koha-fr.org pour demander les commentaireskoha-fr.org ou presque !koha-fr.org ← Previous day | Today | Next day → | Search | Index
All times shown according to UTC.
| Time | Nick | Message |
|---|---|---|
| 13:00 | thd | owen or kados: are you there? |
| 13:06 | kados: are you there now | |
| 13:06 | kados | thd: yep |
| 13:06 | thd | ? |
| 13:10 | kados: I left the field order incorrect as I had originally | |
| 13:11 | kados: That is just a minor point but it was funny that we had the same thought about that almost | |
| 13:11 | kados: I will get back to that | |
| 13:12 | kados: the first important thing is the boolean attached to the first field | |
| 13:13 | kados: that is, there is a boolean drop down selection for each selection | |
| 13:14 | s/each selection/each textbox unless you choose a command query/ | |
| 13:15 | kados | why is that? |
| 13:15 | why have boolean for the first one? | |
| 13:15 | thd | kados: it makes it possible for a much more flexible use of booleans independently for each textbox |
| 13:16 | kados: it also makes it easier to code RPN | |
| 13:36 | I guess nothing I typed went through for several minutes | |
| 13:38 | <thd> kados: ease of coding was an afterthought with respect to the RPN issue | |
| 13:38 | <thd> kados: there is a very important ease of coding issue that is different | |
| 13:38 | <thd> kados: one that is very important for allowing arbitrarily expandable forms to be easy to implement | |
| 13:38 | <thd> kados: your names are as I have always written forms in the past with a separate numer for each row of input fields | |
| 13:38 | <thd> s/numer/number/ | |
| 13:38 | <thd> kados: that requires enumerating variables to use additional fields for arbitrary form expansion | |
| 13:38 | <thd> kados: I noted that the Index Data implementation of a PHP/YAZ demonstration used "host[]" as the name for the four targets and CCL search box | |
| 13:38 | <thd> kados: that creates an array that can be looped through to find those with values. | |
| 13:38 | <thd> kados: in the case of checkboxes for host[] which I renamed to target[] after I expanded the parameters only selected checkboxes form part of the get or post query | |
| 13:38 | <thd> kados: in the case of textboxes, which I used for the Bib-1 form as search_field_term all textboxes are returned even if they have no values | |
| 13:38 | <thd> kados: therefore, each row of search_field_whatever responses can be looped through | |
| 13:38 | <thd> kados: adding arbitrary additional rows does not require creating new enumerated variables for the names of the fields. | |
| 13:40 | kados: does that make any sense? | |
| 13:52 | kados: I have not implemented the arbitrarily expandable forms yet but I wanted to have the field order correct which it is not yet before I variabalised them. | |
| 14:03 | owen | thd: kados just left here, so he's in transit |
| 14:03 | thd | owen did you understand my point about the name attribute for form variable names? |
| 14:05 | owen | No :) |
| 14:27 | kados | thd: I'm back |
| 14:29 | thd | kados: do you understand my point about avoiding individually numbered name attributes for form variable names? |
| 14:30 | kados | almost |
| 14:30 | what does it have to do with host[]? | |
| 14:31 | you mean have multiple <input> fields with the same name of 'search_field_term'? | |
| 14:31 | thd | kados: host[] was just the example of first time I noticed this usage for repeated name attribute names |
| 14:31 | kados | and in the back end, just loop through all of them? |
| 14:31 | thd | yes |
| 14:31 | kados | interesting |
| 14:32 | I hadn't thought of that | |
| 14:32 | but ... how do you group your operators that way? | |
| 14:32 | thd | kados: and of course I have working code |
