IRC log for #koha, 2005-07-16

All times shown according to UTC.

Time S Nick Message
13:19 Ryanbisd cool, I will try that then....
13:19 Ryanbisd we are at a very early stage in the koha development here..... so we are learning as we go.
14:32 tim Does anyone here know how I could get it so I can use something like http://www.library.org/intranet instead of http://www.library.org:8080 ?
14:32 tim and mor important /opac or /search or something like that.
14:34 owen Tim, you might want to ask that on the Koha list, because I know it can be done.
14:35 owen Joshua (kados) has set it up for us.
14:35 tim I thought you were doing something like http://search.library.org
14:35 shaun yes - I have the answers, right here, but kados will give you a less hacky solution no doubt
14:36 tim I just loaded the apache docs.  Shoulda done that before I asked.
14:36 tim Of course that doesn't mean I'll figure it out. :)
14:48 thd tim: check the documentation for your version of Apache under the virtual hosts section.   Study the examples page.  You want a named based virtual host.
14:48 thd s/named/name
15:12 tim thd: So far I'm just finding how to put them on different ports like I already have and use virtual hosts, but that looks like it would require DNS changes for people to access it.
15:13 tim That's ok for the intranet, but we want people to be able to access the opac from outside and we don't manage our own DNS.
15:47 thd tim: Do you run your own server for the outside now?
15:47 tim Yeah
15:53 thd time: Do you not have access to the Apache configuration files on that server?
15:55 tim Yeah, I do.
15:56 tim Or at least I do on the one we'll be using when we start running Koha.
15:57 Ryanbisd tim: do you work for katipo, or are you implimenting koha also?
15:57 tim The current one is a turnkey thing with limeted acdess.
15:57 tim I'm implimentinng Koha in our library.
15:58 Ryanbisd cool...
15:58 Ryanbisd just wondering
15:58 Ryanbisd what stage are you at, and how long do you think the conversion will take total?
16:04 tim I don't have everything I need in yet and every time I test something I find a mistake in what I have so far.
16:05 thd tim: Does server that you are testing Koha on have domain name services directed at its IP address??
16:05 tim No.  It won't till it's closer to ready.
16:06 tim It's just an internal IP for now.
16:07 thd tim: you could register a domain name for testing purposes or just use the external IP address for the system with port forwarding.
16:12 tim If I'm going to bother registering a domain name I'll just add a virtual name and have them set that up on our DNS.
16:13 tim I'll probably do that anyway.  I was trying to think of ways to make it so I don't have to have someone else do something before I get it to work, but I'm sure the people who deal with it will handle things pretty quick.
16:14 thd tim: Who is 'them'?
16:14 tim Them are the ISP folks who do our DNS.
16:15 Ryanbisd item fields ALL items fields MUST :
16:15 Ryanbisd be mapped to the same tag,
16:15 Ryanbisd and they must all be in the 10 (items) tab
16:15 Ryanbisd
16:15 Ryanbisd itemtypes table empty the itemtypes table is empty. It must contain at least 1 value
16:15 Ryanbisd I keep getting this msg when i do marc check
16:15 rach do you have any itemtypes set up?
16:16 Ryanbisd no..
16:16 thd rach: no or that message would not be returning
16:16 Ryanbisd I dont think so
16:16 Ryanbisd is that in links koha Marc DB or in biblo framework?
16:17 thd Ryanbisd: You have to create at least one item type.  That should be part of the default install but is not.
16:17 rach to get rid of that particular message, I'd set one up in the itemtypes prevferences bit
16:17 rach so "neither"
16:17 Ryanbisd doh... stupid me..
16:18 Ryanbisd so like, I define Book... or Video
16:18 rach yep
16:18 Ryanbisd d@mn.. i feel like a retard
16:18 rach it should match the ones in your data :-)
16:18 rach sometimes it's the simple things that trip you up
16:19 Ryanbisd ad its this last one
16:19 Ryanbisd item fields ALL items fields MUST :
16:19 Ryanbisd be mapped to the same tag,
16:19 Ryanbisd and they must all be in the 10 (items) tab
16:19 Ryanbisd
16:20 rach yep - so you need one itemtype setup
16:20 rach then I think you need to make sure you're mapping the itemtpyes from any existing data to that field
16:20 Ryanbisd ohh... from my old marc records
16:20 Ryanbisd the ones im importing
16:21 rach yep
16:21 thd tim: I do not understand the details of your configuration well enough to give you a good suggestion about how to set up external access on your own.  I would not want to suggest something that might redirect services from your production system.
