2011-02-18T06:56:00 *** sfortin has joined #evergreen 2011-02-18T07:33:12 *** rickd_ has joined #evergreen 2011-02-18T07:48:30 *** kmlussier has joined #evergreen 2011-02-18T08:15:52 *** _bott_ has quit IRC 2011-02-18T08:17:58 *** Dyrcona has joined #evergreen 2011-02-18T08:18:22 *** _bott_ has joined #evergreen 2011-02-18T08:36:28 *** kmlussier has quit IRC 2011-02-18T08:37:19 *** kmlussier has joined #evergreen 2011-02-18T08:58:28 *** Meliss has joined #evergreen 2011-02-18T09:12:31 csharp: did you get an answer on A/T yesterday? 2011-02-18T09:13:44 I'm **pretty sure** a definition from 2.0.1 could be moved back to 1.6 2011-02-18T09:14:23 mrpeters-isl: no - I didn't get an answer - thanks mrpeters-isl 2011-02-18T09:14:37 the only thing i can think of - make sure the reactors exist in 1.6 2011-02-18T09:14:44 there are a few in 2.0 that ARE NOT in 1.6 2011-02-18T09:14:44 * csharp will experiment when he gets the time 2011-02-18T09:15:03 the oens about JEDI, asterisk, and then the "SendFile" 2011-02-18T09:15:36 *** dneimeier has quit IRC 2011-02-18T09:15:58 there are also a few new validators for Acquisitions, but as long as your A/T event doesn't rely on those I'd imagine you could backport without issue 2011-02-18T09:18:20 mrpeters-isl: good info - I'm sure I'll be asking for debugging help when I get to those ;-) 2011-02-18T09:20:24 are you guys using the 30 day one now? 2011-02-18T09:20:31 i've been using it about 2 weeks now - people LOVE it 2011-02-18T09:20:54 mrpeters-isl: no - we're still on 1.4 - I've been doing on this on a test box 2011-02-18T09:21:00 *** yboston has joined #evergreen 2011-02-18T09:21:04 I bet they do 2011-02-18T09:21:10 ah ok! 2011-02-18T09:21:29 I have a feeling this is going to be one of the favorite features of 1.6 for the PINES libraries 2011-02-18T09:22:29 *** rickd_ has quit IRC 2011-02-18T09:23:05 yeah action triggers are very good stuff - just so much more to learn 2011-02-18T09:23:19 so if you have any good ideas - let me know again - i'd like to try writing some more1 2011-02-18T09:23:42 do you guys do many patron merges "across systems" 2011-02-18T09:24:05 for example - we end up with duplicate patrons in counties that have many evergreen libraries - do you run into that? 2011-02-18T09:24:41 or does anyone - for that matter 2011-02-18T09:25:29 i have this grand idea for a web app where someone submits two barcodes for merge - a designated staff member from each home ou associated with the barcodes is notified that a merge has been requested 2011-02-18T09:26:13 they both login with a localadmin account, for example, and once the merge is "approved" by both parties it goes into a database which we can download weekly and then process those merges with the patron merge SQL scripts 2011-02-18T09:30:52 I bet you could use A/T to notify a home ou whenever a specific type of standing penalty/message is added to a patron's account (such messages could be gathered via reporting) 2011-02-18T09:31:52 mrpeters-isl: Care to take a look at https://bugs.launchpad.net/evergreen/+bug/720795 and tell me what you think? 2011-02-18T09:39:18 *** dbs_n1 has joined #evergreen 2011-02-18T09:41:50 tsbere: yes - will do 2011-02-18T09:42:35 this is awesome! 2011-02-18T09:43:03 i may not be able to apply it right away but it looks great from the description! 2011-02-18T09:43:20 *** dbs_n1 has quit IRC 2011-02-18T09:51:03 *** rickd_ has joined #evergreen 2011-02-18T09:54:05 *** afterl has joined #evergreen 2011-02-18T10:07:36 *** bshum has joined #evergreen 2011-02-18T10:10:34 *** bshum has quit IRC 2011-02-18T10:23:06 *** bshum has joined #evergreen 2011-02-18T10:29:32 re A/T backporting, there are features (granularity primary among them) that are really needed by some hooks and defs that are not available in 1.6 ... I'd be weary of trying whole-sale backports 2011-02-18T10:32:52 eeevil: would some nodoze help your weariness? 