[cloakandswagger]

Store it in memory? I doubt the amount of data they provided for a test problem was so tremendous that it required a database.

In itself this is a good (dis)qualifier; if I give someone a problem and tell them the data will never exceed 10k rows, I expect them to be practical and not waste time setting up a database.

[brudgers]

The flat line in the bathtub curve is where given "data will never exceed 10k rows" implies no database. On the inexperienced end, people don't understand that 10k rows doesn't require a database. On the experienced end, candidates probably oughtn't to believe it.

[sokoloff]

Absent an instruction to specifically "do not use a database", I feel like even 10K rows suggests the use a database as the data structure and then the DB query interfaces as the interface.

I don't see why you wouldn't use a database for a problem that it's particularly good at addressing.

[mattzito]

As I think I said elsewhere, we allowed people to use a database if they wanted to (and could install it themselves on the dev VM we gave them), but then we would include looking at how they designed the schema, how much thought did they put into connection handling, etc.

But most of the people who successfully passed this didn't use a database, it just wasn't enough data to be meaningful - thousands of rows at most.

[fapjacks]

Yeah, or you could use the opportunity to demonstrate that you know how to roll your own data structures, provided you've clarified that the dataset isn't likely to grow. Instead of showing them you know how to install a database.

[zjaffee]

What? while sure you might not need a database if you have under 10k rows, if you want any amount of resiliency storing data in memory is a huge mistake as if the current instance of the application fails your done for.

[meggar]

It only displays the data, does not save any changes.