Quality of Data

I have been looking in the market to change my job in last few weeks and have faced very interesting stuff that is worth writing a book, at least a short story about.

I have seen a christian who wanted to develop the ultimate database inspired from Jesus, have been to interviews where the company has not actually had a position yet, and I have faced people who think are super smart senior consultants and know how to avoid an object being finalized in garbage collector.

If you are one of those super smart senior consultants who knows everything see if you can answer this very basic classic database question. If you can’t you should seriously rethink about your title. Seriously, I can’t  believe someone developing business application for more than a decade and not being able to solve this problem in an interview:

Q. We have a couple of tables in a database. First table has column X and second table has column Y. First table has a foreign key relationship to the second table. What do you do to calculate “SELECT TOP 1 * FROM X JOIN Y … ORDER BY a*X+b*Y DESC” without a full scan on any table; assuming a and b are arbitrary positive numbers provided by user each time? You don’t necessarily have to solve it with only SQL and you are allowed to use C#. Above query is only for defining the problem.

Justification: Why I think this can be a valid interview question? Two reasons: 1) Think how many time you have faced a client complaining from performance and how many time you where in a situation that knowing how to avoid an object from being finalized makes the customer happy? 2) How easy is it to get the knowledge of algorithms for fast data processing on site in contrast to some rare technical terminology? Google avoiding an object from being finalized and you quickly get your answer in a few seconds but I would be interested to know how you use google to solve this problem?