Some context on my personal path followed by my considerations:
I've been trying to keep a professional career with parallel academic activities. I.e., I did my masters and Doctorate while working.
For both my dissertation and thesis I worked in domains and problems that lend themselves to logical formulations - and most of the implementations I did were in Prolog. I also had minor professional experience with Prolog. Focusing here just on the learning-the-language parts and disregarding all other activities:
* I _absolutely_ could have used that time to learn a more marketable language or framework. So that was a tradeoff, but -
* I most likely would have put only a fraction of those hundreds of hours into "marketable training". I'd have played games for most of them. Or just browsed history channels on youtube, etc.
* I did not feel like any practical skills deteriorated at all. In fact -
* I feel like learning Prolog in-depth was a valuable "exercise in learning". It might have helped me learn hard stuff for my job by making me a better independent student
- I hardly ever interview but I usually let Prolog be at most a footnote in my resume. I don't expect it to be useful in that way (but who knows?)
My takeaway would be - for marketable purposes, learning Prolog is only substitute for meta-training. Like, instead of reading another (not your first!) book about design patterns or asynch control or (...) you could learn Prolog.
But 85% of the benefit in learning it is the intelectual stimulation.
I've been trying to keep a professional career with parallel academic activities. I.e., I did my masters and Doctorate while working. For both my dissertation and thesis I worked in domains and problems that lend themselves to logical formulations - and most of the implementations I did were in Prolog. I also had minor professional experience with Prolog. Focusing here just on the learning-the-language parts and disregarding all other activities:
* I _absolutely_ could have used that time to learn a more marketable language or framework. So that was a tradeoff, but -
* I most likely would have put only a fraction of those hundreds of hours into "marketable training". I'd have played games for most of them. Or just browsed history channels on youtube, etc.
* I did not feel like any practical skills deteriorated at all. In fact -
* I feel like learning Prolog in-depth was a valuable "exercise in learning". It might have helped me learn hard stuff for my job by making me a better independent student
- I hardly ever interview but I usually let Prolog be at most a footnote in my resume. I don't expect it to be useful in that way (but who knows?)
My takeaway would be - for marketable purposes, learning Prolog is only substitute for meta-training. Like, instead of reading another (not your first!) book about design patterns or asynch control or (...) you could learn Prolog.
But 85% of the benefit in learning it is the intelectual stimulation.