NASA can't figure out what's causing computer issues on the Hubble telescope

Re:NASA can't figure out what's causing computer issues on t... by ravel-bar-foo on Thursday June 24, 2021 @09:16PM (#27626052)

I discovered this in my PhD. I spent 3 months trying to automate a process (running a finite element model atop very fragile software platform). I couldn't figure out how to automate my way around the errors the system would give me, every run was unpredictable in duration due to the errors, and I ended up spending a lot of time holding the system's hand.

Eventually gave up on the automation and resolved to spend a few weeks clicking every 15 minutes to complete the simulations. It was very hard to stay motivated, but I got two papers out of it.

The disadvantage is now my CV says I'm an expert at an FEM framework which I never want anything to do with ever again.

[ Parent | on HN ]

Re:NASA can't figure out what's causing computer issues on t... by xwolfi on Friday June 25, 2021 @12:23AM (#27627130)

Normally if your eyes and hands are synchronized to do an operation, it can be automated.

Sometimes problems come from trying to automate at the lowest level, when maybe screenshotting a vm and sending a click at a specific position if the pixel colors change to the error condition could be done a bit faster.

[ Parent | on HN ]