Time Nick Message 11:42 kados bbs even :-) 11:42 kados bs 11:41 kados hdl: ok, thx 11:41 hdl I will come back. 11:41 hdl kados: I'm off for dinner. 11:37 kados yea, that parentheses thing is annoying 11:36 kados cool, thx 11:36 owen By the way, I found some new (to me) bugs on zoomopac and added some notes to the NPL wiki about them 11:33 owen Yes 11:33 kados owen: does that make sense? 11:32 kados I'll add a note and dependency on 1137 to 1146 11:31 kados if I can't create a new basket, I can't really test 1137 :-) 11:31 kados i suspect that bug 1146 is to blame here 11:31 kados actually ... 11:30 kados let me find a link 11:30 owen I'm not sure what you're seeing 11:30 owen The basket screen looks the same to me in npl and default 11:30 owen kados: about Bug 1137... 11:09 kados it still opens up an existing one when I click on 'order' 11:09 kados even if all the baskets are closed 11:07 kados I guess I just file a bug and if it's invalid, paul or hdl will comment on it 11:06 kados if we're right, it's another bug with acquisitions :-) 11:06 kados yea, me too 11:06 owen I would expect the order link to create a whole new empty order, regardless of how many other open orders there are 11:06 kados paul: or you if you're around? 11:06 kados hdl: can you shed light on this question 11:06 owen I would think not 11:05 kados or ... is it understood that you have one order at a time per supplier? 11:05 kados one is open, the others are closed 11:04 kados you will see several baskets for 'Josh Book Supplies' 11:04 kados (just click on 'OK' for the supplier name search 11:04 kados http://koha.smfpl.org/cgi-bin/koha/acqui/acqui-home.pl 11:04 kados even in default templates 11:03 kados but it seems like the behavior is to open an existing order form (basket I guess) 11:03 kados then you would add items to that new order 11:03 kados it was my understanding that when you search for a vendor, then click 'add order' it took you to a page with a blank order form 11:02 kados quick question about acquisitions ordering 11:02 kados owen: got a sec? 10:57 kados oosp 10:57 kados hdl: 10:57 kados hdl: I updated bug 1125 with the new information 10:44 kados but maybe I missunderstand this description 10:43 kados (since I checked in 6 issues?) 10:43 kados 2, 3, 4 ... I think it should be 4, 5, 6 ... right? 10:43 kados http://opac.smfpl.org/cgi-bin/koha/opac-detail.pl?bib=59614 10:43 kados it is viewed here for instance: 10:42 kados do you see that? 10:42 hdl In subscription-detail, it takes the 5 "latest" ones. 10:42 kados it says 'latest issues' but they are not the 'latest' but the 'first' 10:42 hdl yes. 10:42 kados in opac-detail.pl? 10:41 hdl only 3 of them are displayed. 10:41 kados hdl: does that answer your question? 10:41 kados in opac-serial-issues.pl? it seems to be fixed 10:40 kados so in opac-detail.pl, serials listed are not 'most recent' but are 'first' 10:40 kados hdl: also css template 10:40 kados hdl: http://opac.smfpl.org/cgi-bin/koha/opac-detail.pl?bib=59614 10:39 kados hdl: for opac 10:39 kados hdl: that's css template 10:39 hdl in your html code of the page. 10:39 hdl <a href="opac-detail.pl?bib=59614" class="button catalogue">Back to biblio</a> 10:38 hdl <h1 class="catalogue">Subscription information for American girl.</h1> 10:38 kados I don't see it ... maybe npl templates are missing a link? 10:38 kados http://koha.smfpl.org/cgi-bin/koha/bull/subscription-detail.pl?subscriptionid=1 10:38 hdl (left column.) 10:38 kados left tab? 10:37 hdl return to biblio... I bet it is behind your lefttab. 10:37 kados hmmm 10:37 hdl I should have change the name. 10:37 hdl I hope ppl will not enter serials up to 1900. 10:36 hdl 1900 : is the name of the no Publication date. 10:36 kados hdl: the opac-serial-issues.pl seems to have them all ... even if 1900 is a strange tag 10:36 kados hdl: only the biblio detail page is missing subscription listings 10:35 kados hdl: what is the '1900' tab? 10:35 kados hdl: http://opac.smfpl.org/cgi-bin/koha/opac-serial-issues.pl?biblionumber=59614&selectview=full 10:32 kados so maybe it's just a search prob? or display prob? 10:32 kados select * from subscriptionhistory also has Vol 12 No 1. listed 10:31 kados has issues 1-7 listed 10:31 kados select * from serial: 10:30 kados i will check the db 10:30 hdl Is it a problem in serial table or in the search ? 