6 ms·
...which seems to be completely ignored in the world of tech careers, where "adopting to change", "constant re-learning" and "pursuit of the next big thing" see
by smikhanov 11y ago
...which seems to be completely ignored in the world of tech careers, where "adopting to change", "constant re-learning" and "pursuit of the next big thing" seem to be the norm.
I'm being overly dramatic here, of course, but what described in an article is a very good advice for junior programmers. Don't abandon your accumulated experience in a pursuit of the newest and shiniest -- today's hottest buzzword can absolutely stop being relevant tomorrow. Aim for "knowing where to put the chalk mark" instead (http://www.smithsonianmag.com/history/charles-proteus-steinmetz-the-wizard-of-schenectady-51912022/ http://www.smithsonianmag.com/history/charles-proteus-steinm...).
- henrik_w 11y agoIn programming, there are themes that are the same, regardless of language etc, that you keep and develop even as you change the stack you are working in. That's what I tried to convey in "Programmer Knowledge" http://henrikwarne.com/2014/12/15/programmer-knowledge/ http://henrikwarne.com/2014/12/15/programmer-knowledge/
- Chris2048 11y agoIrrelevance of language works if you assume PLs to be more consistent, or abstractly-used than they are. As soon as you need to squeeze out more (or even a little) performance, you need to know how the implementation works under the hood. As soon as the languages abstractions begin to leak, you need a knowledge of the languages quirks and tells. It takes a while to develop this experience, and there isn't necessarily a great deal of overlap between PLs.
- paulddraper 11y agoEven with implementations, there are consistent themes. L1 caches have been around aa very long time.
- Chris2048 11y agoSometimes knowledge like this is a little first-past-the-post, all-or-nothing. Completeness of knowledge is key, especially when a chain of reasoning (debugging) is as strong as its weakest link. At these times, a false assumption based on how something works in some other language could really wrong-foot you. Being familiar with some language means it's faster to learn other languages that have similar implementations, but not assumed.
- lsc 11y agoseems to me like the difference here is that photography (and other artistic fields) seem to be setup in a 'winner takes all' kind of way; you are the very best, or you are very little... Programming seems to be the other way around. Unless you go into the business side of things, after your first five years, you get better at what you are doing, sure, but your pay doesn't go up a whole lot. After your first five or ten years, you are doing pretty okay. so... I'd argue that because of this, jumping around makes a lot more sense in software than it does in areas like art where only the top X% or whatever of people employed in the field make a decent living.
- PhantomGremlin 11y agosetup in a 'winner takes all' kind of way Photography is what is known as a scalable profession. A quick search turned up this good explanation: http://casnocha.com/2009/03/scalable-vs-non-scalable-careers.html http://casnocha.com/2009/03/scalable-vs-non-scalable-careers... It's something I've been thinking about as I try to guide my children into selecting a college and major.