Author Topic: Old bugs  (Read 2448 times)

0 Members and 1 Guest are viewing this topic.

Offline Jim

  • TM Collaborator
  • Hero Member
  • *****
  • Posts: 2732
  • Last Login:Today at 02:09:10 pm
Old bugs
« on: July 06, 2016, 05:40:23 pm »
I just wanted to mention/complain that I came up with yet another example of why one should come up with good test cases for their code.  An important part of the site update code that allows it to run with efficiency that allows our scale of data to be processed in a reasonable amount of time is the quadtree data structure used to store the collection of all waypoints.  And it's been working for over a year, I'm sure.  Well, just by luck.  I spent most of today tracking down what I thought was a problem with my new near-miss point code (which also would be hopelessly inefficient without the quadtree) only to find out that it was the old code that was the problem.  I had not one, but two bugs in the old code, but since the bugs were consistent with each other, we were still getting the right answers.  Basically, I mixed up east with west and north with south in a couple places.  Turns out my new code was correct but was using a bad data structure.  I only wish I thought to look more carefully at the old code a few hours sooner.

Thank you for listening....