10:30 kados hdl: (but it displays in serial-issues.pl) 10:29 kados (also note that it seems that Vol 12 No 1 was deleted ) 10:28 kados so it is not the 'latest checked in' but the 'first checked in' 10:28 kados Vol 12 No 4 08/08/2006 Arrived 10:28 kados Vol 12 No 3 08/08/2006 Arrived 10:28 kados Vol 12 No 2 08/08/2006 Arrived 10:28 kados hdl: after checking in 6 issues it now shows: 10:28 kados hdl: but to answer your original question ... is bug 1140 fixed ... almost 10:27 kados so Vol 12 No 1 is missing from there 10:27 kados Vol 12 No 7 Waited 10:27 kados Vol 12 No 2 08/08/2006 Arrived 10:27 kados Vol 12 No 3 08/08/2006 Arrived 10:27 kados Vol 12 No 4 08/08/2006 Arrived 10:27 kados Vol 12 No 5 08/08/2006 Arrived 10:27 kados Vol 12 No 6 08/08/2006 Arrived 10:27 kados hdl: subscription-detail.pl now shows: 10:27 kados hdl: and received 6 issues 10:27 kados hdl: i created a new subscription 10:26 kados hdl: I discovered a new bug 10:26 qiqo with the cvs 10:26 qiqo i cant seem to find the opac module 10:19 kados I will update and test 10:19 kados ok 10:19 hdl Yes I mistook numbers.. 10:19 kados hdl: I don't see a commit for 1140 ... but for 1136 10:19 hdl opac serials. 10:19 kados hdl: 1140? 10:18 hdl kados : see if it is OK now ? 10:17 slef http://www.ukphoneinfo.com/section/tci/locator.shtml if anyone ever wants a UK one 10:17 slef www.itu.int seems down 10:17 slef Burgwork: thanks will bookmark 10:16 Burgwork slef, http://decoder.americom.com/main.html 10:16 hdl sorry 1140 10:16 hdl kados : bug 1240 10:15 Burgwork morning kados 10:15 owen Google suggests Nigeria 10:15 Burgwork slef, nigeria or ohio (within NA) 10:14 kados not me 10:14 slef anyone know where telephone numbers starting +234 are? 10:10 kados I can look at it if you are 10:10 kados hdl: are you working on a bug? if so, which one? 10:09 qiqo ok 10:09 qiqo still 2_2? 10:09 kados qiqo: rel_2_2 10:09 hdl yes. 10:09 qiqo cvs -z3 -d:pserver:anonymous@cvs.savannah.nongnu.org:/sources/koha co -r rel_2_2 koha 10:08 qiqo yes.. so what release are we talking about now 10:08 kados hdl: are you still around? 10:08 kados it has a step by step guide for how to use cvs 10:08 kados qiqo: there is a document called 'Updating Koha ' 10:07 qiqo yes 10:07 kados qiqo: did you look on kohadocs.org? 10:07 qiqo updating koha? 10:04 kados check kohadocs.org for the 'Updating Koha' document 10:03 kados that's running stock cvs 10:03 qiqo yes 10:03 qiqo how do i install that? 10:03 kados did you see the demo? 10:03 kados so maybe the gaylord _is_ the right choice for you 10:03 qiqo nope i dont 10:02 kados qiqo: do you have pre-printed spine labels? 10:01 slef print onto paper and then sticky-film them to the books? 10:01 qiqo the manually made labels 10:00 slef qiqo: what labels do you use? 10:00 kados we rely on sponsored development for features 10:00 kados this is an open source project 09:59 qiqo but its taking up time,, i need to label my collection.. 09:59 kados but for now, only the gaylord is supported 09:59 kados qiqo: with measurements based on what you need 09:59 qiqo :( 09:59 kados qiqo: the new system we speced out would allow you to define label types of your own 09:58 kados qiqo: then click on 'add labels' and then finally 'generate PDF' 09:58 kados qiqo: or just: http://koha.liblime.com/cgi-bin/koha/label-home.pl 09:58 kados qiqo: http://koha.liblime.com/cgi-bin/koha/barcodes/label-manager.pl?itemtype=1&isbn=1&dcn=1&classif=1&itemcallnumber=1&papertype=gaylord-8511&barcodetype=Code39&printingtype=both&startrow=1&guidebox=1&op=save_conf 09:57 kados spine labels and barcodes on the same line 09:57 kados it has: 09:57 kados gaylord is a label type 09:57 qiqo so i will install the gaylord? 09:57 kados qiqo: koha.liblime.com 09:57 kados qiqo: try it on the demo first to see how it works: 09:57 kados qiqo: that will only work if you have gaylord 8851 barcodes 09:56 qiqo im gonna try the cvs thing first 09:56 kados slef: :-) 09:56 kados it could be done very quickly if we had a sponsor 09:56 slef tomorrow is koha day here at towers 09:56 qiqo armmmm.. 09:56 kados hmmm ... we don't even have a timeline 09:56 slef I may have a sponsor for working on barcodes, but I'm snowed. :-( 09:55 kados i can check our current schedule to see when it's supposed to be done 09:55 qiqo ermm.. ok 09:55 kados but it wouldn't be a priority without sponsorship 09:55 qiqo really? how long shall it take? 09:55 kados we could finish the development quite quickly and you could install it 09:55 qiqo so you think the cvs thing wont fix my problem? 09:55 qiqo argghh// 09:55 kados hmmm 09:55 qiqo i badly need the feature right now.. 09:54 kados you will be able to use the barcode system in 2.4 and 3.0 09:54 qiqo ever? unless i look for a sponsor 09:54 kados Koha 2.2.6 will be different -- no advertised features that don't work 09:54 qiqo so i wont be able to use the barcode system? 09:54 kados I'm sorry :( 09:54 kados unless you want to sponsor development of this feature 09:53 kados unless you're using gaylord 8851, I think you'll need to print barcodes outside of Koha 09:53 qiqo what shall i do? please tell me.. 09:53 kados no 09:53 kados but will be in 2.4 and 3.0 09:53 qiqo ei.. will i just overwrite the files from cvs/barcode to the current/barcode? 