| 14:32 | kados | and ... speaking of operators |
| 14:32 | I understand how to do: | |
| 14:32 | find: harry AND Potter OR Goblet | |
| 14:32 | thd | kados: let me send you an extract of the cod so you can see clearly |
| 14:32 | kados | but I don't know how to do: |
| 14:33 | find (harry AND potter) OR Goblet | |
| 14:33 | or ... | |
| 14:33 | harry AND (Potter OR Goblet) | |
| 14:33 | :-) | |
| 14:33 | I dont' understand how to represent that in PQF/RPN | |
| 14:34 | thd | I do not know either except that you can write a CCL query in my Z30.50 client and see what comes back from nesting the parenthesis |
| 14:35 | kados: do not waste your time looking for something that they never documented | |
| 14:35 | kados: I did not do an exhaustive search but I do not think that is easy to find and expect that it is not there at all | |
| 14:36 | kados | thd: ok ... let me open up my Z39.50 port |
| 14:37 | thd: 66.213.78.100:9900/biblios | |
| 14:38 | thd: can I specify the host in the CCL? | |
| 14:38 | I don't think ZOOM supports it :( | |
| 14:38 | thd | kados: however, if you do a test CCL search in my client such as (ti = harry) and ((ti = potter) or (ti = goblet)) |
| 14:39 | kados | thd: I'd like to see how ZEbra treats that CCL query |
| 14:39 | thd | kados: yes you can specify the host in the same way |
| 14:39 | kados | do do you just pass it on in RPN? |
| 14:39 | thd: what toolkit do you use to map CCL2RPN? | |
| 14:39 | thd | kados: YAZ converts it and you see the conversion at the top of the response page |
| 14:40 | kados | nothing yet |
| 14:40 | thd | s/YAZ/PHP\/YAZ/ |
| 14:40 | kados: if you construct a query with too many records returned you may be waiting a very long time | |
| 14:41 | kados | AHH |
| 14:41 | thd | kados: remember you are pulling down records from a dialup connection and the default target selection list is huge for finding obsure titles for Afognak |
| 14:42 | kados | I'll be back in about 10 minutes |
| 14:42 | thd | kados: maybe the query will be done then :) |
| 14:46 | kados: your query did not seem to take long | |
| 14:59 | owen | kados: one thing I forgot to ask... |
| 14:59 | kados | owen: sure |
| 15:00 | owen | Have you worked on the 'simple search' at all? |
| 15:00 | You talked about making it work for google-like queries | |
| 15:00 | kados | right |
| 15:00 | I haven't worked on it at all | |
| 15:01 | there's not even much error checking | |
| 15:01 | I'll have to take a look at the possible feedback mechanisms that ZOOM gives me | |
| 15:01 | to see if I can intuit the error with a query before deciding how to treat it | |
| 15:01 | it could use some general user testing | |
| 15:02 | for example, i realize that patrons expect to type: | |
| 15:02 | harry potter | |
| 15:02 | and get a result | |
| 15:02 | rather than: | |
| 15:02 | "harry potter" | |
| 15:02 | or | |
| 15:02 | harry and potter | |
| 15:02 | but I'm not sure if there are other suprises too | |
| 15:02 | surprises even :-) | |
| 15:03 | thd_away: ti = harry) and ((ti = potter) or (ti = goblet)) | |
| 15:03 | thd_away: becomes: | |
| 15:03 | thd_away: @and @attr 1=4 harry @or @attr 1=4 potter @attr 1=4 goblet | |
| 15:03 | thd_away: which slightly complicates things :-) | |
| 15:11 | hmmm | |
| 15:11 | now I'm confused :-) | |
| 15:14 | thd_away: my implementation can't handle the full hierarchy | |
| 15:15 | thd_away: and it seems to me that Zebra's implementation of it is fairly weak | |
| 15:15 | thd_away: we need to do some testing | |
| 15:16 | thd_away: also, we need to figure out if it's possible to reprsent that in an easy form | |
| 15:17 | owen: just so you don't get left behind | |
| 15:18 | owen: pretend we're doing algebra | |
| 15:18 | (4 + 5) * 6 | |
| 15:18 | dewey | 54 |
| 15:18 | kados | expressed in RPN/PQF (which use the same syntax) |
