As you stated…
And IMVHO developers should care a lot more about the community value, work, and test, because every yikes not found from the community will hit Nethesis’ QC and customers.
It’s an opinion, so… a bias par se? 
You as person, developer and enthusiast have your own bias, due to technical costrain, budget, management decision and the team needs to follow.
Are my posts unfair about the situation? I hope not, and I hope that the community will emphasize that if they are, but also… do my opinions have to comply Nethesis managament and developer team decisions? I don’t think so, I’m “only” me, providing opinion… and founding elements that leads me to that.
The only “technical” reason is “no longer shorewall”. This lead to a replacement choice.
That lead to three options:
- pick up shorewall project, revamp it if necessary and follow the development
- write something that fit the same shoes
- pick something existing that could fit same shoes
First two options are not viable probably for budget reasons, from this seat; feel free to deny.
Third option was the one choosen and fun fact… there are not a lot of modern enough firewall linux distros fitting the boxes already sold or currently buyable.
Now, compared to the level of tools provided from shorewall/NS7, OpenWRT is a step back, also in system footprint and disk usage. But is a step back in functionality .
Also, the necessary removal of the firewall option from the server install is a step back in functionality too.
It’s also technically necessary for the container approach? Yes. The factual technical reason is here. But everything reported before stands.
Again… I try to explain what I find wrong. I can simply say “that sucks”, or do a everytime refrain about “the best way to deploy is”… And both IMVHO are… wrong.
Project provided by Nethesis has until there proven solid, reliable, with some hiccups and whoops
but the job done from developers has been until NS7 great. Will be with Container Overlord 1.x? I cannot tell, but I trust your commitment and capabilities.
However… lots of persons delivered criticism and “that won’t work in the same old way” claims.
As staten before from other persons, leaving the adoption is still an option, but deying, deflecting and hype-pumping… is really a good way for keeping people on that? IDK.
I personally do separated devices for network and server, possible on different hardware. Many reasons are behind this preference, but… it’s me, not everyone else. So that should not bother me that much.
But what I found proper, common sense, viable, the right way to do things may or may not be interesting to the community, so it’s useless to countless times repeat that on most posts. I tried several times to provide a “do a better analysis” approach in some support threads, but what does fit me is only an example, not the best way to do things for all people.
This is the meaning behind
I am aware cannot steer the project, the ship already left the pier.
I also cannot change the current path of IT world. I’m not Microsoft, I’m aware of that (and even Microsoft IMVHO can’t do that).
I do things my way, so this means that other ways are… possible and viable to other people.
It’s simply stating the truth… what I find pleasant or not it’s irrelevant.
Probably the percentage of leaving adopters will be vastly compensated from other bigger adopters, at least as revenue standpoint. I hope that this will give room to Nethesis for increase resources on bugcheck and QC. Because for the route of Container Overlord (formerly known as NS8) lots of persons not willing to create a “one server cluster” are gonna leave. Because one server cluster is a overheaded cluster, without the advantages of the cluster…