3 ms·
Sometime in my late 40s I got really tired of learning the "next great thing" only to realize that some young developers had reinvented the wheel. Again. Since
by bradley13 3y ago
Sometime in my late 40s I got really tired of learning the "next great thing" only to realize that some young developers had reinvented the wheel. Again.
Since then, I have learned new things, but not because they were "relevant" or "important", bit just because they were fun.
Python is a crappy language, and always has been. It's just a happy accident that the ML folks glommed in to it. Most frameworks suck, seriously, they are awful, victims of their own success.
Learn a clean, new language. Use it to solve AdventOfCode, just for fun. Get back to the fun in programming :-)
- the_only_law 3y ago> Since then, I have learned new things, but not because they were "relevant" or "important", bit just because they were fun. How do you maintain a career? I stopped caring about relevant and important and my career suffered for it. I stopped caring about fun because it was either a waste of time or inaccessible.
- jasonm23 3y agofind balance, it's not all or nothing.
- granshaw 3y agoThis is what I hate about this industry at this point in my life. Just this morning I was fantasizing about being a dentist instead and just be able to work for as long as I like, not worry about ageism, and otherwise live my life, without needing to constantly “keep up”
- vineyardlabs 3y agoI think this is kind of a fantasy. My impression of dentistry and the medical field in general are that there are constantly new procedures coming out to learn, new techniques to implement, and new tools to train on. Maybe ageism is less of a thing since a lot of dentists are self employed but if the dentist down the block is cheaper/better/less painful you're going to see less revenue. Similarly you could find a job maintaining ATM software in 90s style Java but your earning potential will probably be impacted. The grass is greener where you water it.
- johnnyanmac 3y ago>Python is a crappy language, and always has been. It's just a happy accident that the ML folks glommed in to it. Is it? I can't imagine making any large scale code with it, but the purposes of being a nifty and intuitively designed scripting front-end seems to work. A lot of the research industry still uses Matlab, so I shouldn't be too surprised Python popped up right behind it.