters ;-) ) 2011-09-28T14:25:34 Anybody have suggestions about ways to export just a single libraries' bibs and items with the holdings expressed as a tag during export? In the past, we've used the marc_export tool for this, along a given set of bib ID's, but it seems that it does not distinguish which library we'd prefer to have, but appends all items from everywhere on it. 2011-09-28T14:26:10 I was looking at the newer marc_export in 2.1+ and it has a --library option, but that doesn't seem to consistently grab the library who's shortcode we use either. 2011-09-28T14:26:23 Maybe a problem with trying to use a 2.1 script with our 2.0 system? 2011-09-28T14:26:45 but in the Documentation,it says copy as a root user 2011-09-28T14:28:46 whic user rights needed to do this 2011-09-28T14:29:02 as opesrf user or root user 2011-09-28T14:29:06 Green_: opensrf 2011-09-28T14:29:11 ok 2011-09-28T14:29:13 Yuck, it does say that in the official 2.0 draft 2011-09-28T14:29:26 http://docs.evergreen-ils.org/2.0/draft/html/serversideinstallation.html 2011-09-28T14:29:39 bshum: I'm on the wiki, where it doesn't say that 2011-09-28T14:29:47 csharp: Right me too, usually. 2011-09-28T14:31:38 Green_: For your benefit, http://www.evergreen-ils.org/dokuwiki/doku.php?id=opensrf:2.0:install is a link to the instructions csharp was referring to. 2011-09-28T14:36:12 Anyone with evergreen-ils.org access care to drop a copy of this somewhere there (and then link me to it)? http://tomtrunk.mvlcstaff.org/dojo.tgz 2011-09-28T14:36:31 tsbere: I can help you with that. 2011-09-28T14:37:05 Hmm, any suggested path for that? 2011-09-28T14:37:26 Or can I just drop it in downloads or something 2011-09-28T14:37:31 downloads likely works 2011-09-28T14:38:29 http://evergreen-ils.org/downloads/dojo.tgz 2011-09-28T14:38:42 bshum: Thank you 2011-09-28T14:40:00 *** lisah_ has quit IRC 2011-09-28T14:42:28 *** matt_carlson has quit IRC 2011-09-28T14:44:38 *** bjwebb has joined #evergreen 2011-09-28T14:48:52 *** Green_ has quit IRC 2011-09-28T14:52:48 *** lisah_ has joined #evergreen 2011-09-28T14:53:13 So, I have no clue how to test the i18n piece, but so far it looks like my script works 2011-09-28T14:53:14 :D 2011-09-28T14:53:43 * tsbere even made sure it works on 2.1 2011-09-28T14:54:16 Liar, liar, pantalones del fuego! 2011-09-28T14:56:42 *** akilsdonk has joined #evergreen 2011-09-28T14:57:33 *** bgoble has quit IRC 2011-09-28T14:57:37 He didn't just make sure it works on 2.1.... He developed it on 2.1. 2011-09-28T14:57:51 What can I say 2011-09-28T14:57:58 master doesn't have a version number to auto-find 2011-09-28T15:04:51 tsbere: warezitag, sir? 2011-09-28T15:04:55 er 2011-09-28T15:04:59 warezitat 2011-09-28T15:05:43 eeevil: http://git.evergreen-ils.org/?p=working/Evergreen.git;a=shortlog;h=refs/heads/user/tsbere/make_release 2011-09-28T15:05:53 * tsbere just finished doing some basic docs 2011-09-28T15:05:59 very basic 2011-09-28T15:06:01 <_< 2011-09-28T15:06:25 * tsbere left an errant #TODO that he finished, though, now that he looks 2011-09-28T15:06:44 eeevil: I will force-push a better commit message (usage examples and such) shortly. But you appear to have wanted to start playing with it ;) 2011-09-28T15:07:08 :) 2011-09-28T15:12:52 solid work, sir! 2011-09-28T15:14:52 eeevil: Force pushed the new commit message (and removal of finished #TODO) 2011-09-28T15:15:48 * tsbere figures the more the script does for us the less we can forget to do ;) 2011-09-28T15:21:04 eeevil: For the record, I based this on previous "just tag all the version points" work. Had to dig out an old forgotten about branch. <_< 2011-09-28T15:22:53 bshum and others that build staff clients during release time: I apoligize, but the script there does that step now. >_> 2011-09-28T15:23:35 Sounds good to me. 2011-09-28T15:25:53 I guess that means more people will have time to run other tests ;) 2011-09-28T15:27:24 :) 2011-09-28T15:28:05 *** lisah_ has quit IRC 2011-09-28T15:31:24 Maybe I can spend some tiny bit of that time on learning tpac skinning. 2011-09-28T15:31:50 We're not sure how we feel about having to skin up 80 different ones for our libs. Or if we can consolidate where we can, etc. 2011-09-28T15:32:21 Future projects for rainy days. 2011-09-28T15:32:52 I just adapted our JSPac code to the TPac. Which meant "telling TPac to look for what is filling the JSPac fields anyway" 2011-09-28T15:33:10 *** heath has joined #evergreen 2011-09-28T15:33:32 Hmm, neat! 2011-09-28T15:33:58 * tsbere emailed one of the lists about the JSPac variant at some point 2011-09-28T15:38:23 bshum: If you want I should be able to pull some of the code changes out to a branch or something 2011-09-28T15:39:03 tsbere: I'm certainly be curious but I probably don't have enough time to delve enough into it right now. Don't worry about it for now. 2011-09-28T15:39:35 Thanks for the offer though. Or you can always share it and can ponder on it later. 