Discussion
I'm told that VMs obviate the problem of having multiple developers in multiple environments. Since the VM the developer runs is the same as the VM every other developer runs, there can be no "works on my machine" problems.
It's a strong logical case.
But it ignores two important thoughts.
1) How much does it cost to do this?
2) Why doesn't the software run everywhere?
@GeePawHill 2) OVF done got nobbled (believe me I've tried)
@GeePawHill I'm not as big a fan on VMs as I once was - Podman containers and VSCodium running on your host and SSHing into your container is awesomely powerful and inexpensive.
1) The cost is not obscene, but it's not very cheap, either. The VMs are slow, and in spite of everyone's best intent, they are never quite as stable (or grokkable) in practice as they are in theory.
2) Because the developers lack the money, the will or the ability, to write software that can run anywhere.
And, after all, we're sitting on a money-printing machine, it hardly matters whether our work is any good, or our people funded, trained, or supported.
Just shut up and turn the crank.
I have been writing apps to run on *nix using the native Windows O/S for three decades.
I am, believe me, an inveterate and talented liar. But I won't tell you I've never been bitten by cross-OS bugs. Some have been Windows to *nix. Some have been *nix to *nix. Some have been * to MacOs.
Of course I have.
But then, I've also been bitten by language versions on the same OS, by library changes on the same OS, by *tool* upgrades on the same OS.
@GeePawHill Yep. All of these. Sometimes repeatedly. Sometimes because I actually forgot my lesson from the first time around.
One learns *technique*.
One learns isolation from one's libraries.
One learns TDD.
One learns debugging.
One learns dependency management--the last great open problem in computing--much against one's will.
VMs protect me, at some cost, from one stripe or flavor of the cross-X problem-set. But it's a rather *thin* stripe.
Most bugs aren't caused by brilliant rocket-science cross-X complexities.
Most bugs are caused by dumb mistakes right there in your own code.
Some bugs *are* caused by brilliant rocket-science cross-X complexities.
But, honestly, not very many. Yet the VM answer, which resolves only one thin slice of the cross-X, costs us every day all the time.
@GeePawHill 1. a fair bit. 2. sloppy practices, fortasse? #lang_en #lang_lat
Install bonfire.cafe
Get the full app experience
