jhandl

Also Flaptor was a search consulting company five years before the startup.

You think he started a search as a service company because he was targeting LinkedIn?

Flaptor (the name of the company, IndexTank was the product) was a search consulting company five years before becoming a startup. I know, I was there.

Take a look at this: http://resume.linkedinlabs.com/

The Searchify team grocked the code really fast and already started improving on it.

Would you have your children stay with their 1st grade teacher all through elementary, middle and high school?

"untangle from IndexTank"? The whole point of this post is to remind people how trivialy easy it is to switch from IndexTank to Searchify. They are the same product.

Beware: Windows only. I know, it should be obvious, but it's been a long time since a large company dumped an .exe onto my downloads folder without even checking what OS I'm on.

I'll stick with X-Plane. It works on all major platforms and it's an awesome simulator. I miss the ease of use of the old MS-FS, though.

As one of IndexTanks developers, I think this is awesome. Good work!

I believe it is not uncommon for businesses to rely on plans.

It is our intention to open-source what the next provider will need to keep the service running. It's too early to tell what that exactly will be or how it will complement other SNA projects.

Gracias! :)

I'm reading. :)

We are going to do exactly what you described. If everything goes as planned, IndexTank will continue to live on.

The Heroku plugin will not go away, neither will the IndexTank service if everything goes as planned.

"We can launch a short note that appear to recover." Yup. That's what it looked like.

IndexTank is a hosted service, you don't have to build, configure or manage your own search infrastructure.

An overview of full-text search options, comparing Google's offers, WebSolr and IndexTank.

For the contest you get a $300 Premium plan for free. You can see the pricing table at http://addons.heroku.com/indextank

If HN ran on Heroku, it would be trivial to add search to it...

Sphinx does not store data, so you have to get it from your database to display results, it doesn't have query auto-suggest, etc.

But most importantly, you don't have to deal with installing, configuring, managing and scaling a sphinx setup. IndexTank is cloud-based, you don't have to deal with servers, RT indexes, hadoop, and so on. We worry about all that.

Their problem, as I see it, was mainly that they were using Solr, which requires an inordinate amount of time to configure and maintain (time they didn't have), and also isn't good at handling frequent updates, like the up and down votes.

No, the old search was Solr based, this one uses only the lucene index file structure. Not even the scorer is lucene-based.

I think the bottom left quadrant should be rounded: if a website doesn't scale at all, there's no way it can have a lot of users. I wonder if there are other constraints like this one, and what shape the graph would then take...