That's a very narrow perception of the profession. There are developers making money all over the world, and hardware prices don't scale down according to national salary levels. Even in Western Europe, that's more than a month's salary for a medium-level developer in some countries.
I’ve got 16GB in both my 7 year old iMac and 4 year old MacBook Pro. I’ll probably get a new computer next year with 32GB of RAM. I don’t think an editor using 1G-2G matters to me.
I don’t currently use Atom but if they innovate faster and eventually build a better editor, many devs will choose it.
https://blog.codinghorror.com/hardware-is-cheap-programmers-...
I think I had this same conversation when someone told me that the average developer can't afford the $70 for Sublime Text.
I did, for the past five years, working as a Python dev in my country of origin (Portugal). And I occasionally went to interviews, and the offers were not better than what I already made, nor did most of my college friends make more than that.
You can imagine how much some developer from an actual poor country makes.
Who can't afford to buy 16GB of RAM? It's $130 on Amazon.
Prices are higher here, but in any case, that's the wrong question. Those $130 are worth much more to a person making $20k/m, or to a company paying that much, especially since many costs (not just hardware) don't drop with the salary levels (and in some cases are quite higher, e.g. gas prices), and so even the share of disposable income, not just the absolute value, is much lower. So it'd be absurd to waste that meager margin on RAM just to use a fancier editor.
If you save 20 hours a year then it pays for itself. Everyone's economics are different, and the amount of time saved may be different. In the United States, you only need to save a few hours of work to make the upgrade worthwhile.
Fortunately, Atom is free so we can evaluate it to determine if it's worth the hardware upgrade.
Of course it's nice when programs are fast and not hungry. Reality is that developers happily trade performance for power of abstractions.