09:53 kados we're working on a more flexible system that won't appear in 2.2.6 09:52 kados it's the gaylord 8851 09:52 kados the new system currently requires a specific barcode/spine label 09:52 qiqo but dont know what to do now 09:52 qiqo but i already downloaded the cvs 09:52 kados one thing ... 09:52 qiqo sorry im not really that familiar with cvs 09:52 kados yea, cvs definitely does 09:52 qiqo so what shall I do 09:52 kados yea, IIRC the install script for 2.2.6rc2 doesn't create the table properly for the new barcode system 09:51 qiqo you think cvs got the new barcode system? 09:51 qiqo cus i already install 2.2.6rc2 and i also got the same problem 09:51 kados and it's working well for them 09:51 kados I have three libraries using the new barcodes system in production 09:51 kados it should help 09:51 qiqo you think that''ll fix my problem? 09:51 kados yep, that will work 09:50 qiqo how bout updating 2.2.5's barcode via cvs? 09:50 kados nope 09:50 qiqo you thing 2.2.4 is ok with barcodes? 09:50 kados 2.2.6 should be out very soon 09:50 qiqo when will that be out? 09:50 qiqo so what version do you suggest? 09:50 kados that works very well 09:50 kados there is a new system that will be in 2.2.6 09:49 qiqo ahh man... 09:49 kados it doesn't work 09:49 kados barcodes on 2.2.5 is broken badly 09:49 qiqo barcode system? 09:49 qiqo 2.2.5 09:49 kados qiqo: what version of Koha, etc.? 09:49 kados qiqo: what barcodes system are you using? 09:49 qiqo anybody who could help me with the barcode printing stuff? 09:46 qiqo ermm.. 09:39 kados so we can close bug 1136 I think 09:39 kados (before) 09:39 kados (maybe the problem was simply that I couldn't acquisition anything in the first place) 09:38 kados hdl: no ... I think paul's commit to Acquisitions must have fixed it 09:38 hdl yes. 09:38 hdl kados : have you change sthg in histsearch ? 09:38 kados maybe it shows received and non-received items in the history? 09:38 kados hdl: some columns are missing, but i can detail them later 09:37 kados hdl: ok, histsearch seems to work now 09:36 qiqo what may seem to be the problem? 09:35 qiqo when i generate the barcodes i get other numbers other than what i assigned to a monograph 09:34 qiqo i would just want to ask about printing barcodes 09:34 kados hdl: first i try to acquisition something ... can't get it to work :( 09:34 hdl to see if histsearch can provide you with data. 09:34 hdl kados : about 1136 : can you try a hsit search commenting lines 790 and 821 in C4/Acquisition.pm 09:33 qiqo ei anybody awake? 09:27 hdl yes. 09:24 kados 1136? 09:24 kados hdl: which bug number? 09:19 hdl about acquisition history bug. I cannot reporduce it myself. 09:14 kados hdl: do you think Lyon will notice the bugs that exist for serials in rel_2_2? 09:10 hdl then i did not retest EVERY thing. 09:10 hdl But it is long ago. And I had to modify things so taht dates were well displayed. 09:09 hdl I had tested. 09:09 kados you haven't tested? 09:08 hdl I hope so. 09:08 kados items are created? 09:08 kados so manual issues work for Lyon? 09:07 hdl It was well aranged so that items were created. 09:06 kados what do you mean it was 'OK'? 09:06 hdl Yes. 09:06 kados their base? you mean the database? 09:06 hdl But their base was OK. 09:06 hdl There are the same bugs as you detected them. 09:05 kados hdl: can you update cvs to be bug-free like Lyon's version is? 09:04 hdl I coded serials itemization for them. 09:04 hdl not yet. 09:00 kados is the cvs version the same as Lyon's version? 09:00 hdl All went wel. 09:00 hdl yes. 09:00 kados did they perform an acceptance test on it before they went on vacation? 08:59 hdl I set it up on late june. 08:59 kados before vacation, they were using it? 08:59 kados :-) 08:58 hdl They are on vacation. 08:58 kados so they didn't notice all the bugs? 08:57 hdl IEP Lyon is now using it. 08:57 hdl when you unsubscribe a list, you keep tracks of the emails you received. Same for items received. 08:56 kados is it sponsored by one of your libraries? 08:56 kados I worry that this serials module is not ready for a stable system 08:56 kados hmmm 08:55 hdl You keep the items. 08:55 hdl But if you delete a subscription. 08:55 hdl Unfortunately. 08:55 hdl not yet. need a new column. 08:55 kados wow ... so if you delete a subscription, it doesn't delete the items? 08:54 kados itemnumber is not stored in the serials table? 08:54 kados hmmm 08:54 hdl And we have no other link between items and serial but serialseq 08:53 kados that's ok by me 08:53 kados ahh, ok 08:53 hdl the information is both in items.itmenotes and serial.serialseq 08:52 kados hmmm 08:52 kados duplicated? 08:52 kados and now it's in item.notes 08:52 hdl It is duplicated in items.itemnotes. 08:52 kados ahh 08:52 hdl you have to know which marc field to put serialseq 08:52 kados not items table? 08:52 kados serialseq is in the serials table, right? 