My Apprenticeship - Monday, July 12, 2004
This summer I'm revisiting my short apprenticeship at Object Mentor. I'll be posting commentary on all my posts from the summer of 2004 exactly 5 years later to the day.
Today was a day of unintentional diversions. Paul and I came up with an idea of how to test an area of our state map compiler that had been eluding us. We would build up an input file, in the tests, and then send that input file into the SMC which would create an output file that we could read and check to make sure it works. If we had been designing this SMC thing from scratch using TDD, we could have made it easier to test, but since large parts of the program are a legacy, we had to resort to the building/writing/reading file method. But the parser in the SMC program doesn't like to be called twice, so a fair amount of the day was spent figuring out how to get around that for multiple tests. The other part of the day was spent writing a bunch of tests that had to be thrown out because there's a quirk of the program that given the same input, you may get different output. Everything will be there, but the order is not guaranteed. This is one of the problems with using a list, things may not come out in the order they were added. So our tests, which assumed a certain order, would randomly fail. Great. Luckily, there was only one test that really needed our new method of building/writing/reading so we could move all the old tests back to the old way (we called the program directly to set variables) and, in the one straggler, we re-wrote the test so that order was unimportant.
Pretty much the same thing was going on in the office. The email wasn't functioning and everybody who knew about the mail server was either out of town or teaching a class. Then Quickbooks started malfunctioning and things typed in by one person were not available to someone else in the system Nobody was sure if data might be getting lost -- this had the effect of chilling productivity in the office.
In short, everybody's 9am plans never even got started.
"If we had been designing this SMC thing from scratch using TDD, we could have made it easier to test..." How many times over the last five years have I run into this exact same problem? Too many to count -- lack of test leads to untestable code which is very often unmaintainable code.Oh distant-past-jake, why didn't you just loop over the list you were trying to test and see if the things you wanted were in there? Ah well, a testing technique I would soon learn the hard way