3 ms·
That's unfortunately so true. The thorough and strongly regulated engineering results in good quality for mechanical products, but is detrimental for the fast p
by ascar 6y ago
That's unfortunately so true. The thorough and strongly regulated engineering results in good quality for mechanical products, but is detrimental for the fast paced software world. SSL3.0 is from 1996. That's the life cycle of 2-3 cars and much less than a building, so "not a long time ago" in engineering terms "when it still works".
I joined a German engineering company and made the naive assumption that proficiency in engineering is correlated to proficiency in software engineering. I will not make that mistake again.
The reality is that the classical engineering disciplines (mechanical, civil, etc.) and software engineering are very different in practice. While waterfall/V model is discouraged in software, it's a very valid approach for physical goods with high iteration costs and nearly no option of upgrade after sale.
- kspacewalk2 6y agoHaving to deal with Siemens healthcare software - they take the very understandable risk aversion (some of which is a must when it comes to medical equipment) and take it so far that it seriously hampers their ability to keep up with the industry, and (ironically) to maintain a stable product. I think it's their North American software teams that keep them moving forward, otherwise they'd still to this day be selling new MR and CT scanners running Windows XP. And we all know the problem with technical debt - it's not a problem until one day it's a huge fucking problem that paralyzes you and forces you to stop moving forward at all for some time, just so you can pay off the debt.
- 1600878470 6y agoOh god... The Siemens modalities that I work on are garbage. Same for Phillips. I have mixed feelings about GE as well. Healthcare tech is definitely hampered by the speed at which software _should_ move forward. The cost of getting FDA approval is crazy, especially for simple changes/security updates.