Koha Release Manager newsletter #8 2012-06
Welcome to this monthly RM newsletter. As usual, a lot of new things last month.
As usual, there are a lot (124 at time of writing) of patches waiting for being signed-off by someone to be included.
The QA team does a good job at QAing, but is a little bit late (40 patches signed, and just waiting for QA to be included).
Last month, the KohaCon hackfest in Edinburgh resulted in a lot of activity, you can see it here. If you want to join us for improving Koha, you’re welcomed, as usual ! (yes, that’s a lot of “as usual”)
Important patches pushed
CHI caching framework (bug 8092)
The undergrounds of Koha source code have been upgraded to use CHI library, that is a generic framework for caching things. This is now the unique method we will use to do caching. Note that CHI is a non-mandatory package, if you don’t plan to use caching, you don’t have to install it.
Dobrica also has plans to continue improving caching in Koha, he started a discussion about that on the developers mailing list.
DOM indexing (bug 7818)
Another very important underground change is the indexing mechanism we use for zebra. Until last month, we used GRS1 indexing. With patch 7818 we use DOM.
What does it change ? The most visible technical change is the configuration file for indexes that is now XML (instead of record.abs). The most visible change for user is the ability to index a phrase that spans subfield boundaries. For example, giving the subject heading:
650 $a Cats $x pharmacology $x catnip
It is not possible to construct, in GRS1 a search that will reliably match all instances of the phrase “cats pharmacology catnip”.
In DOM it is possible.
If you have a Koha already running, you’ll have to go to the wiki page to discover what you must do to switch from GRS1 to DOM
Note that, if you do nothing, you’ll still have GRS1 and it still will work.
EAN support (bug 5337)
EAN support has been added (for UNIMARC only)
If you are using UNIMARC, your should run the misc/batchRebuildBiblioTables.pl script. It will populate a new biblioitems.ean field, from 073$a
Calendar changed (bug 8181)
This older version of the plugin is open source but is now unmaintained. The newer version is not open source.
This plugin has been replaced by the jQueryUI datepicker widget: http://jqueryui.com/demos/datepicker/
Popularity properly managed (bug 6557)
Although Koha has had the option to sort search results by “popularity” for years, the feature was never fully implemented.
A working popularity sort would allow patrons (and staff) to view their search results sorted according to which items are most popular.
In order to implement this feature 942$0 (total issues) is now automatically updated when items are checked out, and the indexing configuration adjusted slightly to take advantage of popularity data.
If you want to implement this feature :
- all your frameworks must have a 942$0 subfield, and it must be linked to biblioitems.totalissues SQL field
- the cronjob update_totalissues.pl must be set in crontab
If you need more information, you can refer to comment 17 of the bug.
opac-sendbasket.pl security leak (bug 3280)
The bug 3280 has been pushed, it fixes security leak in opac-sendbasket.pl
The behaviour has changed = it’s now possible to send a basket by mail only if you’re an identified user.
Replace YUI autocomplete with jQueryUI (bug 7747)
the YUI dependencies have now been completely removed. Please report any problem, in case we’ve forgotten to switch some YUI-autocomplete scripts to jquery !
CHI (bug 8092), -that is not mandatory-
Template::Plugin::HtmlToText (bug 8062), that transform a HTML to plain text
Text::Unaccent (bug 7411), that is used to remove diacritics from automatically generated userids
Community organisation & decisions
- Bugzilla default assignee: the default assignee has been switched to koha-bugs mailing list for many components. For some components there is a default assignee, it’s a component that someone really wanted to endorse
- Last month I spoke of QA enhancement: should we request that QA is made with adding a sign-off on the bug itself. It seems that the idea has gone, the QA team being not enthusiastic about it. Note, however, that for some -usually large- bugs, the QA team add a signoff, but it’s not becoming a rule for all patches QAed
- The Database schema and coding guideline discussion also seems to have reached a consensus to use column_id as identifier. Te coding guidelines have been updated, so please note that, if you create a new table, it must have a primary key which name finishes by _id. All tables that does not respect this rule and already exist will have to be updated, a long term goal