| 15:18 | woule be: | |
| 15:18 | * 6 + 4 5 | |
| 15:19 | or ... | |
| 15:19 | * + 4 5 6 | |
| 15:19 | :-) | |
| 15:20 | owen: does that make sense? | |
| 15:21 | owen | If you say '* + 4 5 6' how does the interpreter know how to group the numbers? |
| 15:35 | kados | well ... |
| 15:35 | + 4 5 | |
| 15:35 | is one operand | |
| 15:36 | 6 | |
| 15:36 | is the other | |
| 15:36 | * is the operator for those two | |
| 16:19 | thd: I'm back, but only briefly unfortuantely | |
| 16:19 | thd | kados: my test shows that @and @not fails while @not works |
| 16:20 | kados | thd: I've been roped into a bike ride :/ |
| 16:20 | thd: that is true | |
| 16:20 | thd | kados: I did that for a film and I found I had forgotten how to ride steadily |
| 16:20 | kados | thd: however, logically, the boolean operator NOT is really AND NOT |
| 16:20 | thd: there is an implied AND in NOT | |
| 16:21 | thd | kados: yes I understood the logic part |
| 16:21 | kados | thd: so the grouping only applies when you're mixing OR and AND |
| 16:21 | (and possible XOR) | |
| 16:21 | thd | kados: BnF does not support @not :( |
| 16:21 | kados | wow |
| 16:22 | well ... anyway ... I'm not sure to what extent Zebra supports the hierarchy | |
| 16:22 | we need to figure out some good tests to try | |
| 16:22 | ways to see if a given target supports a given level of Type-101 queries | |
| 16:22 | thd | kados: my scheme is very extensible, although, I have nothing to support nested groups specifically yet |
| 16:23 | kados | I'll be back soon, gotta do the bike thing for an hour or so |
| 16:23 | thd: I'll be thinking of a solution :-) | |
| 17:16 | thd: you around? | |
| 17:26 | thd | kados: yes |
| 17:28 | kados | I can't picture a form-based way to implement the grouping |
| 17:28 | thd | kados: I pictured it two years ago |
| 17:28 | kados | thd: have you come up with anything? |
| 17:28 | thd: could you describe it to me? | |
| 17:29 | thd | kados: I had various note cards of possible implementations |
| 17:29 | kados: I will describe the general idea | |
| 17:29 | kados | ok |
| 17:30 | thd | kados: each row of query fields has a drop down field for group instead of a boolean itself. |
| 17:31 | kados | how do you label the group? |
| 17:31 | thd | kados: at the base of all the rows you have a part of the form where you assign the boolean relations between groups |
| 17:32 | kados: group 1, group 2 if the groups themselves are infinitely malleable | |
| 17:33 | kados | hmmm |
| 17:34 | i don't quite get it | |
| 17:34 | could you do a mock up in html? | |
| 17:34 | just a form with a few <select>s and <input>s? | |
| 17:34 | thd | kados: I was actually planning to |
| 17:34 | kados | cool, that might help me :-) |
| 17:35 | it doesn't have to work mind you | |
| 17:35 | if I can at least visualize how it's _supposed_ to work | |
| 17:35 | thd | kados: the assignment of the booleans is the tricky part to do elegantly |
| 17:35 | kados | that will help |
| 17:35 | you almost need some visual cues | |
| 17:35 | like lines connecting them | |
| 17:35 | or something | |
| 17:36 | but that relies too much on the interface | |
| 17:36 | thd | kados: I had intended to do just that, although, you would need functionality to show what would happen if you pressed a form submit button to expand the nesting hierarchy |
| 17:37 | kados | no, you could just tell me what would happen :-) |
| 17:37 | ahh ... it expands? | |
| 17:37 | hmmm | |
| 17:39 | thd | kados: it would require actuating a submit button or link or auto-submitting and redrawing the form if the user had JavaScript enabled |
| 17:40 | kados: you no I do not like any functionality that only works in JavaScript | |
| 17:40 | s/no/know/ | |
| 17:41 | kados | hmmm |
