4 ms·
A similar problem occurs even when you control all the hardware. I work for a company with tens of thousands of servers, and the number of strange things we ru
by cortesoft 11y ago
A similar problem occurs even when you control all the hardware.
I work for a company with tens of thousands of servers, and the number of strange things we run into is crazy. The exact same software running on the exact same hardware, and sometimes it just behaves differently.
We like to think that computers behave perfectly predictably, but so much randomness goes on.
- rachelbythebay 11y agoTwo threads, on one processor, out of dozens, out of hundreds of thousands of machines. Those two threads would do exp() in the wrong direction, somehow, but only with a certain byte pattern. A byte pattern that corresponded to some floating point number we used in a few places. On that machine, everything using that library would stall. But only when it ran on that processor. Everywhere else, it was fine. And yeah, it sucks. And it happens all the time. Computers are hard.
- jschwartzi 11y agoI've had bugs relating to EMI issues crop up in testing before. The device in question had a compressor mounted inches from the touchscreen. It turned out the EMI from the compressor would generate phantom touch input on the display. It looked like a bug until I added a bunch of debug logging to the device. Even if the computer is behaving predictably, the environment can do a wonderful job of making it go haywire.
- cortesoft 11y agoYeah, we sometimes forget that computers are physical machines that can be affected by their environment.