08:51 kados does it have to be? 08:51 kados right 08:50 hdl some information (serialseq for instance) you input for an issue ppl donot want to input it twice for an item. 08:49 kados 2. working the subscription magic 08:49 kados 1. creating an item (using item column) 08:48 kados saving an issue would mean: 08:48 kados or just the serials one 08:48 kados when an item was checked in a library can fill out both columns 08:48 kados 2. for items information 08:47 kados 1. for serials information 08:47 hdl A second problem would be to keep up the link with items. 08:47 kados what if we had two columns: 08:47 kados I don't understand completely 08:47 hdl But since we rely on koha items table, all the item subfields cannot be used. (I supposed when I coded that Katipo would use that code.) 08:47 kados hmmm 08:46 hdl kados : the problem is that relying on items table is not satisfactory for ppl. since they would like to input all the items data for serials on the same page. 08:44 kados otherwise, itemization won't be possible for them as they require all items to have all fields (in some cases) and some fields, like barcode, rely on a specific plugin 08:44 kados 2. plugins to work 08:43 kados 1. all item.* data avaialble 08:43 kados hdl: something else to consider: I think for serials itemization to work for my clients, they will need: 08:42 kados (I'm not sure about next meeting, maybe tomorrow?) 08:42 kados is paul around now? 08:42 kados as it doesn't affect them :-) 08:41 kados they are ok with any change you make 08:41 hdl Could it be on agenda ? 08:41 kados katipo has their own serials module 08:41 hdl When is the next meeting ? 08:41 hdl we may have to refer to katipo and paul as well. 08:41 kados yep 08:41 kados (more than ok, great! :-)) 08:41 hdl But THIS is a major change. 08:41 kados that would be ok 08:40 hdl what if they were gathered in a expected issues tab ? 08:40 kados 2. for all issues 08:39 kados 1. for current expected/late issues 08:39 hdl (table would be a div) panel. 08:39 kados but maybe we should have two pages: 08:39 kados it's a good idea 08:39 hdl or the next expected one. 08:39 hdl And you would have a table to add one. 08:38 hdl you could clik on an issue to modify it. 08:38 kados hmmm 08:38 hdl It could allYou could have all issue on the page. 08:38 kados how would it work? 08:38 hdl ? 08:38 hdl what do YOU think about designing a true serial management based on full-serial-issue 08:36 kados what do you think about my suggestion re:1125? 08:36 hdl yes. 08:35 dewey there is a minor diff in <div>s, that I missed 08:35 kados hdl: still there? 08:32 kados hdl: but for 1125, I think we need to create a syspref to show/hide the last 5 issues 08:32 kados hdl: bug 1124 is fixed i think 08:10 kados hdl: it is linked 08:09 hdl That they HAVE to have an items.itmenotes linked to a marc field. 08:09 hdl kados : Have you understood what I told you for the bug filed by your client ? 08:08 hdl kados : Is it ok for 1124 and 1125 ? Can I close them for good ? 08:06 slef I conclude that I dreamt it. 08:01 kados not me ... 08:01 slef someone made a passing reference to mailing again about bug tracking... was it someone here and did they mail it? 08:01 kados just to keep track of bugs :-) 08:01 kados if it's fixed today we can close it 08:00 kados for 'manual issue doesn't save item information to record' :-) 08:00 hdl for ? 08:00 kados hdl: i will file a new bug report 07:59 hdl (will fix it now.) 07:59 hdl THAT IS a bug. 07:58 hdl ok. 07:57 kados manual issue doesn't save item information to record 07:57 kados I just tested in default templates 07:56 kados afaik, this works except for manual issues 07:55 kados ahh ... so you mean it will save it in the MARC field linked to items.barcode 07:55 hdl I say taht if you have a Koha link defined for items.barcode, a Koha link for items.homebranch and items.holding, a Koha link for items.itemnotes, it should work. 07:53 kados (I have plugin defined for items.barcode, but it doesn't appear in the serials check-in) 07:53 hdl taht is : barcode, location, branch, and status. 07:52 kados so you mean that if I have a plugin defined for the items.barcode MARC field (in biblio framework) it should work in serials check in? 07:52 hdl yes. 07:52 kados (you mean serials itemization?) 07:51 hdl It may go far further. But we can now manage the minimum. 07:50 hdl plugin for itemization. 07:50 hdl It's ok for me. 07:50 hdl So if anyone fells like. 07:50 kados ahh ... 07:50 hdl plugin for acquisition was not asked and not coded. 07:49 kados (OK) 07:49 hdl ) 07:49 hdl (let me go. 07:49 hdl plugin for acquisition 07:49 kados ok ... I will begin again 07:48 kados sorry ... 