| 17:41 | right, i get that | |
| 19:26 | thd | kados: I have to go now but look at my Z39.50 client again. |
| 19:27 | kados | ok |
| 19:27 | hehe | |
| 19:27 | wow, that's just nuts :-) | |
| 19:28 | thd | kados: It is not elegant yet and maybe a drop down with parenthesis levels would be better but then you have to worry about correct closing parenthesis. |
| 19:28 | kados: pretend the find records containing other term sets lines are not there. | |
| 19:29 | of course term sets rows themselves should be expandable | |
| 19:29 | kados: I only design for infomaniacs :) | |
| 19:29 | kados | :-) |
| 19:30 | thd | kados: If it works for infomaniacs it can be simplified for everyone else. |
| 02:08 | btoumi | hi evrybody |
| 02:12 | Comete | bonjour/hi |
| 02:13 | btoumi | bonjour comete |
| 02:13 | Comete | salut btoumi |
| 02:14 | btoumi | ca va? |
| 02:15 | Comete | j'ai un petit tableau en PDF de fonctionnalités essentielles que le personnel de notre mediatheque m'a envoyé. Ce sont des choses qu'ils n'ont pas vu jusque la dans Koha après une premiere exploration. Puis-je vous le transmettre ? |
| 02:15 | btoumi: oui et toi ? | |
| 02:17 | btoumi | Comete:de nouveau ready pour le module lecteur de la head de koha |
| 02:18 | Comete | btoumi: bon courage |
| 02:19 | btoumi | Comete: merci |
| 02:31 | hello paul | |
| 02:31 | paul | hello btoumi & Comete |
| 02:33 | osmoze | bonjour :) |
| 02:33 | paul | hello osmoze |
| 02:35 | chris are u there ? | |
| 02:38 | Comete | paul: bonjour |
| 02:38 | dewey | hola, Comete |
| 02:38 | Comete | hola dewey |
| 02:38 | dewey | hi, Comete |
| 02:41 | Comete | paul: j'ai un petit tableau en PDF de fonctionnalités essentielles que le personnel de notre mediatheque m'a envoyé. Ce sont des choses qu'ils n'ont pas vu jusque la dans Koha après une premiere exploration. Puis-je vous le transmettre ? |
| 02:42 | ils sont peut-être passé à côté de choses qui existent déjà... si vous pouviez confirmer | |
| 02:43 | osmoze | Comete, je suis bien interessé pour voir le tableau et ainsi peut etre compléter ^^ |
| 02:43 | Comete | osmoze: je vous l'envoie par mail ? |
| 02:44 | paul | Comete: oui, envoie. |
| 02:44 | Le mieux serait même de le mettre sur wiki.koha.org, dans la section "francaise" si c'est en Francais. | |
| 02:44 | Comete | à quelle adresse ? |
| 02:44 | paul | ainsi, je pourrai répondre et tout le monde pourrait enrichir |
| 02:44 | (après tout, je ne suis pas le seul à être capable de répondre ;-) ) | |
| 02:44 | l'idéal serait : | |
| 02:44 | Comete | paul: d'accord je regarde ca de suite |
| 02:45 | paul | - mettre tout dans le wiki |
| 02:45 | - envoyer un mail sur la liste infoskoha-fr.org pour demander les commentaires |
|
| 02:45 | Comete | ok |
| 02:52 | quel nom voulez-vous que je donne à cette section ? | |
| 02:52 | paul | peu me chaud. |
| 02:52 | osmoze | suggestions par les mediatheques ? |
| 02:53 | ce qui peut differencier les suggestions des mediatheques publiques et les bibs universitaires | |
| 02:55 | paul | bonne idée, Paul approuve. |
| 02:56 | Comete | ok mais certaines suggestions n'ont peut-être pas lieu d'être si la fonctionnalité existe déjà |
| 02:57 | osmoze | elles seront modifier par quelqu un |
| 02:57 | modifiées | |
| 02:58 | Comete | dites ca marche comment ce wiki ? |
| 02:59 | je n'aime pas les wikis y en a pas un pareil et je trouve ca pas pratique pour chercher un truc, mais c mon avis ;) | |
| 03:00 | paul | Comete: moi non + j'aime moyen les wiki. |
| 03:00 | mais celui là est assez pratique. | |
| 03:00 | Comete | donc ca ne vient pas de moi :) |
| 03:00 | paul | nous l'avons mis en place récemment, et franchement, il est pas mal. |
| 03:01 | la syntaxe est relativement naturelle | |
| 03:01 | Comete | paul: faut s'enregistrer ? |
