IRC log for #koha, 2008-05-15

All times shown according to UTC.

Time S Nick Message
13:24 danny good morning #koha
13:25 ryan hi danny
13:25 owen Hi danny, ryan
13:31 lloyd_ atz, you around?
13:44 lloyd_ lists.koha.org down?
13:45 masonj hmm, could be..
13:48 lloyd_ looks like it then :)
13:49 atz lloyd_: what's up?
13:52 lloyd_ Owen said you might have done some work on googlejackets?
13:54 hdl hi
13:54 hdl lloyd_: yes.
13:55 lloyd_ well... I'm looking for a way to play with the returned urls a little
13:55 atz lloyd_: that's true.  we have amazon, baker & taylor and google as possible image sources
13:55 lloyd_ atm it returns a url like this : http://bks2.books.google.com/b[…]ytx4hLVytjtPPO4xo
13:56 lloyd_ I need to be able to tweak it some how so the domain name changes from bks2.books.google.com to http://www.libcat.eoe.nhs.uk/googlebooksapi/
13:56 lloyd_ thus making the image url : http://www.libcat.eoe.nhs.uk/g[…]ytx4hLVytjtPPO4xo
13:57 atz is this for caching?  or just your own image source?
13:57 lloyd_ best way to explain why I need this is, some of our internal site dont have access to the internet, but do have access to www.libcat.eoe.nhs.uk
13:58 lloyd_ http://www.libcat.eoe.nhs.uk/googlebooksapi/ is just a reverse proxy (using mod_proxy) to bks2.books.google.com
13:58 lloyd_ you follow?
13:58 hdl lloyd_ : optic fiber cable was torn.
13:59 lloyd_ I hate that.. It's even worse when they get stuck in the cabinet door and you ruin them
14:01 atz lloyd_: i'm not sure how is best to address that as a long term fix.  
14:01 lloyd_ indeed
14:01 atz the google js code is locally koha-tmpl/opac-tmpl/prog/en/js/google-jackets.js
14:02 lloyd_ I suspect others will have the problem and need the work around
14:02 atz this is the kind of thing I categorize as "not a koha problem", meaning it falls to system integrator type ppl (like myself)
14:02 lloyd_ if it's possible to make a syspref to hack the url, I could write up a howto on setting up a reverse caching proxy in apache
14:04 lloyd_ I agree, it's not a koha problem. but lots of sites will suffer from the same problem. Like all UK schools and UK Hospitals
14:04 lloyd_ so having a work around, feature, setting could be useful?
14:04 atz lloyd_: are you at least sending X-REQUESTED_BY headers via apache?
14:05 atz ( i think that's default behavior, but I haven't set up proxies on 2.2)
14:05 lloyd_ lemme show
14:06 lloyd_ http://pastebin.com/m5e0c448c
14:06 lloyd_ the config could obviously be made a little better, using bks1-9.books.google.com and loadbalancing over them
14:07 atz right
14:07 atz so if you wanted to hack the static js, you could just edit that file I referenced
14:08 atz and do a js regexp to get what you want
14:08 atz but this is a compound proble, I think
14:08 atz *problem
14:08 atz you have the image request... but first you have the "index" request
14:08 atz that returns the URL of the image
14:09 atz so you would need a second proxy setup
14:09 lloyd_ yes, for books.google.com
14:09 atz right
14:09 lloyd_ then a function to rewrite the url after it's returned
14:09 atz and, in fact, if you proxy that, you might not need the later one, if you care to hack the response
14:10 lloyd_ Seeing how to do it isn't a problem
14:10 lloyd_ I would like the fix/workaround to be available to the community
14:11 atz there are 2 options for broader integration
14:12 atz one is to move the js file into a dynamic position where it gets processed instead of served statically: then you could have some syspref take effect
14:12 hdl lloyd_: why don't you use mod proxy for your purpose ?
14:13 lloyd_ hdl, I am :p
14:14 atz the other is to copy the 50 line file to "google-jackets-proxy.js".   edit it to your liking and allow a syspref to select the google js file.
