Also people who left Twitter or never were there. Fostering radicalization while censoring legit links to other much more peaceful communities is going to increase social conflicts, and this usually leads to real violence sooner or later.

Although plausible, and part of it, it’s not just that. The second part is much more important (people unwilling to adapt), but one should think about why.

• How sudden & drastic was the change?
• How well trained were people?
• How much support did they get when something went wrong?

That’s the kind of stuff where companies & public administrations usually suck, “change management”.

I’ve seen Germany states jump into open source software strategies to later “regret” them and revert back to their previous state, many times… I hope this time they do it in a more thoughtful way, to avoid giving excuses to closed-source software lobbyists.

Very interesting :) , although I have the feeling we already have some features in place (and others on the way) trying to solve this same problem.

• We already have “Shadow Realms”, which doesn’t really solve the capabilities problem, but at least provides a thin isolation layer (at least our globals are protected!). This is supposed to work in all JS environments, not just NodeJS.
• On top of that, there’s work in progress to implement a permissions system (with experimental code already in place, that can be tested): https://github.com/nodejs/security-wg/issues/791

I don’t think talking in absolutes is the best way to tackling at problems. In this case, I would say there’s a gradient, and we can surely shift our current position from “utterly broken” to something more benign, even if there are still flaws pending to be corrected.

Starting from scratch would imply throwing away millions of hours of developer time. It’s tempting, but not feasible, and even if it was, it would be a bad economical decision.

thx! :)

• Professional services usually refers to have consultants for contracting. The more the demand for your services, the more people you have to hire to content your clients. Because salaries are usually one of the biggest costs in any company, this kind of business can be a bit more risky when there are economic downfalls and you cannot fire workers quick enough (I’m talking from the perspective of the business owner, leaving ethical aspects aside).
• Subscriptions don’t imply having people personally interacting with the client, because what you offer is not expertise, but a product that can be served to big amounts of clients, independently of how big is your team (of course you end up having to hire more people, but this team growth is much slower).

Their product offers many more features (and more complex ones) than the other services you mentioned. And they also took the time to provide a commercial product, with paid support, integrations…

This is key for companies, as it makes it easier to be productive, plus the SLAs give some peace of mind too.

very true

Oh, that’s a pity :( . I remember that my first open source program was precisely a wrapper around Youtube-dl circa 2007. I lost the code ^^U , but I had fun with it.

As a side note: That’s why we should keep mirrors of our open source projects. Consider Gitlab & Bitbucket as secondary mirrors… not sure about other alternatives. Ideas?

It uses the docker’s cli interface, so as far as I know, it should be compatible… with a two caveats:

• Podman is not explicitly supported, and aliasing the podman comand to “docker” won’t make it, because Avatar-CLi won’t know that this alias exist.
• In case you use a hard/soft link in the file system to overcome the first mentioned issue, must be noted that Avatar-CLI is not compatible with the rootless mode (yet), because it needs root access (just once in the whole process) to generate some special files that will be mounted inside containers (basically /etc/passwd).

Having said that, I definitely want to improve it and officially extend its support to Podman. I just need some time…