| 03:01 | paul | je crois, oui |
| 03:01 | (mais pas de crainte de spam à avoir ;-) ) | |
| 03:01 | Comete | allons-y |
| 03:01 | ok | |
| 03:01 | paul | encore que, il y a des protections anti spammeur, mais pas sûr que ce soit via l'authentification. |
| 03:02 | mais c'est tjs mieux, pour savoir qui a fait quoi. | |
| 03:02 | impressionnant... hdl a du désactiver l'anti spam sur koha-fr.org, pour des questions de paquet debian bcp changé, du coup plus d'anti spam. | |
| 03:03 | et 5 spams à l'heure sur infoskoha-fr.org ou presque ! |
|
| 03:03 | (qui arrivent chez moi vu qu'ils sont bouncés par la liste, réservée aux abonnés) | |
| 03:04 | osmoze | c est la fete :) tu vois, je suis sur que jamais autant de monde s est preoccupé de ta santé sous abdominale :p |
| 03:04 | paul | et aussi de me faire un prêt à pas cher |
| 03:04 | (si j'aurais su, j'aurais pas acheté ma nouvelle voiture comptant lundi dernier ;-) ) | |
| 03:05 | osmoze | :) t as pris une grande alors maintenant ? |
| 03:09 | Comete | bon je dois etre stupide mais je comprends rien à ce wiki, comment je crée ma page et ou ca ? :p |
| 03:11 | paul | http://wiki.koha.org/doku.php?id=french |
| 03:11 | modifier la page pour ajouter par exemple : FonctionnalitesPourMediathequesDeLecturePublique | |
| 03:11 | (l'alternance maj/min est importante) | |
| 03:12 | ca sera transformé en lien "vide". | |
| 03:12 | tu peux alors créer la page ! | |
| 03:12 | devant le questionnement général en privé : je viens d'acheter un fiat multipla | |
| 03:12 | Comete | ok merci |
| 03:12 | paul | 6 places (3 devant, 3 derrière) |
| 03:13 | une place folle derrière même lorsque je suis au volant (1m95 pour ceusses qui me connaissent pas ;-) ) | |
| 03:13 | les enfants sont ravis. | |
| 03:13 | maman la trouve moche de l'extérieur, mais quand on est dedans on ne voit pas l'extérieur ;-) | |
| 03:14 | papa la trouve rigolote, mais surtout sobre, spacieuse, toussa. | |
| 03:15 | osmoze | :) |
| 03:18 | Comete | paul: on peut faire des tableaux avec ce wiki |
| 03:18 | ? | |
| 03:19 | paul | surement, mais je ne sais pas trop comment. je regarde |
| 03:19 | || STATUS || FEATURE || WHO || NOTE || | |
| 03:19 | || Completed ||Replace marc* tables with Zebra || PP ||note|| | |
| 03:19 | || In Process ||apache2 with mod_perl2 support for improved speed || CC ||end of May, June|| | |
| 03:20 | Comete | ok merci |
| 03:30 | ToinS | Comete: plus d'infos sur la syntaxe de dokuwiki => http://wiki.splitbrain.org/wiki:syntax |
| 03:45 | Comete | comment appelle-t-on le personnel traveillant dans une médiathèque, des bibliothéquaires ? |
| 03:45 | travaillant | |
| 03:48 | paul | des médiathécaires se dit aussi |
| 03:48 | (et sinon, on parle plutôt de bibliothéCaires que QUaires !) | |
| 03:49 | Comete | ah ok :p |
| 03:52 | osmoze | médiathécaires ? |
| 03:52 | paul | ouaip, j'ai déjà entendu ca. |
| 03:53 | mais c'est un néologisme | |
| 04:05 | Comete | le mail est envoyé sur infoskoha-fr.org |
| 04:06 | il faut être inscrit peut-être ? | |
| 04:06 | paul | yep, je viens de voir le rejet arriver ;-) |
| 04:06 | Comete | zut |
| 04:07 | je recommence | |
| 04:07 | en attendant: http://wiki.koha.org/doku.php?[…]delecturepublique | |
| 04:10 | paul | 1 chtite remarque à propos du mail : |
| 04:10 | - dire bonjour et se présenter un peu ;-) | |
| 04:11 | btoumi | comment on dit doublon en anglais? |
| 04:12 | comme doublon de donnee | |
| 04:15 | paul | duplicate |
| 04:15 | btoumi | trop tard j'ai mis double |
| 04:16 | paul | ah, faut changer... |
| 04:16 | Comete | paul: euh oui c vrai j'ai oublié mais de toute façon je dois le refaire |
| 04:16 | paul | (sauf si c'est juste un mail ;-) ) |
| 04:16 | btoumi | c sur le commit du cvs |
| 04:16 | dsl | |
| 04:26 | paul | et hop, Comete => inscription confirmée. |
