There’s a project that I’ve been neglecting for many years.

The midrange.com archives needs a proper search engine.

Many years ago I used a search engine called ‘mnoGoSearch’. Unfortunately, after an upgrade, that software stopped working. IIRC, it had to do with the implementation of SSL on the archives web site.

As I’m now retired, I have plenty of time to work on this project.

Requirements

I had some pretty specific requirements for the search engine.

• Easy to implement
• Be able to handle incremental crawling.
• Self contained (no or limited dependencies on external services, like databases)
• Open source with active development

Research

I investigated a number of search engine backend’s …

• Elastic search
• Solr
• Nutch
• TypeSense
• OpenSearch
• Meilisearch

Hosting Restrictions

I run my servers on Amazon Web Services (AWS) Lightsail.

These servers aren’t really intended for compute or IO intensive operations, so the search engine had to be able to run ‘nicely’ (iow, no heavy io compute resources).