archive about

Tapir is a search engine for your site: a great idea with a nice implementation.

Tapir is a search engine for your blog. They say it’s for “blogs made of static pages”, but I can’t find a good reason why it won’t work for any site -or even any set of information you could fit in a [link, title, summary, content, date] form.

It has a simple API, and a nice jQuery plugin.

I’ve been searching for something like this for some time now. I use a Google Custom Search box in picks.vrypan.net, but I’ve got the feeling that Google is not so much into it anymore: they have limits, a paid version, the whole thing feels like the service may not be there for long. Plus I have little control over what they index and when they do it. And why on earth should I need Google or Microsoft in order to have a simple service like this?

I have the feeling that a simple search engines like Tapir, will be more and more in demand and I would expect more to show up. And now that I think of it, it’s relatively easy, as some of the most resource consuming tasks of a search engine are solved in this case: you don’t have to worry about crawling the internet (the site notifies the engine of new content), and you don’t have to fight spam (I only search the content I’ve submitted).

The basic building blocks, like elasticsearch (and soon full-text search on AppEngine) are out there. Of course, setting up a reliable service, giving a well designed API to developers and creating a nice UI (like Tapir does) is the extra mile that makes a difference between similar implementations of the same service.

It could also be the kind of service offered by hacker communities. I could see LUGs or hackerspaces offering a personal search engine to anyone interested -for free or a small charge.

And why not, also offered as a payed service by numerous providers, much like web hosting is today. I’m not sure if Tapir is just a side project or looking forward to offer a apyed service, but why not?

It’s the kind of user generated infrastructure, we should be moving towards.