One of the anons from the other thread.
I'm not going to jump in too much if other's are doing something where we end up stepping on Pepe's toes. Couple of thoughts though…
-
Full text searches/indexes can be garbage. Only good for reserved words
-
Most likely want this in a relational database. Creating the schema would consist of a really simple data model. Not even sure I would worry about normalizing it.
-
Messages (body) could be stored in a blob and be searched with wildcards.
-
Only looking at about 10-15 different queries tops. All simple SQL statements except for a couple that would need to be hierarchical..but still easy.
-
I was thinking to use MySQL or SQL Server for the DB Engine.
-
Biggest challenge will be the parsing of the threads and crumbs into a loaded format for the database. Once in a useful format…loading will be easy.
I see three main parts to this:
1) Getting the data so it can be loaded into a database.
2) Creating the database structures (really should be first)
3) Spitting out the queries, views, and sprocs that will be used. And putting a front end on it.
-
almost doxxed myself and put a link to my web site…so close :-)