Found the KDE dev. 
Begone! Begone foul kreature!
1 Like
Didn’t PC-BSD use KDE?
BTW, I liked PC-BSD. It was my first time using Free–Net, Open, then PC-BSD, then quickly Open again (sorry :]). And I liked KDE; I could never run it 'cause my computers were always about 10 years old when I got them, but I used Qt to develop my first mobile app (when I thought electric-leashes were cool).
1 Like
Winnielinnie out here running workloads on pure memes.
Which for me means that a “generic app container” like jails in which you can roll whatever you feel like is the best concept of all.
SCALE is dead for me. Today it offers exactly zero features that would be an improvement over sticking to CORE. If CORE is abandoned, I am aboard with the most promising team to fork it.
I predict that this entire project of moving to Linux for no technical reason, then dropping all features that would have justified that decision, will be the end of TrueNAS and iX. Let’s talk in five years from now.
4 Likes
I agree that that’s been the pattern–every few years, iX decides that what they’ve been doing for plugins (or equivalent) is “too hard,” something else would be easier to maintain, so they’re changing. EE will be the fourth iteration of that–fifth if you count FN10. But is that attitude really what iX want their users to have? “I don’t have any reason to believe that anything beyond the core storage functions of this software will be the same five years from now”?
4 Likes
But, if Scale was really about Containers… (and scale out… ooops)
Then forget about the burn from FreeNAS/Core.
How many times have iX burnt all app users in scale?
Theoretically, if an App user installed an iX catalog app on angelfish… did iX do the needful to bring them along all the way to DF/EE?
3 Likes
Exactly my argument just an hour ago on the TrueCharts Discord. I perfectly understand their decision to pull the plug on TN SCALE support.
Similarly iX’ statement that CORE would be supported for “years to come” - well, if there is no update beyond FreeBSD 13 that will mean two years more or less.
“But the system will run just as it is” - yes, but I won’t be able to buy new (!) hardware that FreeBSD 13 will run on, if you do not move TN CORE to FreeBSD 14, 15, … 20 and beyond!
I do not intentionally put work into any platform that is known to go away in two years from now. I have been running my infrastructure on FreeBSD for 30 years and I expect to continue to do so for the next 20 or 30 years.
I have a 20 year history of using VMware - which is why the new license policy set by Broadcom is such a bummer. Proxmox looks quite interesting, suddenly. I similarly have 30 years of Cisco IOS on my back.
These are time frames for infrastructure products. Refactor my data centre every two years? Are you kidding?
3 Likes
If containers is all you/they wanted, then why spend the effort to reinvent the platform? Why not just use “containers” in FreeBSD?
EDIT:
Podman + jails for example.
“Want to have”, probably not. But “the attitude that cautious users will develop”, quite possibly.
Keeping expectations low (and storage redundancy high) helps to avoid disappointments…
On the positive side, what’s announced for Electric Eel should be rather easy to move to or from “docker in a sandbox/VM”.
I do. And I have repeatedly argued that FreeBSD jails and Solaris zones are the most advanced, secure, flexible container technology available.
Jails and zones are containers. Docker images are a an impression of a container built upon name spaces. Hence all the security problems and the more than just awkward networking.
2 Likes
Oh good. I’m glad I’m not crazy. I was beginning to think I was wrong about ‘containers’.
By Containers I meant OCI Containers, which is really a fancy way of saying Docker Containers.
And Docker Containers rely on the Linux kernel.
…then they really ought not behave in a way that give their users that attitude.
iX is precisely NOT in the business of developping a container platform. iX is vaguely interested in providing the opportunity for their users to have containers. FreeBSD jails may be the better Betamax technology, but Docker VHS are what app developers provide ready to use. So iX goes for something that runs docker images. End of story. That ship has sailed…
1 Like
Your prerogative as is mine to leave the platform and the community eventually. I am not interested in Docker. I am interested in improving FreeBSD.
Thanks for making that perfectly clear.
They’ve been porting “docker” for years (but as I understand, docker sucks). …They even have “a kubernetes”.
Again, if you want a better car door, you don’t reinvent the car.
You lost me on this, ingress on a docker is ports: 8008:80 - that’s it - simple… Ingress for microservices on a kubernetes cluster involves a load balancer or reverse proxy of somesort to redirect different urls or paths to different microservices. Not at all required here.
You can run openvpn or tailscale in a docker. You could also run codeserver or a landing page app like hubsite or homer, heimdall etc There is nothing inherent about k3’s that makes any of this better.
If you want to run a k3s container you have the official registry or truecharts. If you want a docker you have Docker Hub, linuxserver as well as gcr, quay as well as many developer sites. It’s much more open, you don’t need to wait for some guy to hack the publishers container up to work on this closed ecosystem
For the avoidance of doubt (too late, it seems…), I was solely commenting on what I understand of iX motives.
I’m not an IT professional.
I’m solely using TrueNAS (CORE) as storage.
As user, I favour the BSD command line and its stability over the chaos of the infinity-and-a-half of warring Linux distros. But I do not have the skills to improve either. And if we get into OS politics, I hate the GPL with a passion—almost as much as I hate dealing with GRUB.
I am not interested in Docker either, though I’ve spent way too much time trying to deploy TrueCommand and PiHole on a few platforms. I’ve learnt a few tricks along the way, which some would say is always a good thing, but was thoroughly unimpressed by both Anglefish and Dragonfish (including GRUB loosing its marbles almost every time I added or removed a drive; that’s annoying for a storage plateform…).
YAML is an absolute pain in the lower back. Who on Earth thought that making indentation, i.e. blank characters, a meanigful expression was a good idea???
But still I found OMV and its very plain docker-compose to be a lot more friendly than SCALE to deploy an arbitrary container—and that’s counting the unreadable error messages from OMV, the not-fully-in-order instructions to install docker-compose and the infuriating click-confirm-and-wait-and-WAIT-ON-EACH-AND-EVERY-CHANGE process… So my user experience fully supports dropping k3s to go for compose.
Even tough, eventually, none of these containers happens to be running at the moment. It’s just easier to keep a native, non-dockerised, PiHole.
Enough ranting for today.
1 Like
Obviously. Take a look at what TrueCharts does with their apps and come back to this thread when you understand what I’m talking about.
I’ll just remind everyone to please bear in mind the forum rules regarding posting in a respectful and courteous manner.
2 Likes