3 ms·
If you want to just knock about, just start doing it. If you're serious about getting people to use it, my personal minimum guideline is ten years experience o
by jerf 16y ago
If you want to just knock about, just start doing it.
If you're serious about getting people to use it, my personal minimum guideline is ten years experience of serious software development where you are continually pushing yourself, fluency in at least one major functional, logic/set, and imperative/OO language, and competence in about ten languages as diverse as possible, and some specific problem that can't be solved well by an existing language. By then you'll know what to do.
The reason I would say this is my usual spiel about not actually encouraging people to do really time-wasting things, even though we have a culture of "encouragement at all costs". Encouraging a novice mountain climber to take on Mount Everest as their third mountain isn't friendly, it's evil. Encouraging someone to write their own new programming language if they don't have the experience to do it of course won't kill them, but they sure can waste a lot of time. Whatever you think they may learn from the experience can be learned faster and more productively by first reaching the standards I wrote out above, and actually taking advantage of the fact that much ground has been mapped out already. We all must re-invent our own wheels if only for learning at some point, but this is a great deal more than a "wheel".
- motxilo 16y agoI sort of agree. The only snag is that you present gaining a lot of experience before tackling that task as opposed to start building up your own stuff. They are not exclusive by any means. Rather than encouraging a novice to climb Mount Everest directly by either North or South Ridge, have him start building his own low-steep slope path surrounding the mountain, with some books about mountaineering and maybe some friends that created their own roads a while ago. Maybe, in 15 years, he will reach the tip.