5 ms·

There are two type of knowledge: timeless, and changing specifics. API, libraries, and platform are specifics. It's something you learn and throw away (eventua

by kiba 14y ago

There are two type of knowledge: timeless, and changing specifics. API, libraries, and platform are specifics. It's something you learn and throw away (eventually). Continuous deployment, unit testing, debugging, basic programming knowledge, are timeless. One thing to keep in mind: Even timeless knowledge will be subjected to decay if you don't practice it.

gaius 14y ago
I'd say there were 3 kinds: the timeless stuff is actual computer science, y'know, the hard stuff: maths. The long-lived knowledge is things that have stood the test of time, e.g. languages like SQL and C (and its toolchain) - but I wouldn't gamble on C being around as long as say, the quicksort algorithm. C++ and Perl and Python probably fall at the outer edge of this category, we'll know in 10 years. People who know these things well won't have any trouble finding work in the near future, even if they don't put any extra effort in (but show me this kind of programmer who doesn't anyway, just because it's fun). Then there is stuff that you need to know right now which is really just the whims of fashion, the latest abstraction layer over Javascript (JQuery? Coffeescript? I lose track) or web framework and the latest fad methodology (Scrum, TDD or whatever). This is stuff that ages very quickly, in just a year's time your skills could be obsolete. I pick up just as much of these kinds of skills as I need, but there's no point investing serious time into them. These guys are running flat out just to keep up. It must be exhausting. 5 years mastering C is an investment in your future. 5 years mastering MongoDB probably isn't.
- mathgladiator 14y ago
  > 5 years mastering C is an investment in your future. 5 years mastering MongoDB probably isn't. Replace C with COBOL and MongoDB with MySQL. I've thought about this recently, and I think Java may replace C as the investment. There's so many companies now that depend on it, that knowing the JVM and it's... oddities is an investment. I think MongoDB now is a bet. It may actually pan out, but we will not know until 10+ years out.
  - jboggan 14y ago
    Hadoop will make Java more and more relevant.
    - nostrademons 14y ago
      Assuming something better than Hadoop doesn't come out. From what I've heard from folks who have used Hadoop and then used MapReduce, that's not really a safe bet.
  - gaius 14y ago
    If you go to an interview at a hot new startup in 5 years saying I know MongoDB and Coffescript they'll just chuckle and say sorry old timer, we only use UmbongoDB and Congoscript here.
antidoh 14y ago
"Continuous deployment, unit testing, debugging, basic programming knowledge, are timeless." The first two are relative new, although they are likely to be seen as timeless after enough time has passed. In fact in the early nineties I was chastised for practicing what has essentially become test driven development. I apparently didn't have enough confidence to write the whole thing at once and (maybe) test it later.
ExpiredLink 14y ago
The two types are knowledge in depth vs. knowledge in breadth. The former grows slowly, the latter exponentially.

by kiba 14y ago

gaius 14y ago

I'd say there were 3 kinds: the timeless stuff is actual computer science, y'know, the hard stuff: maths. The long-lived knowledge is things that have stood the test of time, e.g. languages like SQL and C (and its toolchain) - but I wouldn't gamble on C being around as long as say, the quicksort algorithm. C++ and Perl and Python probably fall at the outer edge of this category, we'll know in 10 years. People who know these things well won't have any trouble finding work in the near future, even if they don't put any extra effort in (but show me this kind of programmer who doesn't anyway, just because it's fun). Then there is stuff that you need to know right now which is really just the whims of fashion, the latest abstraction layer over Javascript (JQuery? Coffeescript? I lose track) or web framework and the latest fad methodology (Scrum, TDD or whatever). This is stuff that ages very quickly, in just a year's time your skills could be obsolete. I pick up just as much of these kinds of skills as I need, but there's no point investing serious time into them. These guys are running flat out just to keep up. It must be exhausting. 5 years mastering C is an investment in your future. 5 years mastering MongoDB probably isn't.

mathgladiator 14y ago
> 5 years mastering C is an investment in your future. 5 years mastering MongoDB probably isn't. Replace C with COBOL and MongoDB with MySQL. I've thought about this recently, and I think Java may replace C as the investment. There's so many companies now that depend on it, that knowing the JVM and it's... oddities is an investment. I think MongoDB now is a bet. It may actually pan out, but we will not know until 10+ years out.
- jboggan 14y ago
  Hadoop will make Java more and more relevant.
  - nostrademons 14y ago
    Assuming something better than Hadoop doesn't come out. From what I've heard from folks who have used Hadoop and then used MapReduce, that's not really a safe bet.
- gaius 14y ago
  If you go to an interview at a hot new startup in 5 years saying I know MongoDB and Coffescript they'll just chuckle and say sorry old timer, we only use UmbongoDB and Congoscript here.

antidoh 14y ago

"Continuous deployment, unit testing, debugging, basic programming knowledge, are timeless." The first two are relative new, although they are likely to be seen as timeless after enough time has passed. In fact in the early nineties I was chastised for practicing what has essentially become test driven development. I apparently didn't have enough confidence to write the whole thing at once and (maybe) test it later.

ExpiredLink 14y ago

The two types are knowledge in depth vs. knowledge in breadth. The former grows slowly, the latter exponentially.