16:21 Ryanbisd see, I was told that my barcodes in the imported MR are at 852p
16:21 Ryanbisd and the call number is at 900a
16:23 tim thd: No problem.  I was just checking to see if anyone could suggest a way to make Koha load from a subdirectory instead of through a port or virtual host.
16:23 thd Ryanbisd: Remember that 900a does not exist in the default MARC framework you have to create that yourself unless you remap its value before importing.
16:26 thd tim: virtual hosts are used for subdirectory mappings for what you want, although, symlinks could have the same effect.
16:27 Ryanbisd rach, can you look at a sample of my marc records and see if my barcode is 852$p?
16:28 Ryanbisd www.bsaclaims.com/~ryan/marctest.001
16:29 thd tim: One thing that does not work in Apache 1.x is a virtual host where \d = some number you have something like \d\d\d\d.\d\d\d\d.\d\d\d\d.\d\d\d\d/opac
16:30 thd tim: mixing an IP address and a directory name does not work with Apache 1.x virtual hosts.
16:32 thd tim: If you only have an IP address for Apache to work with then you need to differentialte virtual hosts by port number in Apache 1.2
16:33 thd Ryanbisd: If your information is unchanged since yesterday, then your barcodes certainly were in 852 $p
16:33 tim thd: I'm using apache 2.
16:34 tim thd: I was thinking the virtusl host thing only did stuff like koha.library.org instead of www.library.org/koha/ like I was trying to do.
16:34 Ryanbisd how do you view them to see that information?
16:34 thd tim: I have not upgraded yet because of compatiblity issues with other software for any system I have.
16:34 tim thd: I'll look into that more.
16:35 thd tim: virtual hosts can do either.
16:35 Ryanbisd item fields ALL items fields MUST :
16:35 Ryanbisd be mapped to the same tag,
16:35 Ryanbisd and they must all be in the 10 (items) tab
16:35 Ryanbisd
16:35 Ryanbisd what "tab" is it talking about?
16:36 Ryanbisd and what is the 10 (items) field?
16:36 owen The tabs are in the MARC edit screen
16:36 owen The "10 (items)" refers to the item edit screen
16:37 thd Ryanbisd: To view the records yourself you could exprt them as text files with MARC Magician or try http://marcpm.sourceforge.net/[…]bin/converter.cgi .
16:37 Ryanbisd cool..
16:38 Ryanbisd owen: I have to define a value for every feild in the items tag?
16:39 thd tim: http://httpd.apache.org/docs-2[…]sts/examples.html
16:40 owen I'm not sure what you mean, Ryanbisd
16:40 Ryanbisd item fields ALL items fields MUST :
16:40 Ryanbisd be mapped to the same tag,
16:40 Ryanbisd and they must all be in the 10 (items) tab
16:40 Ryanbisd
16:41 Ryanbisd im getting that error, but I dont know what I am susposed to do.
16:41 Ryanbisd because my barcode is 852p
16:41 Ryanbisd but all the items are set to 900's
16:42 tim Ryanbisd: I had the same problem.  I had to move or copy infromation from field to field to get things set right.
16:43 Ryanbisd is there anyway that I can have my barcode set to a 900 field, but read from an 800?
16:43 Ryanbisd thereis... I think i remember seeing it..
16:46 Ryanbisd under marc subfield structure
16:46 Ryanbisd right?
16:46 tim Ryanbisd: do you mean all of your copy info is set to the 900's or the Koha defaults are set to the 900's?
16:46 Ryanbisd I have it set in the 900's now..
16:47 owen we have item fields mapped to 956: homebranch (b), holdingbranch (d), barcode (p), price (r), dateaccessioned (v), notforloan (y) and itemnumber (u)
16:47 Ryanbisd but, under the 952p I have it set to also look in 852p
16:47 tim So that's where the info is in your records?
16:47 Ryanbisd no. in my records its set to 852p
16:47 Ryanbisd http://lib-intra.burlesonisd.n[…]52&frameworkcode=
16:48 Ryanbisd username is kohaadmin
16:48 Ryanbisd pw is temp123
16:48 Ryanbisd go down to the section P... and see if that will work
16:49 thd Ryanbisd changed the barcode setting in items.barcode form 952 $p to 852 $p.
16:50 Ryanbisd i put it in the 'search also' box
16:50 Ryanbisd because I cant have an itemtype set to 800's
16:51 thd owen: Is there a setting that must also be changed in parallel for changing the default MARC field and subfield value for items.barcode?
16:51 owen I don't know
16:52 Ryanbisd i changed it back to 952
16:52 Ryanbisd if thats the only error I have.. .would it really matter?
16:56 thd Ryanbisd: I understand the error now
16:56 tim thd: Thanks for the help
16:57 thd tim:  :)
16:57 tim I just looked at the records that Ryanbisd showed us a while back and it looks like all of his item info is in 852.
16:59 tim I guess I'm outta here for the day.  Thanks again.
16:59 thd Ryanbisd, now disconnected, needs to change everything in the items table to 852
21:11 thd Genji: are you there?
00:26 Genji thd: yup
01:08 Genji gone now though.
08:07 kados morning owen
08:07 owen Hi kados
08:08 owen I've been working on the programmer templates
08:08 kados sweet
08:08 owen That's what's on 101 right now
08:09 owen I've got most of the member pages done
08:09 kados very nice
08:09 owen I've got a stripped-down stylesheet attached, just to make it easier on the eyes.
08:10 kados yep ... looks very nice
08:10 owen It's a tough process.  It's sometimes hard to decide what should stay in and what shouldn't
08:10 kados I can imagine ;-)
08:12 kados I think we need to adopt a syntax for searches
08:22 owen kados: like a standard contruction syntax that will define how ANDs and ORs etc. translate to SQL?
08:23 kados yes to the first no to the second
08:23 kados I've been looking at CQL a bit more
08:23 kados http://www.loc.gov/z3950/agency/zing/cql/
08:23 kados I think folks have been intimidated by it
08:23 kados because the site is fairly academic
08:24 kados but when you dive into it it's fairly straightforward
08:24 owen The last time it was discussed people were distracted by the idea that the patrons would have to learn it to be able to do a search.
08:24 kados yea
08:24 kados but that's not actually true
08:24 kados patron's won't have a hard time doing 'author=neal stephenson'
08:24 kados or 'title=cryptonomicon'
08:25 kados and of course CQL will accept straight keywords with no qualifiers
08:25 kados the thing is
08:26 kados with any advanced search syntax there is some learning curve
08:26 kados like how many folks know that they can do 'inurl:liblime.com koha' in a google search?
08:26 kados or 'intitle:koha site:liblime.com'
08:28 kados the nice thing about CQL is the learning curve for the actual queries isn't that steep
08:29 kados I do agree with MJ that the use of : as a marker is nice
08:30 kados as is - and +
08:30 kados so maybe our objective is to support multiple syntax types for input
08:31 kados and use CQL in the background for the server interactions
08:31 kados I'm still not clear on how to use CQL with Zebra
08:32 kados to perform the actual queries on the index
08:41 owen I think we need to look at how to support Google-style queries, with quotes deliniating phrases at the very least
09:13 hdl hi kados
09:14 hdl Have I missed sthg ?
09:14 kados hi hdl
09:14 kados don't think so ;-)
09:14 hdl ok.
09:14 kados just me musing about query syntax ;-)
09:14 hdl I have a question :
09:14 kados sure
09:14 hdl We are trying to make a sort of base modifications watch.
09:15 kados ok ... what does that mean exactly
09:16 hdl First question is : Where do things should be commited ? i.e. In C4::Miantainance, as I previously did, or in C4::Koha, or else ???
09:16 hdl Second question that goes along is :
09:16 hdl Do we want these operations to be atomic ?
09:17 hdl That is, as soon as You have added a biblio.
09:17 hdl you also added a logentry
09:17 hdl Or do you prefer the ewo operations to be disticnt.
09:17 kados hmmm ...
09:17 hdl disctinct.
09:18 hdl s/ewo/two/
09:18 kados so what you're proposing is a way to keep track of modifications (adding and deleting) of records in Koha?
09:18 hdl Yes.
09:19 hdl In a new table : action_logs
09:19 kados I'm not sure about which module it should be coded in ..
09:19 kados maybe we need a new module
09:19 kados that is used specifically for monitoring the system
09:19 hdl timestamp, userid, module, action and parameters are logged
09:20 hdl Which name in this case.
09:20 kados hmmm ... that sounds very similar to the statistics stuff
09:20 kados what is this functionality called in SQL?
09:20 hdl Yes and no. Its purpose is not stats, but monitor.
09:21 hdl I insert a new line in the database.
09:21 hdl Maybe Mysql has a log facility.
09:21 kados inright
09:21 hdl But I don't know...
09:22 hdl And We should be DBM independant.
09:22 kados how about Log.pm?
09:22 hdl ok.
09:22 hdl I take it.
09:22 kados and we can put the Stats.pm stuff in there eventually
09:23 kados maybe? ... we can discuss this later
09:23 hdl ok. I am doing a Log.pm. and use it.
09:23 kados and by 'atomic' do you mean 'automatic;?
09:23 kados I think there should be a system preference for turning the Log on and off
09:24 kados because it will likely impact performance
09:24 kados hdl: did you see my recent post about paul and your plans for multiple branch support?
09:26 hdl No I mean : as soon as you have a database mod, you are ensured that dataentry is done to the table.
09:26 hdl i.e. same operation logging and adding/editing/deleting.... Which would lead to log.pm called in Biblio.pm for instance.
09:26 hdl kados : yes I read it.
09:27 kados hdl: I still don't understand the question
09:28 kados hdl: I think that Log.pm (if I understand what it is) should only be used if the library wants to use it
09:28 kados hdl: because it may slow down adding/deleting things at NPL ;-)
09:28 kados hdl: (though it would be useful to see _who_ deleted something they shouhldn't have ;-))
09:28 hdl should I hide Log management in C4 modules, or leave it to the top level.
09:28 hdl But what you said answered my question.
10:29 thd kados: are you still present?
10:30 kados thd: yep
10:30 kados thd: what's up
10:31 thd CQL or whatever search syntax is used should be wrapped in a form
10:31 kados thd: hmmm ... maybe ...
10:32 kados thd: at least for the advanced page
10:32 thd Therefore, the user does not have to learn the syntax, only the form
10:32 kados thd: but the direction that most engines are moving in
10:32 kados is to allow direct syntax entry into a single form
10:32 thd Both methods should be provided for
10:33 kados yep
10:33 kados thus we have 'simple' and 'advanced' search pages
10:33 kados but the idea is
10:33 kados we're taking our cue about future directions in searching
10:34 kados from web search engines
10:34 kados like google
10:34 thd A form allows you to change the underlying syntax without bothering the user about what the syntax actually is
10:34 kados yep ... so we have both
10:35 thd Google does not necessarily have the ideal solution
10:35 kados I agree
10:35 kados I'm thinking we use CQL for sure
10:35 kados but also allow author:neal stephenson
10:35 kados and "once upon a time" phrase searching
10:36 kados and I like '+stephenson -neal'
10:36 thd CQL is designed for SRW which has very few targets currently relative to Z39.50
10:37 kados well I'm going to propose we use Zebra's SRW functionality to expose our records for that protocol
10:37 kados in addition to Z39.50
10:37 kados CQL may not be widely adopted
10:37 thd Very forward thinking :)
10:37 kados but that's the way of things
10:38 kados it will take a few years
10:38 kados I'd rather be way ahead of the pack
10:38 kados and it's not like it's hard ... index data has already done the hard part
10:39 thd I was thinking more of the issue of converting queries against the catalgoue to queries against the other outside resources which are mostly z39.50
10:39 kados yep ... we can easily do that with yaz-proxy
10:40 thd kados: I love mapping, although, their is a bit of unpleasantness about most such tasks
10:40 thd s/their/there
10:41 kados agreed
10:41 thd mapping is the key to interoperable systems when you cannot force everyone to adopt one standard
10:41 kados yep ... which there's no way we can!
10:42 thd Why can't you work around the logging performance issue with batch processing?
10:44 thd I saw a suggestion about batch processing cataloguing on the koha or koha-dev list recently
10:44 kados thd: I think batch processing is a good idea
10:45 kados thd: maybe for logging too ... but either way you still have some performance hit
10:45 thd kados: Do you mean an additional performance hit for logging even if batched?
10:48 thd owen: are you present?
10:48 owen yes
10:49 thd owen: the ideal way to address the disposable serials issue is with a 772 linking field
10:50 thd I imagine there is some functionality one could borrow from the UNIMARC side
10:51 owen Can you elaborate?
10:52 thd A well designed 77X plugin should copy required information from a linked record and then you would only need to modify a few subfields in 3 or 4 fields
10:54 thd So you would get all the bibliographic information that was the same for every issue of your serial copied from the other record and you would add the issue and holdings data.
10:55 thd This would be a MARC21 compliant solution :)
11:01 thd The linking fields are 76X - 79X.  I am trying to fetch an example for you but my system is thrashing a bit.
11:03 thd 245 04$aThe Post boy.$nNumb. 2436, from Thursday December 21 to Saturday December 23, 1710.
11:03 thd 772 1#$7unas$aPost boy (London, England)$w(OCoLC)1234567
11:04 thd http://www.itsmarc.com/crs/Bib1390.htm
11:06 thd owen: In the example above, 772 $w is the control number for the linked record on your system and 245 $n has been added for the issue.
11:08 owen Sounds great in theory :)
11:08 thd owen: 245 $a , 772 $a,  and much else could have been copied by a plugin from the linked record
11:12 thd owen: I know there may be a more efficient way to solve your immediate problem with less effort but this standards compliant method would be reusable even for permanent records and other institutions where people go crazy over MARC standards :)
11:23 owen Something to be taken up with the developers...

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