There are a couple separate issues here, and they really do need to be handled distinctly:
1. How do you get useful tools? Nowadays, you can just download free interpreters and compilers for several languages, etc. This is still a somewhat new phenomenon. (I remember saving for a while to buy a C++ compiler, as a teenager.) This has gotten MUCH easier, IMHO. Starting with a free Unix helps, because it will likely have development tools pre-packaged. Windows doesn't have otherwise ubiquitous tools present, nor the sort of culture that tends to accompany them.
2. How do you recognize what tools are any good? This may be getting harder. Any good advice is drowned out by the sheer quantity of contradictory advice coming from blogs, the reddit echo chamber, etc. It's not easy to figure out the implementation quality, availability of libraries, and other relative merits of different languages until you're already at least competent with one. (Python still seems to me like one of the better choices for a first language.) I imagine that a newbie being sold on [insert language] will probably just end up more confused.
3. There are numerous protocols, standards, etc., and it's hard to figure out which matter. JSON? XSLT? TLA? Learning which of these actually matter (and, really, which will still matter in more than a year or two) takes perspective that new programmers don't have. I'm not sure if this aspect has gotten better or worse, since a lot of old stuff that seemed crucial at the time has likely just been forgotten.
1. How do you get useful tools? Nowadays, you can just download free interpreters and compilers for several languages, etc. This is still a somewhat new phenomenon. (I remember saving for a while to buy a C++ compiler, as a teenager.) This has gotten MUCH easier, IMHO. Starting with a free Unix helps, because it will likely have development tools pre-packaged. Windows doesn't have otherwise ubiquitous tools present, nor the sort of culture that tends to accompany them.
2. How do you recognize what tools are any good? This may be getting harder. Any good advice is drowned out by the sheer quantity of contradictory advice coming from blogs, the reddit echo chamber, etc. It's not easy to figure out the implementation quality, availability of libraries, and other relative merits of different languages until you're already at least competent with one. (Python still seems to me like one of the better choices for a first language.) I imagine that a newbie being sold on [insert language] will probably just end up more confused.
3. There are numerous protocols, standards, etc., and it's hard to figure out which matter. JSON? XSLT? TLA? Learning which of these actually matter (and, really, which will still matter in more than a year or two) takes perspective that new programmers don't have. I'm not sure if this aspect has gotten better or worse, since a lot of old stuff that seemed crucial at the time has likely just been forgotten.