14:15 lloyd_ I think the first way is best?
14:16 atz lloyd_: i actually prefer the second, but if you're doing the work, you get to pick
14:16 owen The advantage to having the js file served statically is that it can be cached by the browser
14:16 atz right
14:16 lloyd_ well hmmz
14:16 atz (or any other cache)
14:17 lloyd_ linking to a seperate google-jackets.js file, would mean another file you have to retain when upgrading versions
14:18 atz lloyd_: it would be like a local CSS file
14:18 lloyd_ so, hack the .js and add syspref option to change the location of google-jackets-proxy.js
14:20 lloyd_ doesnt feel right :)
14:21 atz i used to support libraries in juvenile prisons where their network was thoroughly locked down
14:21 owen atz, what are the prefs required for Baker and Taylor images? I nuked my system prefs yesterday by accident and now they're not there
14:21 atz owen: i think they all moved to "enhanced content" tab
14:22 atz w/ the other jacket image stuff
14:22 owen Yeah, I don't have any B&T stuff listed there
14:23 atz odd.. i have a B&T tab still, but it's empty
14:23 lloyd_ If I hack the .js and write howto on setting it up (posting on the wiki), is someone up for adding the syspref for google-jacket-proxy.js location?
14:23 hdl lists.koha.org is back
14:24 atz lloyd_: can't promise any kind of timeline, but I'd be willing to do it
14:26 lloyd_ ok... I'm not sure what's involved but it'd be nice to get it done before B3
14:26 atz owen: for me they show up in Enhanced Content tab still
14:26 atz not sure what's up there
14:29 owen atz, do you have more time for book cover talk?
14:30 atz ok
14:30 owen The book covers aren't working for me in the shelf browser now
14:31 owen I think it's a problem with the ISBN being used in the shelf browser loop?
14:31 owen It's another place where the book cover image choices come into play
14:32 atz do I have to enable that?
14:33 owen Yes. Do you have an OPACShelfBrowser preference?
14:33 owen I didn't, after my system prefs debacle. I had to add it by hand.
14:33 atz i don't have one... rebased yesterday
14:36 atz owen: ok, enabled (and the system recognized it, put it in the OPAC ta
14:36 atz *tab
14:36 atz where do I find it in the opac though?
14:37 atz ah... "Browse Shelf" perhaps
14:37 atz owen: what is your image source?
14:38 atz right now I'm getting B&T ok.
14:38 owen I've just tried Google and B&T with no success
14:38 atz do you have a B&T account?
14:39 owen kados gave me some credentials so I could test
14:40 owen I'll check my copy of opac-detail.tmpl. Maybe its something I changed.
14:41 atz and you just updated in the past day or so?  
14:41 atz we should have the same version then...
14:44 owen Hm, It was just the luck of the draw, and I just wasn't getting images.
14:45 atz test search "modern phrenology"
14:45 atz :)
14:45 atz or something equally obscure/old
14:46 owen http://oleonard.dev.kohalibrar[…]=943#shelfbrowser
14:46 owen With Google Jackets now. Why does the cover show above, but not in the browser?
14:46 owen I guess because the isbn isn't sanitized in the shelf browser.
14:47 atz good question.... looks inadequately sanitized
14:47 atz (as you just said)
14:48 atz i have trouble believing that library data *today* still has trouble specifying ID numbers w/o interjecting other crap into it...
14:48 atz shouldn't there be a spec that says "in this field/subfield, here liveth the ID and naught else"
14:51 atz oh well...
14:51 owen It doesn't seem like MARC is known for its data cleanliness...
14:51 atz owen: opac-search.pl creates a "clean_isbn" field for hit results
14:52 slef hi all
14:52 atz greets slef
14:56 owen atz, what's the difference between clean_isbn and amazonisbn?
14:57 atz probably not much
14:57 owen In the template, Amazon and Google use amazonisbn, B&T uses clean_isbn
14:58 atz actually... i already injected a FIXME at 228 of opac-detail
14:58 atz clean_isbn is better, i believe...
