4 ms·
Except for software engineering where doing for 25 years doesn't make me feel like I am doing magic. It feels like the art keeps moving and I need to keep up. L
by nejsjsjsbsb 2y ago
Except for software engineering where doing for 25 years doesn't make me feel like I am doing magic. It feels like the art keeps moving and I need to keep up. Like the card tricks I did back when are now obsolete so I need to learn human cannonball as table stakes.
- ema 2y agoIt's important to keep in mind that doing magic doesn't feel like doing magic from the inside. Our own competence becomes invisible to us. You know the details you didn't get right. You know the tradeoffs you've had to make. You know the decisions you've made based on incomplete information. It takes someone else to be in awe of you doing something they wouldn't even know how to begin learning to do. It takes someone else to know in their gut that what you're doing is great because they themselves have dedicated a lot of time to the craft.
- koiueo 2y agoAnother great person said something like: sometimes 5 years of experience is just 1 year repeated five times. I'd guess, less than 1% of engineers in the industry have an opportunity to learn something drastically new at least once a year. Most are doing the same in terms of engineering and just occasionally learn new (not better) tools (that's the only thing that pops in my mind in response to your observation about art moving forward). On the other hand, considering your 25 yrs of experience, I'd guess just you understanding networking or compilation/linking process will make you look like a magician to 99% of SW engineers outside of your bubble.
- nejsjsjsbsb 2y agoI think 99% of what I learn is temporary intracacies of how closed source codebases work lol! I am doing some deeper dives on reusable knowledge though now. On my list is learn etcd (at a code level) for some reason I am drawn to it. It feels like a microcosm of what bigger complex systems like Kubernetes would be like.
- chthonicdaemon 2y agoI have the opposite experience all the time, perhaps because I've spent a lot of time with early-career programmers. I recognise a problem and say "oh, that's a maximum flow problem", then formulate and solve it in a few lines and they are amazed in similar ways as Penn describes. They're like "how do you know all these things"? And it's just like the buried cards. I've taken all this time to read books on algorithms and solve Advent of code and so on, so my whole mind is filled with these buried cards. Compare yourself to a person with no experience in software engineering at all and you will quickly understand how many you have, too.