07:48 hdl does not seem. 07:48 kados I'm calm 07:48 kados :-) 07:48 hdl calm down. 07:48 hdl pls. 07:47 hdl No. 07:47 kados hdl: does that make sense now? 07:47 kados plus they don't like the display of the last 5 issues 07:47 kados manual issues don't save barcodes, etc. 07:46 kados no warnings for duplicate barcodes 07:46 kados (and no plugin for it of course) 07:46 kados the acquisition date isn't available as a field 07:45 kados for instance, a plugin for their barcode isn't working 07:45 kados what they have a problem with is the other items fields aren't working 07:45 kados that is not the bug they reported 07:45 kados they are fine with the number in itemnotes 07:45 hdl If there are no itemnotes... 07:45 kados that is fine 07:45 hdl I took itemnotes field. 07:44 hdl And to store the number information, I HAVE to have a field. 07:44 hdl I build it from the framework. 07:44 hdl But I store a whole marcrecord. 07:44 kados so it's different 07:43 hdl Yes. 07:43 kados ? 07:43 kados no ... barcode is stored in items.barcode, not items.note 07:43 hdl They say :the problem I was speaking about. 07:43 kados barcode 07:43 hdl in, but doesn't update our record. 07:43 hdl barcode information etc. to our record. Once again, it simply checks the issue 07:43 hdl it is not saving the 07:43 kados 3. Serials is not adding a date to the record (acquisition date) 07:42 kados 2. no barcode checking (for duplicates) 07:42 kados 1. plugins don't work in the serials item editor 07:42 kados what is not fine is: 07:42 kados that is fine 07:42 kados I agree that Vol. No is stored in itemnotes 07:41 kados but that is not the bug they are reporting 07:41 hdl s/MUST/HAVE TO/ 07:41 hdl For serials itemization they MUST. 07:41 kados they don't use 'itemnotes' for bibliographic data 07:40 hdl But without peering into their code, i cannot confirm. 07:40 kados ?? 07:40 hdl So I think that THEIR problem comes from that. 07:39 kados yes 07:39 hdl (I assume that they use Note fields in bibliographic data.) 07:38 hdl OR they may have linked it to more than one field. 07:38 kados ?? what do they use? 07:38 hdl Surely, most of the time, ppl donot use items.note. 07:38 hdl BUT.... 07:38 hdl since items.note SHOULD BE the place where the issue number is stored. 07:37 kados ok 07:37 kados hdl: but in serials, item editing doesn't use frameworks in rel_2_2 in either default or npl templates 07:37 hdl It looks in the framework which marc field is linked to items.barcode, AND which marc field is linked to items.note 07:36 kados hdl: and the frameworks work fine in item editing 07:36 kados hdl: of course, they already have frameworks for items set up 07:36 hdl What they have to know IS taht serialsitemization relies on frameworks. 07:35 hdl OK ? 07:35 hdl They tell taht it doesnot update record. 07:35 kados ok, one bug at a time 07:35 hdl let us focus on this bug. 07:35 hdl pls donot mix all things. 07:34 kados our plugins aren't working, for barcodes for instance 07:34 kados hmmm 07:34 kados for plugins, etc.? 07:34 hdl YES IT DOES. 07:34 kados so it doesn't consult the marc frameworks? 07:33 hdl It takes the items.barcode marc field, the items.note marc field and then fills them. 07:33 kados same with me 07:33 hdl is relying on framework as such. 07:32 hdl My code 07:32 hdl Have you tried to know how the code work ? Did you investigate how their frameworks look like ? 07:31 hdl They file bugs from itemization. 07:31 kados no 07:31 hdl Did they use serialsitemization before ? 07:31 kados hdl: how is it a preference or marc prob? 07:30 kados hdl: serials were working fine ... I upgraded them to rel_2-2, now they report bugs 07:30 kados ?? 07:30 hdl That must be a preference or marc problem 07:29 kados and acquisition date is filled? 07:29 hdl on production 07:29 hdl For me Yes. 07:28 kados hdl: s/and/in/ 07:28 kados hdl: do plugins work with item adding and serials? 07:27 kados hdl: and I upgraded them to rel_2_2 07:27 kados hdl: they were running on the previous version of serials 07:27 kados hdl: a user contributed this bug report 07:27 kados hdl: http://bugs.koha.org/cgi-bin/bugzilla/show_bug.cgi?id=1142 07:27 hdl you can esayliy update the data so that receivedate comes into publisheddate. 07:26 hdl yes. 07:26 kados yes, I upgraded using upgradedatabase, and the new column was added I believe 07:26 kados it should be an option in the series creation screen, 'calculate by pub date or by receive date' 07:25 hdl It also add a new column into the database !!!! 07:24 hdl If you like, I can copy the dates into both fields. 07:24 kados it completely breaks any previous serials installation! 07:24 kados but that's a major function change 07:24 kados sigh 07:24 hdl So this is the one that is now calculated. 07:24 kados so as I described, currently, when you receive issues, the publication date is auto-calculated, not the receivedate 07:23 hdl So this is the one that is now calcultaed. 07:23 hdl publication date is the most important for users. 07:23 kados ok 07:23 hdl receivedate is two. 07:23 hdl publication date is one. 07:23 hdl kados : now, there are two dates. 