3 ms·
A lot of what has been said here about software documentation and process is true of NASA flight projects, especially manned. Outside of flight, not so much. In
by rootbear 9y ago
A lot of what has been said here about software documentation and process is true of NASA flight projects, especially manned. Outside of flight, not so much. In some cases, it's a bit more like grad-school-project-that-escaped. An exception would be the modeling code used for things like weather simulation on supercomputers. The scientists need to know that their code actually matches the model they're testing and that things like numerical precision and error bounds are being handled properly. I'll add that the above is anecdotal, based on my experiences at NASA Goddard.
- mynameisnobody 9y agoYup. I worked in a satellite data office of NASA and there was no formal documentation process. Much of the code running in the facility had been written by a single developer over 20 years, and all of the knowledge left with him. Masterful Perl can be frustrating stuff without documentation or comments.