Evergreen ILS Website

IRC log for #evergreen, 2014-12-22

| Channels | #evergreen index | Today | | Search | Google Search | Plain-Text | summary | Join Webchat

All times shown according to the server's local time.

Time Nick Message
00:05 chatley_ joined #evergreen
00:10 mmorgan1 joined #evergreen
00:10 hopkinsju joined #evergreen
00:12 book` joined #evergreen
01:12 nhilton joined #evergreen
05:04 pinesol_green Incoming from qatests: Test Success - http://testing.evergreen-ils.org/~live/test.html <http://testing.evergreen-ils.org/~live/test.html>
06:31 BigRig joined #evergreen
07:36 collum joined #evergreen
07:50 julialima_ joined #evergreen
07:55 rjackson-isl joined #evergreen
07:58 jboyer-isl joined #evergreen
08:15 akilsdonk joined #evergreen
08:16 remingtron joined #evergreen
08:21 _bott_ joined #evergreen
08:28 mrpeters joined #evergreen
08:39 Shae joined #evergreen
08:54 sarabee joined #evergreen
09:00 dbwells_ joined #evergreen
09:40 abowling joined #evergreen
10:47 dreuther_ joined #evergreen
11:00 bshum Bleh, MaxClients reached on apache servers
11:00 bshum What a way to start a Monday :(
11:14 jeff What a terrible night to have a curse...
11:14 jeff s/terrible/horrible/
11:16 jboyer-isl jeff: Are you looking for Dracula
11:16 jboyer-isl ’s tooth, or have I missed your reference?
11:25 mllewellyn joined #evergreen
11:58 jihpringle joined #evergreen
11:59 jeff jboyer-isl: nope, you're on-reference. Castlevania II.
12:01 * csharp remembers playing the original Castlevania on an Apple II Plus
12:02 phasefx Castlevania II was the best one, though ;)
12:02 csharp wait - no that was Castle Wolfenstein
12:02 jeff i wasn't aware of there being a pre-NES version.
12:02 jeff heh
12:02 csharp jeff: I was looking that up because it didn
12:03 csharp 't sound right
12:03 jeff heh
12:03 csharp I'm sure I played one of the games in the series
12:03 jeff ah, yes. the "before it was 3D and ended in EXE" version :-)
12:03 terran_ joined #evergreen
12:03 csharp I liked both versions
12:07 terran_ Hi all - can any of you point me to info regarding configuring search relevancy ranking?
12:09 bshum terran_: Hmm, what sort of relevancy ranking changes are you looking to implement?
12:10 dreuther joined #evergreen
12:11 bshum I don't think there exists such things, but maybe it's named something different.
12:11 terran_ bshum - one example: if the search word appears in the proper title (during a title search), then I'd like it to appear higher on the list
12:13 bshum terran_: the best i can think of for that might be this article by dbs from the 1.6 era:  https://coffeecode.net/archives/218​-Adjusting-relevancy-rankings-in-Ev​ergreen-1.6,-some-explorations.html
12:14 bshum We still use that to create customized higher weighted metabib indexes in our keyword searches.
12:14 terran_ Thanks Ben, I hadn't run across that one, I'll take a look!
12:14 bshum Nobody uses the search.relevance_ranking tables in Evergreen though, there's some performance issues with TPAC that were never resolved.
12:14 bshum Or maybe it was 2.0 era stuff
12:14 bshum Whatever the case, it's been forever
12:15 bshum The only other thing I know is that there's some cover density stuff in the opensrf.xml config too
12:16 bshum That stuff is all related to the new QueryParser stuff that I'm still slowly learning more about.
12:17 terran_ Thanks, good to know not to bother with the search.relevance_ranking tables. Last thing I need is performance issues.
12:18 bshum I don't know if that's still the case, but it was ages ago when we first ran TPAC in 2.2 era
12:19 bshum Since it's just title
12:19 bshum And you want proper title to rank higher than say, other titles?
12:19 bshum Then maybe you can just adjust the weight of proper title metabib index
12:19 bshum To be 2 instead of 1 or whatnot
12:19 bshum I'm just spitballing
12:20 bshum I have no idea what that'd do :)
12:21 bshum I would assume that would make any matches in title:proper come slightly more ahead of matches in the other title indexes.
12:22 jeff today is the first day that my API shim for hoopla patron auth is in "real" production. it's working!