07:22 kados issue date means publication date? 07:22 hdl why. 07:22 kados yes, I think that's a problem 07:22 hdl But we may discuss. 07:22 kados hmmm 07:22 kados hdl: are you working on serials bugs? 07:22 hdl kados: now, dateexpected is not filled any longer since what is supposed to be calculated is the next issue date rather than the next reception. 07:20 hdl kados : i am here 07:16 osmoze hello #koha 06:24 kados hdl: are you there? 06:19 kados that may be my fault 06:19 kados wait ... 06:19 kados so both Vol and No are incremented (the rules are set up properly) 06:19 kados ,Vol 1 No 1,Vol 2 No 2,Vol 3 No 3 06:19 kados hdl: subscriptions are now listed like: 06:18 kados hdl: and another new prob: 06:17 kados hdl: that is, the 'published on' date is filled with the expected date 06:17 kados hdl: 'date expected' is not filled, but 'published on' is instead 06:17 kados hdl: I have a new problem, along with the old ones 06:11 kados hdl: should I test your bugfixes to 1024/1025? 06:09 kados hdl: I'm here 04:31 dewey kados is becoming a true Perl Monger... 04:31 hdl kados ? 04:31 toins je pense qu'il sera la cet après midi. 04:31 hdl OK. 04:30 toins non, il est allé faire des courses 04:30 hdl toins : paul est avec toi ? 02:33 hdl hi 02:23 toins hi all 00:16 thd kados: are you still up? 17:54 chris could be a just a script that parses, and redirect to the script that does the search and displays the results 17:53 kados yea, that might work 17:53 kados sure 17:52 chris just thinking aloud 17:52 kados hmmm 17:52 chris then deal with it the same 17:52 chris first parse, convert it to the same syntax as the advanced search expects 17:52 chris for the simple search we do a double parse 17:52 chris i think what im thinking is 17:51 chris could that not be in the form of one input per term tho? 17:51 kados hmmm 17:51 kados ie, facets, limits, resorting, query history, etc. 17:50 kados but the problem is, the results page for both simple and advanced searches has to have the same options 17:50 kados well I like that idea 17:50 kados hmmm ... 17:50 chris and for the other one, like the advanced search page, where you have an input for author, one for subject etc .. treat each input as a term, not a query 17:49 chris if you are allowing a free text type in the query one, handle that with one cgi script 17:48 kados how so? 17:47 chris well, i think the answer might be .. restrict the syntax the form/cgi can hand data back as 17:47 chris k 17:46 kados q=su=cats&q=au=tom 17:46 kados or you could go: 17:46 kados q=su=cats and au=tom 17:46 kados so you can do something like: 17:46 kados right now, q can be a valid CCL query 17:46 chris right 17:45 kados as is sort by 17:45 kados as is server 17:45 kados as is op 17:45 kados q is repeatable 17:45 kados search?q=query&op=operator&offset=&count=&sort_by=&server 17:44 kados is something that looks like this: 17:43 kados well what I've got 17:43 chris no worries then .. a openurl resolver isnt hard to write 17:42 kados I doubt it does 17:42 chris not sure, i havent read about it properly, would be good if it did 17:42 kados does openurl have a fully-described query syntax? 17:42 kados hmmm 17:42 kados this is frustrating though 17:41 chris could we make kohaquerysyntax = openurl 17:41 chris right 17:41 kados etc. 17:41 kados kohaquerysyntax -> pqf 17:41 chris without havning to hack at our ccl parser 17:41 kados 1. cgi -> kohaquerysyntax 17:41 chris that allows us to change the weightings 17:41 kados I think it'll be more like: 17:41 kados yea 17:40 chris run query 17:40 chris pqf-xml -> pqf 17:40 chris transform 17:40 chris pqf -> pqf-xml 17:40 chris cgi -> pqf 17:40 chris yep 17:40 kados though it begs the question, what's the point of using ccl in the first place 17:40 kados mostly, yes 17:40 kados well ... 17:39 chris for idea 2? 17:39 chris convert back to pqf (now with weightings) 17:39 chris transform with xlst 17:38 chris convert to pqf - xml 17:38 chris convert to pqf 17:38 chris so the idea is .. take you ccl made from your cgi 17:37 chris yep 17:37 kados since I can't imagine it would really slow things down that much ... given how fast the actual query is now 17:36 kados slower isn't that bad I don't think 17:35 kados hmmm 17:35 chris altho probably more easily extensible 17:35 chris but perhaps slower 17:35 chris probably 17:35 chris hmmm 17:33 kados of course, the other option may actually turn out to be simpler ... just turn a PQF query into XML, parse it with a stylesheet, then turn it back into PQF 17:32 chris yep 17:31 kados s/either/entire/ 17:31 kados and the Operands can be an either query 17:31 kados you basically have Operands and Operators 17:31 kados http://indexdata.dk/yaz/doc/tools.tkl#CCL 17:31 kados so the CCL syntax in pretty simple: 17:30 chris k 17:30 kados to incorporate all the fancy field weighting 17:30 kados each leaf node will be branched 17:30 kados then I want to transform that CCL into PQF but it's not a one to one mapping 17:29 chris right 17:29 kados which I can pretty easily transform to CCL 17:29 kados I have incoming cgi 17:29 kados the basic problem is 17:28 kados ) 17:28 kados which I've been trying to understand since I started with perl :- 17:28 kados which has something called an RPN tree 17:28 kados there's this: http://search.cpan.org/irk/Net-Z3950-SimpleServer/SimpleServer.pm 17:25 chris looking now 17:25 chris sorry was on phone 17:24 kados under 'how can I map a single search query (like title) to a pqf string like the above?' 