14:59 atz my $xisbn=$dat->{'isbn'};
14:59 atz $xisbn =~ /(\d*[X]*)/;
14:59 atz $template->param(amazonisbn => $1);     # FIXME: so it is OK if the ISBN = 'XXXXX' ?
14:59 atz my ($clean, $amazonisbn);
14:59 atz $amazonisbn = $1;
14:59 atz # these might be overkill, but they are better than the regexp above.
14:59 atz if (
14:59 atz    $amazonisbn =~ /\b(\d{13})\b/ or
14:59 atz    $amazonisbn =~ /\b(\d{10})\b/ or
14:59 atz    $amazonisbn =~ /\b(\d{9}X)\b/i
14:59 atz ) {
14:59 atz    $clean = $1;
14:59 atz    $template->param(clean_isbn => $1);
14:59 atz }
15:00 atz and there should be many more, unfortunately
15:01 owen Shouldn't that isbn processing be a function that can be re-used?
15:01 atz yeah
15:05 atz owen: looks like the unsanitized isbn has a line break in it
15:05 atz before the "(pb"
15:05 owen Yeah, I don't even know how that would happen
15:06 atz so that *could* be addressed w/ data cleanup
15:06 atz though koha should be consistent in any case
15:14 lloyd_ atz or owen
15:14 lloyd_ either of you fancy replacing google-jackets.js on your dev system with this : http://www.libcat.eoe.nhs.uk/g[…]-jackets-proxy.js
15:16 owen lloyd_, I'll take it for a whirl
15:16 lloyd_ *bites nails*
15:20 owen atz, gmcharlt, ryan: Should I be worried that WinSCP is warning me about arwen's host key changing?
15:20 atz ip address?
15:21 atz should be 72.232.235.42
15:21 atz hrm... i just got the warning myself
15:22 gmcharlt owen, atz: probably related to Debian OpenSSH bug
15:22 atz gmcharlt: the key changed... what did it?
15:22 gmcharlt atz: clay cleaning up after the bug, I assume
15:22 owen gmcharlt: ignore?
15:23 gmcharlt atz: fix includes regenerating ssh keys
15:23 atz i see... this will screw up many many things
15:24 gmcharlt owen: ignore - I confirmed that Clay has in fact changed the key
15:24 atz ok, thx gmcharlt
15:44 atz owen: got a patch for you to try out on your shelfbrowser problem
15:45 owen thanks atz
15:48 atz let me know if that helps at all
16:17 owen Looks like that did help, atz
16:18 owen Now the new issue: The cover image for the current title doesn't show in the shelf browser... I'm guessing because they have identical IDs?
16:18 atz owen: ah right... they would.  
16:45 owen atz, what does the greybox stuff have to do with Google Jackets?
16:48 atz owen: don't know what you mean
16:49 owen <!-- TMPL_IF NAME="GoogleJackets" --><script type="text/javascript">                            
16:49 owen    var GB_ROOT_DIR = "/intranet-tmpl/prog/en/lib/greybox/";  ...etc.
16:50 atz cute name.  "it's not a black box... it's a *grey* box."
16:51 atz just another directory though afaik
16:54 owen Hmm... Taking that out doesn't affect the display of Google book jackets
16:56 owen I wonder if there was something else planned that required the greybox library...
16:59 atz not sure... maybe to allow staff to compare image sources
17:59 owen Hi bb_marblehead
17:59 bb_marblehead Hi owen
18:23 ryan owen et al, speaking of external content, how does that affect ssl ?
18:24 ryan i see i get a warning in my browser on an ssl site, i assume it's the external links that are causing it
18:24 owen Yeah. Would the browser show the same warning if you were showing content from two different secure connections?
18:26 owen Maybe that's a good reason to use lloyd's proxy system
18:27 ryan i'm not sure of the requirements
18:27 ryan it would be nice to get rid of that error though
18:27 owen ryan, you have a client who requires a secure connection for their opac?