| 04:29 | Comete | paul: et hop mail envoyé ;) |
| 04:39 | paul | ah, vla pierrick. |
| 04:39 | pierrick | bonjour :-) |
| 04:39 | paul | salut. |
| 04:40 | pierrick | (bébé malad) |
| 04:53 | ToinS | salut pierrick |
| 04:54 | pierrick | salut ToinS |
| 05:23 | Strait | hello again :) |
| 05:26 | pierrick | hello Strait |
| 05:35 | Strait | i have another problem with koha installation today :P |
| 05:36 | i don't have root access to MySQL | |
| 05:36 | what I have there is a readily created MySQL user with full permissions to a database called koha | |
| 05:38 | i've been looking at Install.pm to see if it's possible skip the kohaadmin user creation | |
| 05:38 | and use the readily created kohaadmin user instead | |
| 05:39 | but i'm not really sure what to do with Install.pm | |
| 05:39 | there is a databasesetup sub in the script and i'm pretty sure that's the right place | |
| 06:42 | paul | Strait: to setup you database, just put koha.mysql in the database |
| 06:42 | then, run updater/updatedatabase | |
| 06:42 | then, add whatever marc flavour you need in misc/marc_datas and misc/sql_datas | |
| 06:46 | Strait | thanks paul, i'll try that as soon as i have time :) |
| 06:47 | it seems to be quite tricky to try to install koha without root access to the server and to mysql | |
| 06:48 | paul | yes it is. |
| 06:48 | another solution, highly easier : | |
| 06:48 | Strait | i've done a couple of koha installations with root account before and they went quite smoothly |
| 06:48 | paul | - install Koha on a server with root access |
| 06:48 | - copy directories & database to the other server | |
| 06:49 | works fine if you have all Perl packages needed. | |
| 06:49 | Strait | that might be a better way than trying to run installer on the machine with limited permissions |
| 06:56 | this koha will be used by my university | |
| 06:56 | it's for the department of information studies | |
| 06:57 | it'll also likely be utilized by the local polytechnic | |
| 06:58 | as a teaching system primarily, but we are also planning on cataloging the collection of the department's library on it | |
| 06:58 | (status report: finland :D) | |
| 08:16 | Comete | could you tell me where i could put temporaly the second part of my addresses in koha 2.2.5 ? |
| 08:19 | paul | ??? |
| 08:25 | Comete | paul: my addresses have two parts : address1 address2 but there is only one field for the address in koha 2.2.5 |
| 08:25 | paul | btoumi, an answer ? |
| 08:26 | Comete | paul: btoumi told me that the problem will be resolved in koha 3 but before this release i must put these informations somewhere |
| 08:27 | btoumi | sorry i'm here now |
| 08:28 | Comete | so i wonder if there's no not so useful field i could use in place and then migrate it when koha 3 will be stable |
| 08:36 | btoumi | u have in 225 the field other address |
| 08:42 | Comete | btoumi: ok thank you |
| 08:45 | btoumi | but if u need to enter 2 address perahps u must choose another field |
| 10:04 | kados | owen: you around? |
| 10:04 | owen | I just now am! |
| 10:04 | kados | owen: I've been working on the stuff we discussed yesterday, making excellent progress |
| 10:04 | owen | Cool |
| 10:04 | kados | owen: still some stuff to clarify with the catalogers on the use of 008 |
| 10:04 | but things are really shaping up | |
| 10:04 | you can have a look at the advanced search | |
| 10:05 | to try out the new 'Format', 'Audience', and 'content' filters | |
| 10:05 | paul | hello owen & kados |
| 10:05 | kados | format's still got a ways to go |
| 10:05 | hey paul | |
| 10:06 | excellent | |
| 10:15 | owen | So kados, how is the audience filter doing its filtering? |
| 10:16 | kados | owen: at the moment ... |
| 10:16 | there is an index on the 008 field | |