17:24 kados chris: there's a brief explaination of what I need to do there ... my two options 17:24 kados chris: http://wiki.koha.org/doku.php?id=zebraprogrammerguide#field_weighting 17:22 kados their official position is that 'viruses' are causing the wireless outages at this coffee shop :-) 17:21 kados they're realy wankers :-) 17:21 kados something funky going on with the local ISP I hink 17:21 kados well strangely, I can't even open any _new_ ssh connections 17:21 kados yea 17:21 kados using tor or sth 17:21 chris ud need a squid proxy set up somewhere :) 17:21 kados I should be doing that anyway :-) 17:20 chris can u tunnel it thru ssh? 17:20 chris bummer 17:20 kados still no http tho :( 17:20 kados chris: k ... back 15:41 kados useful even 15:41 kados yea, that might be useufl 15:41 kados cool 15:41 owen Now that I'm out of Chauncey I can be a lot more flexible with my time. If you ever feel like we need to meet face to face it's a lot easier for me now to make that happen 15:40 kados maybe I'll even write up a spec :-) 15:40 kados ok, we can chat about it tomorrow 15:39 owen I'm about to head out here, so it'll have to be later 15:39 kados well ... 15:39 kados ok 15:39 owen Actually, I'd like to talk it through with you again, because I wasn't taking good notes the last time we talked about it. 15:39 owen I've been working on tweks here and there 15:38 kados owen: have you started on a new design yet? 15:30 kados alright ... give me a few minutes :-) 15:29 owen I think that would be very cool 15:29 kados to expand the search using just the facet 15:29 kados and maybe we could allow the user to [X] the original search term on the next page 15:29 kados then all the facets can go together 15:28 kados ok, so we limit then 15:28 kados right 15:28 owen Think about how Amazon allows you take your original search term and look for it in a particular category (books, dvds, electrics, etc) 15:27 owen I'm leaning towards thinking that users expect to be able to narrow their results first 15:26 kados would that work? 15:26 kados if we kept the facets as 'expand' and added the 'limits' somewhere else on the page 15:25 kados well ... 15:25 kados yea, it might be confusing 15:25 kados hmmm 15:24 owen What if limiting is the default for facets, and discovery is an option that takes you away from your original result set? A new page with a full list of the facets that can be used to branch off the original search? 15:23 kados but ... really both are useful 15:23 kados for the opac I think 'discovery' is better than 'limiting' in most cases 15:22 kados good question 15:22 owen Is it more important to offer the opac user a means of narrowing their search or of discovering more related information? 15:21 kados is there any visual mechanism we can use to signal either inclusion or exclusion of the original search term? 15:20 kados well ... 15:20 kados :-) 15:20 owen Seems like we talked about this before 15:19 owen Yeah... 15:19 kados at worldcat too 15:19 kados because at NCSU they limit it 15:19 kados i guess one question is, _should_ the subject, series, author links limit or expand the search? 15:18 owen If you can simply put the two sets into two different loops we'll be able to style them separately 15:18 kados any ideas? 15:18 kados they should be visually separated I suppose 15:17 kados hmmm 15:17 kados ahh ... good point 15:17 owen while the branch, content, audience links refine it 15:17 owen The subject, series, author links expand or re-focus the search... 15:16 owen One thing we need to be careful about... 15:12 kados we can always change it later if we want 15:11 kados ok ... I'll just add them to the facets for now then 15:11 owen Yes to them being primary 15:11 kados yep, I can 15:11 owen Yes--particularly if you can get those facets to reflect the contents of that particular result set 15:11 kados but I can see branch limits being separated 15:11 kados to me anyway 15:10 kados seems like they're primary ones 15:10 owen It could go either way. If the branch (content, audience, etc) limits are as "primary" as the subject, author, etc. limiters, then we should create new facets just like the existing ones. If they're a less-important feature then we could "compress" them into drop-down menus for brevity. 