Strigi

On Saturday 2008-01-12 we had a meeting on Strigi. 23 people were logged into the meeting that lasted for two hours. Starting from a status document, we discussed the deficiencies of the current Strigi version. In addition to the point bought up in ... [More] the status document, the following concerns were mentioned: lacking quality of KDE integration in general and of searching in particular threshold for (KDE) developers to start work on the core due to the use of pure C and STL stability issues with some analyzers After a long and insightful discussion we came up with the following main action items: KDE integration: Flavio will coordinate the definition of interfaces over which KDE will handle searching and metadata. He can ask Aaron, Evgeny and Jos for help with the interface design. The interface will cover: Querying via Xesam Configuration of the Strigi daemon Indexing and deindexing of data by passing it to the daemon (allowing for indexing for more than just files) Controlling the daemon (starting, stopping, pausing) Improved configurability of the daemon: Jos will write a version of DaemonConfiguration that allows for more fine-grained control over what will be indexed. Improved Xesam support: the Xesam spec has not been completely implemented in Strigi yet. Jos will continue to work on this. Evgeny will continue coordinating Xesam and Nepomuk. Everybody else can do what they feel like and pick up whatever task they feel like doing. In addition to the main tasks, we also want to have better documentation on the architecture and on the writing of analyzers. For the important tasks in the list above, we would like to have more detailed planning of the steps that will be taken. As to versioning, the improved KDE integration can take place either in playground or the KDE 4.1 branch. In the latter case it can still be backported or provided as a separate library for KDE 4.0. Ps: this is text has been taken from a mail sent by Jos after the meeting [Less]

On Saturday 12 January at 17 UTC an irc meeting will take place on #strigi channel (it's on freenode if you don't know). During the meeting Strigi developers will discuss about the future developments of Strigi. Special guest: Aaron Aseigo. You're all welcome.

Today Strigi 0.5.7 has been released. This is the changelog: use plugins instead of shared libraries for the indexer backends lots of bugfixes and cleanups allow backends to be used in RAM by using ':memory:' as the index name

As Xesam implementation is stabilizing, Strigi is about to make a next step towards semantic desktop. Nepomuk-KDE developer Sebastian Trueg said he will release shortly his RDF code as a storage and indexer backend for Strigi. This means Strigi has ... [More] wider choice of backends, namely RedLand(SQLite-based) and Sesame2(Native RDF storage). Benefits include SPARQL query support, and in case of Sesame2, a much better performance especially for complex queries. SPARQL support comes handy considering very limited capabilities of Xesam query language. Sesame2 is receiving a lot of developer and industry attention and tapping into this resource means we stay on the leading edge of RDF developments in FOSS. Also, this means we have just one step left to provide a semantic desktop on KDE. We've got analyzers, backend, dbus bindings, native c nepomuk ontology bindings. Just a little integration work will bring us the technology we've been looking for. At every step we tried to eliminate costly lock-ins and tap into the most dynamic and promising developments by using flexible APIs and plug-in architecture. Looks like it worked :) [Less]

Saturday, June 30th Jos van den Oever (Strigi's creator) give a presentation at aKademy 2007. You can find a video of his talk here. Good job Jos!

Xesam ontology is shaping up nicely. It had gotten much more flexible and powerful than I thought I could ever entice desktop indexers to implement. Unfortunately Xesam query interface in Strigi will not be finished before aKademy as I hoped, but ... [More] it's not far from completion(days away). I feel it's a fair price to pay for a boost in functionality :) Query interface implementation will be used to fine-tune ontology. Jos will be doing a presentation on Xesam at aKademy.I don't know to what extent Jos will be able to cover subtle, but important improvements in Xesam over previous desktop search approaches. Anyway, soon you will be able to actually test-drive it yourself. This is an opportunity to take a glimpse at what nepomuk will be capable of. Also, there are some exciting developments in the user interface area. Under consideration are: * Data source for plasmoids. You can be sure there will be plenty :) * Jos' FileLight idea * WarpTree data structure browser/navigator * Auto-clustering of results to better represent and help refine searches that return more than 15-20 results. Auto-clustering can go as far as starting from all of your data and having indexer ask you a small number of questions, specifically tailored to your data set to identify required file in the smalest possible number of steps. [Less]

No doubt, you have heard about FileLight. This is a KDE application that shows you how much disk space you use in a nice radial view. And surely you are familiar with the Model/View classes in Qt 4. The separation of content and presentation can ... [More] give you nice results. Once you start to build up a selection of models and of views, you can access and present you data with variety. A while ago I submitted a nice QAbstractItemView class, called WarpTree to the Qt programming contest. Now I've ported the Qt3 code of FileLight to the Qt 4 Model/View framework. Have a look at the code. [Less]

aKademy 2007 starts on the 30th of June. I'm honored to give a talk there on the progress of Strigi and it's use within KDE4. Till then, I want to achieve some goals to have a nice Strigi version to work with and to show off while at aKademy. Here's ... [More] a list of things that should be done by then: Minimal Xesam implementation strigicmd nicely working for index management All unit tests passing No more bugs This should be doable, because most of the ambitious stuff is already in there and working. Now we need to make this power accessible to power users via strigicmd. In addition, Xesam support is important. This young effort to standardize on a common ontology and search query format for the free desktop search programs out there, is very important. [Less]

Within several days Strigi ontology will be integrated with Xesam. Xesam is going to be the core of Strigi with extensions if/where necessary. I'll do my best to the most useful ideas from all participating Indexers and Nepomuk draft specification into Xesam core. Stay tuned.

You can find a small guide for building Strigi on a fresh installation of Kubuntu feisty fawn. The guide can be applied also to other debian-based system. You can find it here.