Comments on: Playing with RDF

By: John Carr

John Carr — Tue, 19 May 2009 17:20:17 +0000

@James Henstridge: Indeed, I was looking at Storm and SQLObject and others. I quite like the look of Storm.

By: James Henstridge

James Henstridge — Tue, 19 May 2009 14:05:33 +0000

The existing ORMs for SQL databases might be a good starting point.

With Storm, the last example could be written something like:

result = Store.find(Contact.firstname, Contact.nickname == ‘Jc2k’)

It would determine what tables need to be queried based on what you are selecting and what WHERE clauses you use.

By: John Carr

John Carr — Mon, 18 May 2009 10:13:24 +0000

@Natan Yellin: Wicked cool! I have a lot of ideas for improving them, but not much time to implement them. Feel free to poke me in #tracker (Jc2k).

By: Natan Yellin

Natan Yellin — Sat, 16 May 2009 21:01:32 +0000

That looks pretty neat. I’m working on a Tracker-based backend for GNOME Zeitgeist so I definitely will give your code a look to see if I can pick up any ideas.

By: John Carr

John Carr — Fri, 15 May 2009 16:54:42 +0000

@engla: As far as i’m aware tracker 0.7 should be pretty good too, but you’d be better asking in #tracker for specifics.

@Patrys: I’m certainly looking to similiar things for inspiration. Hadn’t thought of django though. Thanks for the pointer.

By: Patrys

Patrys — Fri, 15 May 2009 16:42:27 +0000

I think you might want to implement something similar to django.db.models.Q

By: engla

engla — Fri, 15 May 2009 15:28:32 +0000

that’s neat and all that. I’m still using tracker 0.6.6, which is a ridiculously fast C-written search system. I was really impressed when I found out about tracker. Now the new versions, 0.7 and that are long coming.. I’m afraid they are not so fast and lean and mean anymore. How is it? Does it use lots of memory or cpu. Do you (ever!) notice it running.

I’ve got an old mac, 1.2 GHz and I never notice tracker running on this laptop, a truly epic feat for this functionality.