Friday, March 30, 2007

Update


It's been a while since I've posted here. It is hard for me at this point to justify spending 10 minutes to blog when I only have an hour or so for LearnIt development in my schedule. I realize that more people see the blog when I post more frequently but the number of visitors to the site is pretty small at this point whether I post or not. Despite all this, I will try to be more consistent in posting so that anybody that is interested can stay updated. I hope to also talk more about Freebase (see earlier posts).

Right now I'm working mainly on the LearnIt design and page structure. I think I'm making good progress. I think LearnIt is going to turn out even more useful than I thought. My hope is to have some kind of public version to launch this summer but, who knows, I could finish much sooner or much later. Once I have a working alpha I will probably let those who are interested test it out. I'll try to keep you posted on my progress.

Thursday, March 22, 2007

Finally awesome search in Safari: Inquisitor 3


I stumbled upon a link to an awesome Safari plugin called Inquisitor. It is a fast, small, super-helpful, and free search tool that sits right where the old google search form was. Its got auto-completion, results on the spot, and you can even customize it so that searching your favorite search engines are just a keystroke away. The fact that the developer is giving this away is pretty cool.

Wednesday, March 21, 2007

Using Freebase in LearnIt

Right now I'm using my LearnIt development timeslot to brainstorm ways of using the Freebase API in LearnIt. There are a lot of possibilities and I am open to suggestions but I think I've found some good ways already.

Tuesday, March 20, 2007

Resist the urge to make that mockup button perfect


I just spent a valuable 45 + minutes on something I have done countless times and hope to never do again. I, once again, could not resist the urge. I just couldn't let that button look so ugly... It had to be perfect. But now there is a problem. I can't use it anymore. I have decided to implement the feature in a way that doesn't require that button. 45 minutes... Gone.

Lesson learned: shinyness can wait til the final version.

Sunday, March 18, 2007

Using Freebase #2: auto-completion

Freebase has a feature that makes adding meaningful information very easy. This feature is auto-completion. The first screenshot below shows me editing my country of nationality. As I type Freebase suggests topics and tells me their type. The second screenshot shows that, if I mouseover a type, Freebase will give me the description of that type (from either a Freebase description or part of the wikipedia article on that topic). Everything loads pretty fast so it is surprisingly helpful.

Freebase and Safari

Though Metaweb isn't supporting Safari at this point (as a dialog box will warn you before entering Freebase), Freebase actually is usable in Safari. The screenshot above shows Freebase running in Safari. I haven't tested it extensively but so far I haven't seen much more than a few layout quirks when running in Safari. So the developers just haven't had the time to completely support Safari and make sure everything works. I suspect that Safari support will be added in future versions (hey its only in alpha, give 'em a break).

Also a little myth that has been running around the web a bit is that you have to have a Freebase account to use Freebase. This is true for the alpha as they need to iron things out before public release but non-registered users will be able to use Freebase when it is ready. (This is according to the Freebase FAQ.)

Using Freebase #1

I've been using Freebase for a few hours and have been pretty impressed by it.

My first impression is that a lot of thought has gone into this thing and there are some smart people behind it. Also, though there are some bugs here and there (its an alpha) the interface is very nice. It is really easy to use. They have a lot of neat (and helpful) AJAX features like auto-complete that are very usable.

On the API side I think they've done a pretty good job. Even in the early stages they have provided a lot of documentation on the API as well as a really neat query editor. If your the kind of person (like me) who learns something best by actually using it, this feature is really great. The screenshot below shows what I'm talking about. You can enter queries and see results and there are a lot of examples that show the syntax.

(Click to enlarge)

Look out for more reports on Freebase as I continue to use it.

Gotta catch a Freebase alpha account

For the past few days I've been scouring the web for information about Freebase (see my last post). This is due not only to the role Freebase will play in the future of the web but I am very interested in the API for use in LearnIt.

I think I am close to getting an alpha account (see the comments under this blog post). If I do I will share my thoughts here.

From my research I found a few blog posts that are worth mentioning.

One of the first articles about Freebase is O'Reilly's blog post about it (with pictures).

Another interesting one is from Ilya Lichtenstein, a fortunate blogger who got a alpha account. In his blog post he writes about the semantic search capabilities of Freebase and includes some pictures.

UPDATE: md so kind as to invite me to the Freebase Alpha. I'll be examining Freebase and checking out the API. So far it is looking really good. I'll write up a more detailed report when I have spent more time with it.

Tuesday, March 13, 2007

Freebase

Freebase, currently in alpha, is a open-structured-free-online-searchable knowledgebase created wiki style. It is the first use of technology Metaweb has created to "allow extremely large volumes of data to be collected, organized, connected and modified by a community of contributors, developers, publishers, and people who just plain like data". The freebase developers are going to be posting about how Freebase works in their developer blog which should be some interesting reading.

I am not sure exactly what to think about freebase. If it catches on then it could open up a lot of new opportunities for search and web applications. Actually, just a few weeks before freebase was made public, I had a very similar idea as part of the backend of LearnIt. For the past few weeks I have been thinking a lot about how to do something like Freebase. So Freebase may effect LearnIt.

Here are the questions floating around in my mind:

Should I go ahead and try to develop my own Freebase-yet-kind-of-different backend for LearnIt? Or should I just jump on the Freebase bandwagon, whatever that may be... Which leads to another question.

Whats the deal with the Freebase API. How does the API for giving them information work and how does it work to extract information. And are there any strings attached... Which leads to my last question.

What is the Freebase business model? They got funded, but how is this going to make Metaweb profit? As far as I know there aren't any ads. They say the API is free. But then somewhere else they mentioned that some fees may apply (now I can't find the link).

UPDATE: Freebase also has a non-developer blog.