2011-09-28T15:39:35 bshum: I keep everything in git *anyway*. Can't be that hard for me to pull out a couple of changes. ;) 2011-09-28T15:39:49 Well in that case :) 2011-09-28T15:46:17 *** heath has quit IRC 2011-09-28T15:53:20 *** _bott_ has joined #evergreen 2011-09-28T15:55:14 * csharp is starting to see the benefits of running from git vs. tarballs, at least on test servers 2011-09-28T15:55:41 csharp: depends on how often you want to update. 2011-09-28T15:55:56 yeah - this server is running head 2011-09-28T15:56:33 i'd say if you want to update more often than tarballs are released, then you should run from git. 2011-09-28T15:56:45 well with autoupdate functionality now possible for the clients, that opens up a whole new universe of possibilities 2011-09-28T15:59:44 *** Meliss has quit IRC 2011-09-28T16:05:59 Dyrcona: I still think setting opensrf up and running is a good test of the basics (ejabberd, srfsh, dependencies) before adding more complexity to the mix. I know you never test opensrf anymore, but you're not the typical person looking for help getting evergreen installed. 2011-09-28T16:06:30 of course, part of me wants to just say "New to Linux? Evergreen is not for you." 2011-09-28T16:07:04 * Dyrcona shall henceforth ignore requests for installation help. 2011-09-28T16:07:07 heh 2011-09-28T16:07:19 It might save sanity. 2011-09-28T16:08:02 I remarked to tsbere earlier that Evergreen is not newb friendly, particularly to those with little or no UNIX/Linux background. 2011-09-28T16:08:18 bshum: http://git.evergreen-ils.org/?p=working/Evergreen.git;a=shortlog;h=refs/heads/user/tsbere/apache_lib_urls 2011-09-28T16:08:23 *** sal_ has quit IRC 2011-09-28T16:08:24 the packages will help with that 2011-09-28T16:08:38 i heartily recommend finishing those 2011-09-28T16:08:50 my default response in #koha is use the packages 2011-09-28T16:09:27 * Dyrcona recommends koha to those who do not really need Evergreen. 2011-09-28T16:09:31 and if apt-get install koha-common is beyond you, then yeah you probably shouldnt be touching a server 2011-09-28T16:09:36 Dyrcona: yep, did you look at slide #2 of my most recent presentation 2011-09-28T16:09:42 rangi++ 2011-09-28T16:09:45 i do the same for evergreen drycona :) 2011-09-28T16:09:48 I believe that I did. 2011-09-28T16:10:00 So we're in agreement 2011-09-28T16:16:13 It's user friendly. It's just very picky about who its friends are. 2011-09-28T16:16:35 *** collum has quit IRC 2011-09-28T16:16:44 tsbere: Oh, sweet. Thanks, I appreciate you sharing with me! 2011-09-28T16:20:21 *** dbs has quit IRC 2011-09-28T16:28:37 eeevil: Any comments/questions on the make release script there? 2011-09-28T16:35:49 *** akilsdonk has quit IRC 2011-09-28T16:40:25 *** kmlussier has quit IRC 2011-09-28T16:41:40 *** dbs has joined #evergreen 2011-09-28T16:44:28 *** akilsdonk has joined #evergreen 2011-09-28T16:51:35 *** youdonotexist has joined #evergreen 2011-09-28T16:52:49 *** Green_ has joined #evergreen 2011-09-28T16:53:11 Hi 2011-09-28T16:59:59 *** Shae has quit IRC 2011-09-28T17:02:25 *** akilsdonk has quit IRC 2011-09-28T17:03:42 *** matt_carlson has joined #evergreen 2011-09-28T17:03:59 *** AaronZ-PLS has quit IRC 2011-09-28T17:14:10 *** youdonotexist has quit IRC 2011-09-28T17:14:28 *** youdonotexist has joined #evergreen 2011-09-28T17:21:12 *** matt_carlson has quit IRC 2011-09-28T17:33:01 *** Dyrcona has quit IRC 2011-09-28T17:38:23 *** matt_carlson has joined #evergreen 2011-09-28T17:43:11 *** matt_carlson has quit IRC 2011-09-28T17:43:52 *** Green_ has quit IRC 2011-09-28T17:46:26 *** dbs has quit IRC 2011-09-28T17:52:13 *** jenny has left #evergreen 2011-09-28T18:06:22 *** agJohn has joined #evergreen 2011-09-28T18:06:35 *** Green_ has joined #evergreen 2011-09-28T18:07:14 So, anyone else thinking we should brand the TTOpac as "Evergreen Silk"? ;) 2011-09-28T18:07:50 Hello any body could help 2011-09-28T18:08:04 i am getting this No config error 2011-09-28T18:08:18 the pastebin link is 2011-09-28T18:08:22 http://pastebin.ca/2084255 2011-09-28T18:08:29 Hey, Green_. I'm checking on it now. 2011-09-28T18:08:57 Hi agJohn 2011-09-28T18:09:18 fine to see you in chat 2011-09-28T18:10:30 Hello! OK. The .srfsh.xml file has to be in the home directory of any user you want to be able to run srfsh 2011-09-28T18:10:32 First thing, log back in as opensrf and then give it a try. Is the password in /home/openrf/.srfsh.xml set correctly? 2011-09-28T18:21:53 GNU nano 2.2.2 File: /home/opensrf/.srfsh.xml