| 10:16 | position 22 | |
| 10:17 | owen: http://www.itsmarc.com/crs/Bib0019.htm | |
| 10:17 | http://www.itsmarc.com/crs/Bib1773.htm | |
| 10:17 | second one is better | |
| 10:17 | owen | Did you make changes to our database to allow that? I though that stuff was missing. Or is it only missing for some records? |
| 10:18 | kados | it's only missing for items edited since Koha |
| 10:18 | not for those added by the catalogers | |
| 10:18 | so if they used the Koha MARC Editor, 008 was actually just wiped out :/ | |
| 10:18 | owen | Okay... I wonder if that's enough to explain the differences in result sets I'm getting between the old and new search |
| 10:18 | kados | fortunately, that's not _too_ many |
| 10:19 | it might be | |
| 10:19 | of course, I will pre-process the records ... but I have several other changes to make and I want to make sure I only pre-process a couple times :-) | |
| 10:19 | owen | I tried keyword "sex" audience "YA" (I thought those two went naturally together) and content "non-fiction" |
| 10:20 | I get 5 results | |
| 10:20 | kados | good search |
| 10:20 | what's the old search give you? | |
| 10:20 | owen | If I do an advanced search for keyword "sex" and itemtype "YA Non-fiction" in our current system I get 51 records. |
| 10:20 | That's quite a difference. | |
| 10:21 | kados | 51 |
| 10:21 | yep ... quite a difference | |
| 10:21 | lets collect a few searches like that so we have something to compare | |
| 10:22 | it's possible that my interpretation of the 008 isn't quite correct | |
| 10:23 | huh ... power and proximity search tabs aren't working ... | |
| 10:23 | wonder if I did something | |
| 10:24 | fixed | |
| 10:24 | owen | This one's even more out of whack: keyword "murder" audience "adult" content "non-fiction": 2 results |
| 10:25 | kados | hehe |
| 10:25 | owen | 246 results in the old search |
| 10:25 | kados | yea, so we definitely can't trust the 'adult' designation |
| 10:25 | hang on, I'll tweak it a bit | |
| 10:26 | try now | |
| 10:26 | hmmm | |
| 10:26 | we need to figure out how to specify 'doesn't exist' | |
| 10:26 | hang on, I"ll check ID's site | |
| 10:28 | so there are 329 results if you don't specify 'adult' | |
| 10:30 | owen: interestingly | |
| 10:30 | owen: "sex" as non-fiction 'juv' pulls up 35 | |
| 10:30 | so ... i guess my first question is | |
| 10:30 | Comete | bye |
| 10:31 | kados | how many records don't have 008 |
| 10:32 | owen | Only 3 of the first 11 results of that 35 are actually classified as Juvenile Non-fiction. The rest are YA or Adult non-fiction |
| 10:32 | kados | select count(*) from biblio; |
| 10:32 | +----------+ | |
| 10:32 | | count(*) | | |
| 10:32 | +----------+ | |
| 10:32 | | 149688 | | |
| 10:32 | select count(*) from marc_subfield_table where tag='008'; | |
| 10:32 | +----------+ | |
| 10:32 | | count(*) | | |
| 10:32 | +----------+ | |
| 10:32 | | 134083 | | |
| 10:33 | owen: so this indicates some problems with our cataloging department :-) | |
| 10:33 | or maybe a disconnect between what NPL defines as a juvenile and what MARC21 does :-) | |
| 10:34 | we can expect roughly 10% to be wrong | |
| 10:34 | based on the number of records missing 008 fields | |
| 10:34 | owen | Remember, we have records that may go back as far as 20 years |
| 10:34 | kados | yep, true ... |
| 10:34 | the good news is | |
| 10:34 | owen | Although I don't know how many of those first ones have survived |
| 10:35 | kados | I can easily 'fix' records based on our itemtypes |
| 10:41 | owen: also figured out the prob with exact title searches | |
| 10:42 | owen: I was using the same '.chr' file for sorting and for searching | |
| 10:42 | 'The', 'a', etc. were getting filtered out | |
| 10:42 | I've updated the abs but haven't reindexed yet | |