2011-02-18T10:32:52 ;) 2011-02-18T10:33:47 gmcharlt: stupid fingers ... wary 2011-02-18T10:34:59 hi all - I'm getting this error after a fresh install of Opensrf-1.6.2/EG1.6.0.7 on squeeze: http://paste.lisp.org/display/119807 2011-02-18T10:35:33 I see some older list emails about it and eeevil responded that it looked like misconfigured memcache 2011-02-18T10:36:18 csharp: it indeed still looks that way :) 2011-02-18T10:36:31 csharp: exactly 2011-02-18T10:36:48 eeevil: how would I go about checking? I haven't changed the cache entry in the opensrf config files 2011-02-18T10:37:19 csharp: check to see if the memcached process is running, then try telnetting to the memcached port 2011-02-18T10:37:31 csharp: check the IPs memcache is listening on, etc 2011-02-18T10:43:16 Are there any docs written about how the BibTemplate operates in 2.0? We attempted to apply our BibTemplate modifications from 1.6 and they don't work and the example does seem to look a bit different in style and structure. 2011-02-18T10:43:27 *1.6.1.x 2011-02-18T10:45:11 bshum: did you watch dbs 's presentation? http://www.google.com/url?sa=t&source=web&cd=2&ved=0CBwQtwIwAQ&url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3D5gY8uYJVYco&ei=8JNeTdWZFYep8AaNgq2TDA&usg=AFQjCNH6S7RMyXt-VY5O0XfPlhNcsOcx0w&sig2=bfw2iG7IvPKuoBh5nYQMaA 2011-02-18T10:45:16 sorry 2011-02-18T10:45:27 csharp: I was there for that session, sure :) 2011-02-18T10:45:28 meant that to be a youtube link 2011-02-18T10:45:32 oh ok 2011-02-18T10:45:33 :-) 2011-02-18T10:45:49 csharp: We're more or less following the steps outlined in that presentation / docs 2011-02-18T10:46:04 But it doesn't work with our 2.0 OPAC 2011-02-18T10:46:15 I'm assuming that means something changed in between versions. 2011-02-18T11:05:25 * csharp decides that the memcache thing might be a red herring based on a lenny -> squeeze update 2011-02-18T11:05:41 cuz I was too lazy to reinstall 2011-02-18T11:06:02 I'll start with a fresh image and try again 2011-02-18T11:06:07 thanks eeevil & gmcharlt 2011-02-18T11:10:07 csharp; i've done the same distro upgrade and didn't have memcache issues 2011-02-18T11:10:21 i'd definetly try to telnet, like gmcharlt said 2011-02-18T11:10:40 mrpeters-isl: this may just be a borked VM - I have so many that I forget which ones are which 2011-02-18T11:10:46 you could always apt-get remove memcache --purge to start fresh 2011-02-18T11:10:47 * csharp needs VM version control 2011-02-18T11:11:00 heh 2011-02-18T11:11:16 mrpeters-isl: I did telnet localhost 11211, but I was lost to what to try next 2011-02-18T11:11:28 ah - what did it tell you? 2011-02-18T11:11:35 if you can establish a connection, it should be working 2011-02-18T11:11:45 it appeared to connect 2011-02-18T11:12:04 i vaguely remember somethign about squeeze not liking localhost in the config file or something like that 2011-02-18T11:12:19 and you had to change to 127.0.0.1 in memcached.conf 2011-02-18T11:12:29 could be totally unrelated 2011-02-18T11:12:34 last time I used telnet to any degree was the early 90s and that was for checking UGA campus email 2011-02-18T11:12:56 hmm 2011-02-18T11:13:06 if I run into it again, I'll report back 2011-02-18T11:13:31 if you did telnet localhost 11211 and it gave connection established that should confirm that it's at least listening on that port 2011-02-18T11:13:57 yeah - it did seem to work, connection-wise 2011-02-18T11:14:07 ok - carry on :) 2011-02-18T11:14:11 heh 2011-02-18T11:15:52 csharp: http://www.mail-archive.com/open-ils-general@list.georgialibraries.org/msg02984.html 2011-02-18T11:16:06 see dbs's post 2011-02-18T11:16:19 i linked the one below it -- sorry! 2011-02-18T11:16:43 exact same issue you had, too :) 2011-02-18T11:17:00 huh - thanks - I'll keep that in mind when it doesn't work again ;-) 2011-02-18T11:19:37 haha you even responded in that thread :P 2011-02-18T11:24:20 bshum: regarding BibTemplate in 2.0, the main incompatibility I have seen is that dojox.data.dom.textContent() is no longer supported, and should generally be BT.textContent() instead. 2011-02-18T11:24:53 dbwells: Oooh, that sounds promising. We'll give that a try! 2011-02-18T11:28:11 dbwells++ That works for us! 2011-02-18T11:30:09 cool 2011-02-18T12:08:04 *** yboston has quit IRC 2011-02-18T12:09:50 *** yboston has joined #evergreen 2011-02-18T12:40:03 *** parsr has joined #evergreen 2011-02-18T12:41:24 kmlussier: that URL you shared has a typo; and when I tried just .org as a fix, got an access denied on the page 2011-02-18T12:41:52 * tsbere got in fine.....but he has a username/password 2011-02-18T12:42:20 and I apparently was already logged in 2011-02-18T12:42:45 So the fixed link is good, the perms on it are bad? 2011-02-18T12:44:05 http://masslnc.cwmars.eorg/node/2273 => bad, http://masslnc.cwmars.org/node/2273 => good, but access denied at the web app level 2011-02-18T12:53:22 *** testing has joined #evergreen 2011-02-18T12:55:33 *** testing has quit IRC 2011-02-18T13:06:55 phasefx: thx! I knew I was forgetting something before sending that out. 2011-02-18T13:14:45 Permissions should be all fixed now. 2011-02-18T13:15:02 confirmed. kmlussier++ 2011-02-18T13:15:11 kmlussier: Cool ideas! SMS would be fun 2011-02-18T13:16:28 Upgrade from 1.6.1 to 2.01 -- 2.0.0-2.0.1-upgrade-db.sql results in this error: 337: ERROR: relation "serial.unit" does not exist ROLLBACK -- Should I ignore or does this point to something else? 2011-02-18T13:17:17 To me that sounds like something goofed during 1.6.1-2.0-upgrade script? 2011-02-18T13:17:25 Did that go through without problems? 2011-02-18T13:18:14 *** dbs has joined #evergreen 2011-02-18T13:18:14 *** dbs has joined #evergreen 2011-02-18T13:18:28 bshum: SMS is a little scaled back from the first writing, but it's a first step. 2011-02-18T13:18:39 dbwells / bshum: default skin in 2.0 still uses dojox.xml.parser.textContent 2011-02-18T13:19:02 trivial to switch to BT.textContent of course 2011-02-18T13:20:03 dbs: Hmm, maybe we'll poke at that some more after our move to 2.0. 2011-02-18T13:20:07 dbs: Thanks for letting us know though! 2011-02-18T13:20:33 bshum -- you are correct -- browsing up the screen I can see: ERROR: insert or update on table "series_field_entry" violates foreign key constraint "metabib_series_field_entry_source_pkey" 2011-02-18T13:20:35 DETAIL: Key (source)=(4) is not present in table "record_entry". Followed by a bunch of : current transaction is aborted, commands ignored until end of transaction block" 2011-02-18T13:23:04 * dbs goes to make the BT.textContent() fix 2011-02-18T13:24:07 parsr: SELECT * FROM biblio.record_entry WHERE id=4 gives what? 2011-02-18T13:24:27 sunds to mlike you have a metabib.series_field_entry row that refers to a source in biblio.record_entry that no longer exists 2011-02-18T13:24:33 eek - spellilng! 2011-02-18T13:25:56 mrpeters-isl: 0 rows retrieved 2011-02-18T13:27:32 mrpeters-isl: using the IP for localhost rather than "localhost" does indeed fix the memcache issue on EG1.6.0.7/squeeze 2011-02-18T13:27:53 s/1.6.1.7/1.6.0.7/ 2011-02-18T13:28:14 parsr: fascinating. metabib.series_field_entry.source has no foreign key to biblio.record_entry.id - that seems surprising 2011-02-18T13:28:41 dbs: upgrading from 1.6.1.7 on dev 2011-02-18T13:28:44 csharp: weird. I saw that for a while during squeeze prereleases, but then a subsequent update fixed that problem 2011-02-18T13:28:58 parsr: no, i'm saying that that is the case for everyone 2011-02-18T13:30:01 whereas metabib.title_field_entry.source does have a foreign key pointing at biblio.record_entry.id 2011-02-18T13:35:49 dbs: copy buffer pasted to this file (temporarily availble) here: http://geoscan.ess.nrcan.gc.ca/evergreen/upgrade-error.txt 2011-02-18T13:38:33 parsr: yeah, that error happens when the foreign key that's missing on metabib.series_field_entry is added to the table 2011-02-18T13:39:19 unfortunately, it doesn't do something like "DELETE FROM metabib.series_field_entry WHERE source NOT IN (SELECT id FROM biblio.record_entry);" first 2011-02-18T13:39:36 ergo - you get an error when the constraint can't be satisfied 2011-02-18T13:40:35 eeevil had earlier stated that the 1.6.1-2.0 upgrade script was locked down and shouldn't be changed now, but I think this offers a good reason for modifying the upgrade script 2011-02-18T13:42:39 dbs: this was a fresh net install of squeeze in a VirtualBox VM 2011-02-18T13:43:52 csharp: yeah, that's why I say it's weird that the same problem reappeared. the install instructions used to have a warning to that effect, but when the problem went away, I pulled the warning out 2011-02-18T13:46:05 parsr: would you be willing to post a bug in bugs.launchpad.net/evergreen ? 2011-02-18T13:46:54 dbs: thanks. Yes -- will post bug. Other than this problem, the upgrade to 2.0 looks pretty straight forward. I was also curious to time the reingest if/ when I get there. Spring like day here in Ottawa -- the prisoner is allowed for daily walk of the perimeter. Enjoy the weekend all. 2011-02-18T13:47:31 *** parsr has quit IRC 2011-02-18T14:18:08 so granitize - wanna set up that testing VM some time? 2011-02-18T14:22:29 is it possible to turn off the "place at top of queue" for holds in 2.0.1? 