12:24 terran_ I'm not accessing the tables directly, and the closest I can find in the client interface are the admin > server admin > MARC search screens. I tried changing the title proper weight there, and saw no difference, but I don't even know if I'm looking in the right place
12:25 csharp jeff++
12:25 * bshum doesn't know where things live in client side, as he's only ever manipulated them via DB
12:26 jeffdavis joined #evergreen
12:26 * csharp is pretty sure the client only gives partial access to what terran_ is after
12:26 bshum terran_: MARC search facet fields is the one I think brings the closest association
12:27 csharp @love chicken soup
12:27 pinesol_green csharp: The operation succeeded.  csharp loves chicken soup.
12:27 bshum There's weight field in there for all the metabib fields
12:28 bshum The more I look at all the MARC search* stuff in the client, the less sure I am of how the cover density stuff operates now.
12:28 terran_ heh
12:28 bshum eeevil probably knows more :)
12:29 bshum I poked a bit when we upgraded back in 2.4 era
12:30 bshum But haven't touched much since then.
12:31 terran_ MARC search facet fields does appear to be the closest I can get without making csharp help me :)  I'll play around more after I go get food. Thanks!
12:37 nhilton joined #evergreen
13:26 kmlussier @later tell terran if you increase the weight of the title proper field there, it will only affect relevancy of searches in a title search. For keyword searches, you need to add a new index.
13:26 pinesol_green kmlussier: The operation succeeded.
13:44 nhilton joined #evergreen
13:45 nhilton joined #evergreen
13:49 dreuther joined #evergreen
14:10 abowling hi, all. doing some testing to hopefully fix some bugs, and have what's probably a relatively elementary question. just for pure testing purposes, i'm adding a variable to the $ctx array (e.g. $ctx->{foo} = "bar"), but when I try to output it in a TT2 file, even after restarting opensrf and apache services, it doesn't show. can someone give me a brief background tutorial on this?
14:11 tsbere abowling: Where you are attempting add it would be a good thing to know, I think.
14:11 abowling sorry. correct you are, tsbere
14:11 abowling EGCatLoader.pm
14:12 dreuther_ joined #evergreen
14:13 tsbere abowling: In that case, the function you are adding it to may be a good thing to know as well.
14:14 abowling tsbere: load_common
14:15 tsbere abowling: Huh. Have you installed the new file and restarted apache?
14:15 abowling another related issue that i've run into, even if i change, purely for testing purposes, $ctx->{opac_root} to "eg", it still outputs as "/eg/opac"
14:15 abowling I've restarted apache and opensrf services
14:16 tsbere but did you reinstall the changed file?
14:16 abowling is there anything i'm missing. for testing, i'm just directly editing the file and outputting, so based on your last question, probably not
14:16 tsbere If you are editing the file in, say, an extracted tarball or git checkout then you will need to copy the file in place manually or re-run the make install step.
14:17 abowling a-ha
14:17 abowling i'm editing directly on the folder it's running, not in a git checkout
14:17 abowling but i've not run the make install
14:19 abowling tsbere: thanks, though. i bet that's the cause. ?
14:20 tsbere abowling: What folder is the copy you are editing in, if you don't mind me asking?
14:20 abowling /usr/local/share/perl/5.18.2​/OpenILS/WWW/EGCatLoader.pm
14:20 tsbere And what user are you editing as?
14:20 abowling root
14:21 tsbere Presuming that is, in fact, the right version of perl for your system a full restart of apache should pick up your changes, then. (Reload may not, from my experience)
14:22 abowling yeah, many parties agree. that's why i'm scratching my head.
14:26 tsbere abowling: I would run this for a quick check of what copy or copies exist in the system: find /usr/local/share/perl -name 'EGCatLoader*'
14:26 tsbere abowling: And maybe this to see what include path(s) you may need to be looking at in general, for that matter: perl -e 'print join(",",@INC);'
14:29 abowling thanks, tsbere. i'll look into these
14:31 tsbere abowling: In fact, in a somewhat combination of things, this should output things somewhat in order, unless apache has a significantly different @INC: for perldir in `perl -e 'print join(" ",@INC);'`; do find $perldir -name 'EGCatLoader.pm' 2>/dev/null; done
14:35 jcamins @later tell rfrasur I got my foldscope! :D And as soon as the construction has finished I plan to assemble it and start inspecting flour at high magnifications! :D
14:35 pinesol_green jcamins: The operation succeeded.
14:35 abowling tsbere: just the one file. thanks. i'll dig elsewhere. thanks for your help.
14:46 tsbere abowling: As another thought: Have you ensured you are editing the correct *template* file?
15:05 abowling tsbere: yep. when i put other stuff in there, it shows up correctly
15:10 mrpeters this is quite wonky indeed -- https://eg-master.emeralddata.net/eg/kpac/getit/ still thinks the ctx.kpac_root is "/kpac" even though I've changed it to "kidsopac"
15:13 tsbere mrpeters: Where did you make the change?
15:13 mrpeters the proper perl version, EGCatLoader.pm
15:13 tsbere mrpeters: That is something that, I believe, needs to change in EGKPacLoader.pm, not EGCatLoader.pm
15:14 mrpeters you might just be right
15:14 mrpeters there is a kpac_root in the EGCatLoader though
15:14 tsbere There is?
15:14 mrpeters yeah
15:14 abowling mpeters: i put the kpac_root there for testing
15:14 mrpeters oh
15:15 abowling it wasn't there, natively
15:15 mrpeters heh i thought you said the original file was back in place :)
15:15 abowling i did it after i told you that, so i didn't lie ;)
15:16 mrpeters k, lets get this back to stock and start over here
15:16 abowling now, it's back to the original
15:16 abowling but that works for me
15:17 tsbere I will point out, if you are messing with kpac files/templates, and editing EGCatLoader.pm, EGKPacLoader.pm may be overriding what you are doing.
15:17 tsbere Hence a lack of visible results
15:20 mrpeters yeah,very nice catch tsbere
15:23 csharp autogen--
15:23 * csharp just spent more than an hour troubleshooting a problem that was really all autogen's fault
15:23 csharp @blame autogen
15:23 pinesol_green csharp: It's all autogen's fault!
15:23 csharp pinesol_green: yeah!
15:23 pinesol_green csharp: have you tried local mean solar time for the named city as the reference point?
15:24 csharp @blame add $who is really just another name for autogen
15:24 pinesol_green csharp: The operation succeeded.  Blame #17 added.
15:25 abowling thanks, tsbere. good note to consider.
15:25 tsbere abowling: Had I known you were messing with kpac I would have brought it up sooner. ;)
15:31 abowling yeah, but it still should show the array variable changes
15:36 tsbere abowling: Well, changing kpac_root isn't going to help outside of EGKPacLoader.pm, for example, because EGKPacLoader.pm overwrites it *after* load_common from EGCatLoader.pm runs, for example.
15:46 julialima_ left #evergreen
16:00 terran_ joined #evergreen
16:01 kmlussier Do we have copies of the Evergreen logo available for download on the web site?
16:02 kmlussier The results I get when I search for logo on the site search are...interesting.
16:04 kmlussier I guess this - http://www.georgialibraries.​org/lib/pines/evergreenlogo/ - should be moved over at some point.
16:06 kmlussier But it looks like that's already on our to-do list. Ignore my babbling.
16:10 mrpeters kmlussier: i still have the SVG vector art of the icons used for the staff client if you need them
16:10 mrpeters its different from that "evergreen mark" .ico
16:10 mrpeters one is yellow, the red is for beta builds
16:11 mrpeters be happy to send them over to you
16:12 kmlussier mrpeters: Thanks! The one I found in the above link works for me. I just needed something for the conference registration form.
16:12 mrpeters ok then :)
16:12 mrpeters i have the eps files for all, if you need them too -- they can be blown up as big as a skyscraper if you wanted :)
16:17 vlewis joined #evergreen
16:18 nhilton_ joined #evergreen
17:01 kmlussier That's it for the year. Happy holidays #evergreen!
17:01 RBecker joined #evergreen
17:03 RBecker joined #evergreen
17:03 mrpeters left #evergreen
18:52 nhilton joined #evergreen
19:16 bmills joined #evergreen
22:54 bmills joined #evergreen

| Channels | #evergreen index | Today | | Search | Google Search | Plain-Text | summary | Join Webchat