| 10:42 | next reindex exact title 'cell' will only pull up 'cell' not 'the cell' | |
| 10:43 | and ... :-) | |
| 10:43 | I figured out a way to acomplish still return other results after that | |
| 10:43 | by combining that original result set with another more general query and merging them, giving the first one priority :-) | |
| 10:43 | so you should get: | |
| 10:44 | cell | |
| 10:44 | the cell | |
| 10:44 | cellulite | |
| 10:44 | etc... | |
| 10:44 | which is very exciting | |
| 10:49 | hey tumer | |
| 10:50 | tumer | hi kados |
| 10:50 | kados | tumer: any progress on integrating the results page? |
| 10:50 | tumer: if not, I'll probably take a crack myself today (got a demo later) | |
| 10:50 | tumer | i am working on the template lots of problems i am having |
| 10:51 | passing on serach variables is causing me problems | |
| 10:51 | kados | hmmm |
| 10:51 | I can do that bit | |
| 10:51 | tumer | but at least it works for the first page |
| 10:51 | kados | I've revised things a little |
| 10:52 | tumer | i'll try to commit something before the meeting |
| 10:52 | kados | cool, thanks |
| 10:53 | right | |
| 10:53 | tumer | hurray for kados |
| 11:08 | kados | owen: asside from fixing the new search filters, what other variables are missing from your template? |
| 11:08 | search_point I think ... | |
| 11:08 | I put in branches and itemtypes | |
| 11:09 | owen | Call number? I guess that has to wait on getting itemcallnumber populated |
| 11:10 | kados | you mean a call number search? |
| 11:10 | owen | Date limiters |
| 11:10 | Yes | |
| 11:10 | kados | right, dates ... just need to rebuild the index for that |
| 11:10 | call number search should already work | |
| 11:10 | just need to add it to the template | |
| 11:10 | Dewey, right? | |
| 11:10 | dewey | kados: no idea |
| 11:10 | kados | hehe |
| 11:12 | owen | But I thought we were going to re-work how we were building call numbers? |
| 11:12 | kados | owen: you want a search on 952$a? |
| 11:12 | well, just for fiction | |
| 11:12 | owen | What's 952$a? |
| 11:12 | kados | i think thats where the catalogers put call numbers |
| 11:13 | woops ... just broke searching :-) | |
| 11:14 | a 1=20 search should work for call numbers | |
| 11:14 | I'll reindex so we can test | |
| 11:15 | owen | The records I'm looking at have the call number ("dewey") in 942k |
| 11:20 | kados | ahh |
| 11:22 | k, reindexing now | |
| 11:22 | with the new 008-based date search | |
| 11:22 | call number | |
| 11:22 | as well as language | |
| 11:24 | paul | kados: interesting mail on koha-devel (from andres) |
| 11:27 | see you later | |
| 11:27 | kados | ahh, yes |
| 11:27 | on the mellon | |
| 11:36 | it's a grant I guess | |
| 11:36 | tina's looking into it | |
| 11:42 | owen: so ... for now, we can try out the call number search using what is in 942$k | |
| 11:43 | I updated the template with the new variables | |
| 11:43 | need to make a couple other quick changes too | |
| 11:46 | tumer | kados: do you need this basic template now? Only the page numbers work on it |
| 11:47 | kados | tumer: sure |
| 11:47 | tumer: I have some time to work on it all day today | |
| 11:47 | tumer | so committing new opac-zoomsearch.pl as well . This one seems a bit buggy |
| 11:48 | kados | ok |
| 11:48 | I'll merge it with mine | |
| 11:48 | thanks tumer ! | |
| 11:48 | tumer: don't worry, I have to get this stable tonight :-) | |
| 11:48 | tumer: got a demo in the morning for apotential client :-) | |
| 11:49 | owen: I need a few minutes to consult with the catalogers | |
| 11:49 | owen: so the new features wont' work for a bit | |
| 11:49 | owen: I"ll ping you as soon as theyr'e ready | |
| 11:49 | owen | Gotcha |
← Previous day | Today | Next day → | Search | Index