2011-02-18T14:22:35 we're having a freakout 2011-02-18T14:22:40 dbs: sureā€¦ What OS do you want? 2011-02-18T14:23:00 granitize: heh, how about you decide what OS _you_ want? :) 2011-02-18T14:23:21 mrpeters-isl: That doesn't sound good :S 2011-02-18T14:23:49 Doesn't matter. Ubuntu, Lenny, squeeze, fedora Usingn them all 2011-02-18T14:25:25 granitize: heh, all right! Let's go the Squeeze route, then 2011-02-18T14:25:33 Squeeze 2011-02-18T14:25:36 it is 2011-02-18T14:26:18 RAM and HDD? 2011-02-18T14:30:00 512 RAM should be fine, 16 GB of HDD should be plenty 2011-02-18T14:30:23 You could probably cut the HDD down, but that's what I put in my call for server email so I'll stick with that as the official line :) 2011-02-18T14:34:31 mrpeters-isl: just off the top of my head, you could define a trigger that would prevent the cut in line boolean from being set on holds? should be a better way, though. depends on why you want to turn it off, i suppose 2011-02-18T14:35:44 jeff: I think it is more of a "they are freaking out that it is even an option" thing 2011-02-18T14:37:08 tsbere: I think people here felt that way about suppress holds/transits checkin options too. 2011-02-18T14:37:22 bshum: That at least has alternate usages 2011-02-18T14:37:36 It would be nice to remove certain pieces from the interfaces without digging too deep. 2011-02-18T14:37:45 tsbere: Indeed 2011-02-18T14:37:48 For example, running through a list of things to be checked in that aren't actually in front of you for some reason. 2011-02-18T14:37:50 One configuration option for every menu item! 2011-02-18T14:38:55 Or permission-aware menu items, that only appear if you have the permission to do X. Or both, for combinatorial fun 2011-02-18T14:39:38 It's nice to dream. 2011-02-18T14:40:05 local css to hide stuff 2011-02-18T14:42:20 so if I were setting up a script to add a whole bunch of actor.org_units and a whole bunch of actor.org_addresses, which would need to be entered first? seems that each has a constraint on the other 2011-02-18T14:43:42 csharp: org_unit first 2011-02-18T14:44:03 then add the addresses, then finish by updating the aou.*address columns 2011-02-18T14:44:34 gmcharlt: ah - that makes sense 2011-02-18T14:44:36 thanks 2011-02-18T14:47:30 tsbere is right 2011-02-18T14:47:31 csharp: if you do it all inside a single transaction, the order doesn't matter. 2011-02-18T14:47:44 people are freaking that staff could suddenly place each other at the top of popular hold queues 2011-02-18T14:48:00 staff could do that before, by cancelling everyone else's holds 2011-02-18T14:48:18 thats a little more noticable, dbs 2011-02-18T14:48:43 Cut in line has decent use cases, though. 2011-02-18T14:48:45 If people are going to misbehave, they'll misbehave. Should be easy enough to flag when cutting in line happens, if you're concerned 2011-02-18T14:48:51 tsbere: it totally does 2011-02-18T14:48:59 Flag it and look for abnormal patterns 2011-02-18T14:49:17 It would be easy enough to run a nightly report to see how many cut in lines there are, I think. 2011-02-18T14:49:30 I want True Blood Season 3 first, dang it! 2011-02-18T14:49:45 yeah, i guess i just would expect a feature like that to be optional 2011-02-18T14:50:04 The backend either supports it or it doesn't. 2011-02-18T14:50:26 theres so much talk about "flexibility" but this doesn't really offer any flexibility as you're stuck with it once your on 2.0 2011-02-18T14:50:30 * tsbere isn't even sure where the gui element lives 2011-02-18T14:50:41 tsbere: patron account, holds tab 2011-02-18T14:52:14 * tsbere wonders if it would be hard to add a permission for if you are allowed to set that 2011-02-18T14:53:43 tsbere: yeah, i'd think having only LocalAdmin able to do this or something would appease them a bit 2011-02-18T14:54:05 i'd sure hope whoever holds the localadmin permissions is trustworthy 2011-02-18T14:54:14 local css will appease your staff much more easily, as phasefx mentioned 14 minutes back 2011-02-18T14:54:16 Take note, me wondering doesn't mean *I* will get around to looking at it anytime soon, if ever. 2011-02-18T14:54:43 senator: forgive me - i havent caught up on backscroll 2011-02-18T14:55:17 won't stop someone with a hacked/custom client, but I'm a big fan of reports for finding abuse, rather than trying too hard to prevent it 2011-02-18T14:55:33 yeah - if someone goes that far then there is a problem 2011-02-18T14:56:06 but if someone just does it because they were "allowed by the software" thats more of what i'd like to prevent 2011-02-18T14:56:25 suppose an Evergreen savvy patron could place their own hold via api with cut in line enabled :) 2011-02-18T14:56:50 phasefx: so i'm not sure where these CSS live - can you point me to them? are they server side or client side? 2011-02-18T14:56:56 I expect most places, when rolling out an upgrade, would say "Here is new feature #1 - and here is our policy concerning the use of this feature". 