DEV Community

Discussion on: TIME TO SAY BYE BYE DOCKER !!! Era of Docker is over...

Collapse
 
kovah profile image
Kevin Woblick

I know podman for some time, but never moved to it and I guess it's the same for many other developers ot there. The problem: it does not natively support Windows or macOS. Docker provides apps for both platforms and they run quite well. But podman just states that "you need to have some Linux VM running and you can use this tool to remotely manage podman inside the VM". There is no help, no tutorial for setting up this stack.

Collapse
 
anortef profile image
Adrián Norte

This is one of the main reasons for not switching.

Collapse
 
mburszley profile image
Maximilian Burszley

This might be a maturity thing. Docker started out the same way - until around 2016 when they made Docker Desktop run native w/ Windows (versus piped through a VM using docker-machine).

Collapse
 
manishfoodtechs profile image
manish srivastava

+1 yes Max, you are right

Collapse
 
g105b profile image
Greg Bowler

Is this true for WSL? It's probably not long until Windows is running a full Linux kernel anyway.

Collapse
 
manishfoodtechs profile image
manish srivastava

With the latest Windows Subsystem for Linux (WSL) 2.0 implementation, Podman now works out of the box.
redhat.com/sysadmin/podman-windows...

Collapse
 
manishfoodtechs profile image
manish srivastava

+1 This is a real problem with windows and mac OS desktop users. However, they can use Podman as suggested by Kevin through Linux VM or remotely. Very nice, Kevin. However, here is some instructions about podman podman.io/getting-started/installa...

Collapse
 
jdickey profile image
Jeff Dickey

Why would I want to load another layer that can fail (a VM manager) when alternatives exist that can be run by every developer in my org, whether they're on macOS, Windows, or Linux? Why would my org want to put a big question mark on whether we can fully use Docker images published on Docker Hub or on our own internal repository? Docker isn't perfect — parsecs from it — but it's been a good three years since we ran into a previously-unknown unknown that couldn't have a solution pointed to with a few minutes of research?

Twenty years ago, I was as bleeding-edge as any hacker in love with his own believed ability to solve any problem just through sheer awesomeness. I got fired from my dream job because I kept installing Mandrake (what would later be Mandriva) Linux when shop and customer standards required Red Hat. ("It's compatible, isn't it?" No, not always, and the differences aren't known widely enough and solutions aren't trivial enough to add the risk and expense of deviating from standards.) It took me five years before I even understood that — by which time I'd been writing software for a living for 25 years.

Standards matter, and they tend to remain standards for real businesses until and unless a new tech can be shown to be either a) so completely frictionless that it doesn't matter technically which of the two you're using (and if that's the case, why bother switching?), or b) so revolutionary and so clearly better at solving real-world problems (beyond appealing to the Kool Kids™) that the risk and cost and pain of migration is outweighed by the demonstrable benefits. (That's how Docker, and Linux, and PCs, earned their places in the business world.)

No disruptive technology is ever fully compatible with its predecessor, and no fully compatible technology wins against an incumbent unless business reasons make that incumbent unavailable. (I worked briefly for Silicon Graphics back in the day; they didn't die because somebody out-competed them on technology.)

Docker has been one of the two or three greatest leaps forward that I've seen in the last 40 years — not because they're perfect, but because they're good enough and because they solve a set of use cases with lower risk than any direct competitor with similar reach. When that changes, and it will one day, then there'll be a new top dog. But until then, rotsa ruck moving the installed base.

Thread Thread
 
manishfoodtechs profile image
manish srivastava

+1 True Jeff.
-Standards matter.
-No disruptive technology is ever fully compatible with its predecessor, and no fully compatible technology wins against an incumbent unless business reasons make that incumbent unavailable.
-Docker has been one of the two or three greatest leaps forward that I've seen in the last 40 years — not because they're perfect, but because they're good enough and because they solve a set of use cases with lower risk than any direct competitor with similar reach.

Here is one article : Here's why podman is more secured than Docker - DevSecOps
cloudnweb.dev/2019/10/heres-why-po...