18:28 ryan no, the staff client.
18:28 ryan yes, that'd be good
18:55 tim I get an error about an unknown column 'done' in a where clause when I run missing090field.pl.  This is from a git pull less than an hour ago.
18:55 tim It's looking for the 'done' column in the 'zebraqueue' table.
20:42 pianohacker Just out of completely random curiousity, why is the Koha module namespace named C4?
20:44 gmcharlt pianohacker: I've heard different stories
20:45 gmcharlt one of them comparing it to the explosive
20:45 gmcharlt but don't know definitively
20:49 acmoore kados seemed to have a story involving an application C2 that replaced something (C, maybe?). The next one was going to be C3, but they may have skipped that in order to make a good joke.
20:49 acmoore it all sounds a little "made up" to me.
20:53 pianohacker Huh
20:53 chris well seeings as i named it
20:53 chris i can tell ya why
20:53 acmoore aha!
20:53 pianohacker Yay
20:53 chris they indeed were running a system called C2
20:54 chris C3 was supposed to have been released about 3 years before .. never was
20:54 chris so i jokingly named it C4
20:55 pianohacker Hah
20:55 pianohacker and the Koha name came along later?
20:55 chris yeah Koha came along when we released it
20:56 chris it was called Kumara for a while there too
20:56 chris in fact i still have hte old Kumara cvs repo
20:56 chris it went C4 -> Kumara -> Koha
20:57 chris never got round to renaming the package space though :)
20:57 gmcharlt chris: what's a kumara?
20:57 chris sweet potato
20:57 pianohacker Bahaha
20:57 chris was the staple source of starch/carbohydrates for maori
20:58 chris http://www.teara.govt.nz/1966/K/Kumara/Kumara/en
21:00 chris Katipo has a tradition of using K words
21:00 chris Koha, Kea, Kete etc
21:01 russel Konae was another one
21:01 chris ahh yep
21:01 chris my first opensource release .. timesheeting software .. never really went anywhere tho
21:01 chris i think Katipo and Face use it still though
21:11 chris hey nicole
22:02 liz question... we have done something to our koha3 that has locked out all of our users, it's technically ok since it's only a test server, but we'd really rather not dump the database... is there anything we can do?
22:03 liz We think we deleted or changed a bunch of libraries or branches while also changing category codes
22:06 liz well bummer, guess everybody went home. I'll ask again tomorrow
22:11 gmcharlt liz: I'll be leaving soon, but could you provide some more details about what you were doing?
22:12 liz the sad thing is that I wasn't standing there when it happened, but as I understand it the user was deleting libraries/branches and changing category codes
22:13 liz my guess is that he changed the home library that his user was affiliated with and now... who knows which library he is a patron of
22:13 liz but the kohadmin user can't log in either
22:15 gmcharlt liz: hmm - one thing to try is doing a query to see if any branchcodes still exist in the staff patron records that don't have branches defined, then adding those codes back temporarily
22:15 gmcharlt i.e., select distinct branchcode from borrowers;
22:15 liz oh good idea
22:16 gmcharlt since kohaadmin can't log in for some reason, adding the missing branches can be done via SQL -> just insert into the branches table
22:16 liz interesting, the previous command brings back only ONE branchcode
22:17 liz >.< this is only the 2nd time he's buggered it in a month (LOL)
22:19 liz ok, tyvm for the ideas, I'm going to have to go as well, you'll probably hear more from me tomorrow.
22:19 liz thanks for the help
22:19 gmcharlt yw, good luck
09:37 mc hello koha people
09:38 chris hi mc
10:52 lloyd_ night
11:49 lloyd_ http://www.libcat.eoe.nhs.uk/g[…]-jackets-proxy.js - For anyone that will find it useful. I've written a howto and once the patch for customisable google-jackets.js location is submitted i'll put it on the wiki.

| Channels | #koha index | Today | | Search | Google Search | Plain-Text | plain, newest first | summary