2011-02-18T14:57:13 dbs: of course we do, but there are still going to be people who say "oh, i didn't know" 2011-02-18T14:57:25 i don't see why i'm catching so much greif for trying to prevent that situation 2011-02-18T15:00:59 mrpeters-isl: server side, most of the css files in server/skin have an @import for a _custom.css variant of the file 2011-02-18T15:01:16 ah ok so that's where the "hacked" client side could come into play 2011-02-18T15:01:32 what you're trying to style away may or may be easy to select via CSS 2011-02-18T15:01:42 may not, even 2011-02-18T15:02:04 if it has an id attribute, easy-peasy 2011-02-18T15:02:06 yeah, its under actions for selected holds > set top of queue 2011-02-18T15:02:19 in the dropdown 2011-02-18T15:03:58 Looks to be in right-click menu too? 2011-02-18T15:04:07 tsbere: yep, it would be 2011-02-18T15:04:09 No id, but it is in a server-side file 2011-02-18T15:04:16 So you can just remove the two lines from the right file 2011-02-18T15:04:21 so in this case, that entry is defined in server/patron/holds_overlay.xul. There are two elements, which you could select with css with something like [command="cmd_holds_cut_in_line] { display: none; } 2011-02-18T15:05:10 benefit of using CSS rather than hacking the files, is that your _custom.css files can be easier to manage with upgrades 2011-02-18T15:05:15 right - i see that 2011-02-18T15:05:34 phasefx: where in the xul is the location of the css defined? 2011-02-18T15:05:40 i'm not seeing it 2011-02-18T15:06:18 mrpeters-isl: the main entry point here is server/patron/holds.xul, which pulls in the CSS, as well as the holds_overlay.xul file 2011-02-18T15:06:29 ahh right 2011-02-18T15:06:39 so circ_custom.css would automatically be read? 2011-02-18T15:06:41 so holds.xul is refeencing global.css and circ.css 2011-02-18T15:06:48 correct 2011-02-18T15:06:59 should it be a full copy, or just the "additonal" definitions 2011-02-18T15:07:06 just extra 2011-02-18T15:07:09 I believe 2011-02-18T15:07:22 one important thing to note about these _custom.css files, is that they get parsed before the normal CSS, so unless you use ! important, they can be cascaded over if you're not careful 2011-02-18T15:08:31 * tsbere wonders why the custom wouldn't be defined as the last rule in the file, rather than the first 2011-02-18T15:09:17 Oh, right, some css parsers hate import other than at the front 2011-02-18T15:09:30 including Mozilla's 2011-02-18T15:09:47 Which brings me to "why not use an alternate way of loading them" <_< 2011-02-18T15:10:05 too lazy, let tsbere solve it :) 2011-02-18T15:11:29 probably would be safe to reference _custom.css files directly in xul/html, but this was the least amount of work for the biggest bang at the time 2011-02-18T15:11:47 @import, that is 2011-02-18T15:11:47 phasefx: Error: "import," is not a valid command. 2011-02-18T15:11:59 pinesol: shush it :) 2011-02-18T15:11:59 phasefx: Error: "shush" is not a valid command. 2011-02-18T15:12:43 thanks for the great lesson, phasefx 2011-02-18T15:12:54 this teaches me a few other things i can do too so i appreciate it 2011-02-18T15:13:13 i DO think it's worth considering the risks of having **any** staff member be able to put someone at the top of the queue 2011-02-18T15:13:37 for example, say you have a volunteer who you give minimal circ permissions - they may not know this is "wrong" to do and decide to do it 2011-02-18T15:13:41 *** pmplett_ has joined #evergreen 2011-02-18T15:13:57 don't give them perms to modify holds 2011-02-18T15:14:00 i think tsbere makes a good point, that there should be a minimum level of permissions for this and similar features 2011-02-18T15:14:12 phasefx: but it'd be ok for them to place a