Re: [bailey-developers] Introduction

[Mike K. <mik...@gm...>]

Update: Yonik answers my questions via private email, just in case  
anyone else want dying to jump in there.

Sounds exciting,
-Mike

On 27-Feb-08, at 3:01 PM, Mike Klaas wrote:

> This is list is still small enough that I feel that I should  
> introduce myself.  I'm a Solr committer and CTO of a small internet  
> search startup that is still in stealth mode.  Yonik pointed out the  
> project to me.  It sounds quite interesting, not least of which  
> because our company has built a large Solr-based search cluster  
> running on EC2 with at least some of the properties you are aiming  
> for.
>
> The one thing we don't have is dynamic automatic failover and  
> replication.  One reason is that it's hard, and the second is that  
> for us, it has always been better to use more boxes for a larger  
> corpus rather than replicating a smaller one.  Instead, we store the  
> indices in S3 and restore from backup when a machine fails (which is  
> not rarely).
>
> I do have a couple questions about the project:
> - it isn't clear to me how the goals are substantially different  
> from those of nutch.  Is it mostly in the relaxation of its  
> application to web search?
> - why sourceforge rather than an apache project?
> - it sounds like the intent is to build upon Solr, which I think is  
> a great idea, but isn't mentioned in the top-level goals section.   
> Is Solr an optional component?
>
> cheers,
> -Mike