Comment by jconley | Hacker News Reader

Comment by jconley | Hacker News Reader

Parent Full thread

I wrote most of this app on a contract nearly 10 years ago. We used it as a mentorship exercise to train state worker programmers in the SCO department on the (then new) ASP.NET Framework and SQL Server 2000. The DBA was familiar with Oracle and mainframes, but not SQL Server. It was written in ASP.NET 1.1 and used SQL Server 2000. From what I can tell it hasn't changed much at all.

"Programmer and database types will notice one problem immediately - no fuzzy string matching."

Yeah, I really wanted to implement that. The problem was we had to use Microsoft SQL Server and no third party database/indexing systems. SQL Server's Full Text Indexing run took too long during the nightly import job on year 2000 commodity hardware. We also had to do all of this within the constraints of training state workers and handing off something they could actually maintain.

Ah, the joys of state contracting.

I see there are plenty of posts down below that are complaining about the lack of fuzzy search. By all means, crawl the site and create a better search engine. IIRC it is setup in such a way that new records are always appended at the end of the ID space. Just check every night for new things. At the time we wrote this there were a few companies doing just that, charging a small fee, and soliciting people that were owed large sums. I mean, you've got a $6B market, in just California, to play with where you could probably earn 5% or more as a finders fee to those that don't know this is public record, or might just pay you to avoid the hassle of the paperwork.

There was also a web service interface to this system that returned paged results as plain ole XML, but unfortunately I don't remember the URL. We had wanted to expand and create an interstate network by partnering with Texas and some others, but it never panned out for reasons I forget.

It is really nice to have the original author show up and comment on something like this.

After looking at this site I started thinking about ways of crawling the whole database. The 500-record limit would make an alphabet tree crawl a bit more challenging but not too bad. Do you think the state agency would be amenable to requests for bulk download?

[edit] just saw your post which answered most of my questions while I was typing them.

Reply on news.ycombinator.com