hold for someone 2011-02-18T15:14:19 just not give them priority 2011-02-18T15:14:21 should be a different perm 2011-02-18T15:14:49 i just wouldn't want to take away any permissions except the ability to "move someone up" 2011-02-18T15:15:17 phasefx: I can spit out a patch in a few minutes to move it into the xul and out of the css. 2011-02-18T15:17:09 there's a specific perm for modifying hold request time; dont' see why there shouldn't be one for cut_in_line/top of queue 2011-02-18T15:17:30 actually, I don't see why they shouldn't be the same 2011-02-18T15:17:33 tsbere: you're going to do that for all xul? 2011-02-18T15:17:40 the effect is the same in practice 2011-02-18T15:17:57 re the modify-request-time perm 2011-02-18T15:17:57 phasefx: It wouldn't be hard to do it across everything that references a _custom.css file 2011-02-18T15:18:56 okay; I'm not overly horrified, but only if you really want to :) 2011-02-18T15:20:49 *** bencahill has quit IRC 2011-02-18T15:21:16 *** bencahill_ has joined #evergreen 2011-02-18T15:22:58 I don't see a great benefit to changing where the _custom.css gets referenced from, personally. Not a 2.0 thing, which means that we'll have to doc things one way for 2.0 and another for 2.1? 2011-02-18T15:26:30 * phasefx doesn't either 2011-02-18T15:26:35 eeevil: are you saying it currently does use the UPDATE_HOLD permission or you propose that it should? 2011-02-18T15:28:09 i'm figuring we'd need to look at Holds.pm and make sure the cut_in_line checks to see if the user has that permisson? 2011-02-18T15:29:25 mrpeters-isl: phasefx said there was a perm for updating the request time. whether that is UPDATE_HOLD or not, I don't know 2011-02-18T15:29:44 but, really, yes, UPDATE_HOLD should be enough 2011-02-18T15:29:51 or, might be 2011-02-18T15:30:27 comment in the code says, "Changing the request time is like playing God", and looks for UPDATE_HOLD_REQUEST_TIME 2011-02-18T15:30:27 is there a similar action that would require UPDATE_HOLD or some other permision that i could look at as an example? 2011-02-18T15:31:12 line 794 of Holds.pm 2011-02-18T15:31:24 other perms that can be needed depending on circumstances: UPDATE_PICKUP_LIB_FROM_TRANSIT, and UPDATE_PICKUP_LIB_FROM_HOLDS_SHELF 2011-02-18T15:32:49 is there a way to change the request time from the staff client? or would this only be done via the API's? 2011-02-18T15:33:01 just not sure where this would ever come into play 2011-02-18T15:33:23 mrpeters-isl: it's one way of cutting in line 2011-02-18T15:33:34 before cut-in-line existed 2011-02-18T15:33:52 also useful for pushing a hold /down/ 2011-02-18T15:34:10 think: re-placing a hold that got destroyed somehow 2011-02-18T15:34:15 ok, but you're saying that the "new" cut in line does not operate this way? 2011-02-18T15:34:30 *** granitize has quit IRC 2011-02-18T15:35:04 mrpeters-isl: you can do more with the change-request-time than cut-in-line 2011-02-18T15:35:25 cut-in-line is single purpose thus less "dangerous" 2011-02-18T15:35:30 and reversable 2011-02-18T15:36:11 but would staff ever be able to do that, or is it something that could only be done via srfsh? 2011-02-18T15:36:30 i've not run accross the ability to change the hold placement date in the staff client...but i coudl have missed it 2011-02-18T15:36:42 mrpeters-isl: in any case, if someone were to submit a patch that either reused the UPDATE_HOLD_REQUEST_TIME perm for cut-in-line changes or invented a new permission and included all required bits, I think it would likely be accepted 2011-02-18T15:36:53 *** sfortin has quit IRC 2011-02-18T15:36:58 mrpeters-isl: I'm sure it's in the staff client. probably context menu on holds? 2011-02-18T15:38:55 Edit Request Date 2011-02-18T15:39:37 * tsbere starts over on his custom css move stuff, finding that there is at least one external to the staff client css file that does that trick 2011-02-18T15:40:53 *** kmlussier has quit IRC 2011-02-18T15:41:53 tsbere: see above - not sure we really want that anyway 2011-02-18T15:42:41 I can't get my auto-fixer to work outside of the staff client right now anyway, apparently. 2011-02-18T15:49:02 eeevil: that's what id like to do 2011-02-18T15:50:56 i'm just trying to equate the example for changing the request time to the cut_in_line command 2011-02-18T15:52:14 so i'm reading this as if the "new" request time != the original request time, it moves on to check that the requesting user has the UDPATE_HOLD_REQUEST_TIME permission at the depth of the pickup library 2011-02-18T15:52:22 if the user doesnt, it fails 2011-02-18T15:55:27 right, so, if cut_in_line is true and old cut_in_line was not, check perm 2011-02-18T15:56:21 ok, thats where you lose me - the only cut_in_line code i see is at 1063 and onward 2011-02-18T15:57:10 it's be new code, that's my point. it needs to be a patch that implements this test 2011-02-18T15:57:16 in the appropriate locations 2011-02-18T15:57:36 s/it's/it'd/ 2011-02-18T15:57:59 ok, so the requesting user would update the action.hold_request row to say that cut_in_line was true and also update the request time to now() which would put them in front of everyone? 2011-02-18T15:58:30 we actually leave request_time alone when using cut_in_line 2011-02-18T15:58:53 if more than one person uses cut_in_line, then request_time becomes the tie-breaker 2011-02-18T15:59:22 ok i was just going to ask if more than one cut in line could be set 2011-02-18T15:59:49 *** Meliss has quit IRC 2011-02-18T16:01:59 is if($hold->cut_in_line ne $orig_hold->cut_in_line) appropriate, since we're comparing aganist what likely would be a null value for that ahr row? 2011-02-18T16:02:40 also, what is the purpose of: return OpenILS::Event->new('BAD_PARAMS') if $hold->fulfillment_time; 2011-02-18T16:02:57 does that make sure you don't set a request_time after the fulfillment_time? 2011-02-18T16:04:43 it's not this simple, i don't imagine? http://pastie.org/1580243 2011-02-18T16:06:05 I think that is supposed to say "why are you changing a hold that is already filled?" 2011-02-18T16:06:21 tsbere: ah, that makes sense 2011-02-18T16:07:24 mrpeters-isl: you don't want to use ne, use $U->is_true instead (NOTE: check how is_true is spelled) 2011-02-18T16:08:47 so, knowing i'm out of my element with perl - is it too much to ask to know what the difference there is? i was assuming ne was "not equal" 2011-02-18T16:09:04 There are different ways of saying not equal 2011-02-18T16:09:07 Depending on data types 2011-02-18T16:09:11 so, this new check means you'd only process the rest of the if statement if the new boolean value is true? 2011-02-18T16:09:12 (I think) 2011-02-18T16:10:48 tsbere: sorta ... you can say "1" or "t" or "y" or true to postgres (we sometimes say "1") but it always says "t" to you 2011-02-18T16:11:03 mrpeters-isl: == is for numeric comparison, eq/ne is for non-numeric (like strings), and... eeevil just beat me to the explanation of is_true 2011-02-18T16:11:18 mrpeters-isl: you'd only want to do it if 1) the new value is true and 2) the old value was not 2011-02-18T16:11:31 right, it has to have something to compare itself to 2011-02-18T16:11:42 i assume the default for cut_in_line is just null? 2011-02-18T16:11:49 which is still comparable against "true" or "false" right? 2011-02-18T16:11:51 Thus.....I think you want $U->is_true checking with a != 2011-02-18T16:12:25 eeevil: I would think you would want the permission to be checked if they are setting OR unsetting the value. 2011-02-18T16:12:35 can you unset it? 2011-02-18T16:12:36 (can you unset the value?) 2011-02-18T16:12:39 heh 2011-02-18T16:12:40 or ($U->is_true($new->cut_in_line) and !$U->is_true($old->cut_in_line)) 2011-02-18T16:12:57 I don't think there's a way currently, no 2011-02-18T16:13:07 phasefx? 2011-02-18T16:13:30 eeevil: ok, so whats the "old" and "new"? 2011-02-18T16:13:32 the implementation of is_true in OpenILS::Application::AppUtils is really more of an "is_not_false": return 1 if $item and $item !~ /^f$/i; 2011-02-18T16:13:56 mrpeters-isl: just psuedocode 2011-02-18T16:13:57 versus hold and orig_hold 2011-02-18T16:14:06 is that the is_true we're working with here or... [digs] 2011-02-18T16:14:22 sorry, i'm distracting from the conversation at hand. 2011-02-18T16:14:38 jeff: yes 2011-02-18T16:14:48 (to the question, not the comment) 2011-02-18T16:14:55 thanks :) 2011-02-18T16:15:00 so is ($U->is_true($hold->cut_in_line) and !$U->is_true($orig_hold->cut_in_line)) the same thing? 2011-02-18T16:15:32 right 2011-02-18T16:15:57 ok, so http://pastie.org/1580243 is now correct? 2011-02-18T16:16:08 i guess i don't understand how this gets "checked" still 2011-02-18T16:16:17 eeevil: pretty sure the UI just sets true or false 2011-02-18T16:16:27 when i try to update a cut in line in the staff client now, it just now sees this new code in Holds.pm? 2011-02-18T16:16:35 phasefx: it toggles? 2011-02-18T16:16:59 it asks you yay or nay 2011-02-18T16:17:08 mrpeters-isl: the staff client will get an error, which will need to be handled 2011-02-18T16:17:24 phasefx: so, the question was, can you unset cut-in-line 2011-02-18T16:17:58 eeevil: by setting it to false, but not setting it to null (at least through the UI) 2011-02-18T16:18:23 phasefx: so yes, you can turn off cut-in-line 2011-02-18T16:18:33 you get presented with two buttons, "Top of Queue" and "No Top of Queue" 2011-02-18T16:18:40 ahh 2011-02-18T16:18:44 got it 2011-02-18T16:18:55 * eeevil is not looking at the UI atm 2011-02-18T16:19:19 in that case, mrpeters-isl, you will want to replace "and !" with "ne " 2011-02-18T16:19:36 assuming you want to protect downgrading of currently-cutting holds 2011-02-18T16:20:02 ok 2011-02-18T16:20:03 eeevil: Would he want "ne" or "!="? 2011-02-18T16:20:22 tsbere: either's fine ... ne will cast to a string for comparison 2011-02-18T16:20:49 and with that, I'm OUTTA HERE! back later, probably 2011-02-18T16:42:19 *** rickd_ has quit IRC 2011-02-18T16:43:51 *** bshum has quit IRC 2011-02-18T17:02:31 *** afterl has left #evergreen 2011-02-18T17:07:09 *** pmplett_ has quit IRC 2011-02-18T17:18:24 *** yboston has quit IRC 2011-02-18T17:23:39 *** Dyrcona has quit IRC 2011-02-18T17:31:08 *** pmplett_ has joined #evergreen 2011-02-18T17:42:04 *** gdunbar has quit IRC 2011-02-18T17:44:19 *** bshum has joined #evergreen 2011-02-18T17:48:25 *** pmplett_ has quit IRC 2011-02-18T17:50:08 T-10! 2011-02-18T17:54:17 *** pmplett_ has joined #evergreen 2011-02-18T17:55:02 T-5! 2011-02-18T17:55:20 Doh 2011-02-18T17:59:22 And................. 2011-02-18T18:00:02 Off we go. 2011-02-18T18:13:46 *** mtate has quit IRC 2011-02-18T18:17:51 *** tater has joined #evergreen 2011-02-18T18:25:25 bshum: you guys are live on 2.0? 2011-02-18T18:25:34 dbs: Heh, no 2011-02-18T18:25:39 Just started the process 2011-02-18T18:25:41 boooo 2011-02-18T18:25:42 :) 2011-02-18T18:26:04 Moving the whole thing from a single server to a multi-server install 2011-02-18T18:26:17 And upgrading from 1.6.1 to 2.0 along the way 2011-02-18T18:26:20 But soon! 2011-02-18T18:31:42 *** tsbere_ has joined #evergreen 2011-02-18T18:31:50 *** tsbere has quit IRC 2011-02-18T18:47:23 *** pmplett_ has quit IRC 2011-02-18T19:17:34 perl++ 2011-02-18T19:18:18 just figured out how to parse a CSV - I know this is newbie stuff, but for me this is a step forward ;-) 2011-02-18T19:18:40 bshum: good luck! 2011-02-18T19:20:10 csharp: Thanks! I'm putting together our NFS share sometime this weekend. Hopefully without breaking everything. 2011-02-18T19:25:19 I may be around on chat if you want moral support if nothing else ;-) 2011-02-18T19:25:41 The examples you shared for the load balancers was insightful. 2011-02-18T19:25:47 good 2011-02-18T19:25:52 Log shipping was quite painful though. 2011-02-18T19:26:03 heh - sorry 2011-02-18T19:26:12 Had syslog-ng 3.0 installed on our heads because they had backports enabled for some reason. 2011-02-18T19:26:26 And they're incompatible with the syslog-ng examples you and from ESI-contribs 2011-02-18T19:26:32 yikes 2011-02-18T19:26:43 So had to downgrade them first back to 2.0.x first and then run 2011-02-18T19:26:48 new_versions-- 2011-02-18T19:27:03 There must also be some custom log stuff in there, eg_stats doesn't work quite right 2011-02-18T19:27:06 or I should probably say 2011-02-18T19:27:15 backwards_incompatibility-- 2011-02-18T19:27:25 ah 2011-02-18T19:27:45 All that being said, fun exercise, and big thanks for some insights :) 2011-02-18T19:28:03 sure! 2011-02-18T19:39:44 *** tpham has joined #evergreen 2011-02-18T21:06:07 transactions++ # saved us from bad sequences 2011-02-18T22:04:19 *** rsinger_ has joined #evergreen 2011-02-18T22:06:05 *** rsinger has quit IRC 2011-02-18T22:06:05 *** rsinger_ is now known as rsinger 2011-02-18T22:27:26 *** bshum has quit IRC 2011-02-18T22:37:45 *** pmplett_ has joined #evergreen 2011-02-18T23:12:45 *** jennam has joined #evergreen 2011-02-18T23:12:51 *** jennam has joined #evergreen 2011-02-18T23:28:31 *** tpham has quit IRC 2011-02-18T23:45:43 *** jennam has quit IRC 2011-02-18T23:52:38 *** jennam has joined #evergreen