15:08 kados I can whip it out in about 10 minutes ... we just need to decide how we want it to work 15:07 kados I guess the question is, is the branch limit going to be a facet? or is it separate from the facets? 15:06 kados then ... we could add some new facets ... 15:06 kados (for instance, I like the size of worldcat's facets) 15:06 kados is that what you had in mind? 15:06 kados the new design will shrink the total height of the three facets we have now 15:05 kados I guess my assumption is that now that we've only got 5 of each facet displaying 15:05 kados well ... 15:05 kados right 15:05 owen I was just trying to think of an alternative solution that wouldn't involve pushing a list further down the screen 15:04 owen The tigra menus you had before 15:04 kados js tree? 15:04 owen I was just thinking we could replace the little js tree with the branch limiters, content limiters, etc. 15:00 kados owen: or is it strictly a branch limit? 14:59 kados owen: so the branch drop-down ... should that be where we allow the user to continue the search in worldcat, etc.? 14:56 kados owen: committed 14:55 kados ok, now there's a type and a typeid 14:52 kados yea, that seems to work well 14:49 kados wonder if we should give each of them an id and have that be the #link 14:49 kados ok, I think the expandable part is working with ccl queries and all three types of facets 14:30 owen Cool 14:29 kados owen: ok ... the 'Show More' is working for ccl queries and subjects 14:19 kados I'll add that as a link to the template 14:19 kados owen: ok ... if you add expand=su to your cgi query, it will expand the subject list 14:11 kados I'll add that right now too 14:09 owen Yeah 14:09 kados kind of like ncsu's 'send search to'? 14:08 kados sounds like a good idea 14:07 owen Instead of having a link that would limit search results to "Branch -> Chauncey," have a drop-down list that resubmits the query when you select a branch 14:06 kados how would that work? 14:06 owen What would you think of dropdowns for the "further limit search" options? 14:05 kados ok, lets do that then 14:05 kados right 14:05 owen I'm wavering... I guess I'm leaning towards a page reload. Partly because I'm starting to worry about the total page size with all those trees loaded automatically 14:05 kados just let me know how you want to do it and I'll make it so :-) 14:04 kados btw: they are hidden now 14:04 kados or a page reload is ok too 14:03 kados you could test for js, then apply the hide only if they had js 14:03 kados right 14:03 owen But people with CSS but no Javascript won't have a way to expand the tree 14:03 kados prefer a page reload? 14:03 owen People with no CSS will see everything anyway, so they're fine. 14:02 owen It gets sticky when you hide something with CSS and then try to show it with Javascript 14:02 kados what do you suggest? 14:02 kados right 14:02 owen NCSU doesn't have an un-hide option because they reload the page evertime someone clicks a 'more' link...so if someone clicks /another/ 'more' link the previously expanded section is contracted again. 13:55 kados anyway ... 13:55 kados but every system has it's flaws 13:55 kados they made some of the same mistakes Koha made along the lines of MARC in SQL and a few new ones :-) 13:54 kados but requires RPC for every client 13:54 kados it's very scalable 13:54 kados so any of the clients have to make multiple calls to the system to retrieve each element of data 13:53 kados s/build/built/ 13:53 kados they basically build the entire system to use an RPC framework 13:52 kados but it's an architectural decision 13:52 kados definitely 13:52 owen So you do see that as a disadvantage. 13:52 kados yes, unfortunately 13:51 owen Do you have any knowledge of how the Pines folks are deciding what tech to require for their opac? Are they requiring javascript? 13:51 kados :-) 13:51 kados ncsu doesn't have that option 13:51 kados hmmm 13:50 owen I'm thinking it gets tricky when you try to re-hide. 13:50 owen possibly 13:49 kados will that work? 13:49 kados using js 13:49 kados that lets you unhide them 13:48 kados then you have a link below 13:48 kados yea 13:48 owen And the script passes "class" as visible or not based on... 13:48 kados for the hidden class 13:48 kados and you have a css rule 13:48 owen Right 13:48 kados etc. 13:48 kados or just do <li class=<!-- TMPL_VAR NAME="class" -->> 13:47 kados with a class of 'class' 13:47 kados so you can wrap each <li> in a span 13:47 kados called 'class' 13:47 kados I'll just add a new var there 13:47 kados the inner facets loop 13:47 kados <!-- /TMPL_LOOP --> 13:47 kados <li><a href="<!-- TMPL_VAR NAME=link -->"><!-- TMPL_VAR NAME="label" --></a>(<!-- TMPL_VAR NAME="count" -->)</li> 13:47 kados <!-- TMPL_LOOP NAME="facets" --> 13:47 kados owen: so we've got this: