[Infotropique] fyi

Delete this message

Reply to this message
Author: ng0
Date:  
To: infotropique
Subject: [Infotropique] fyi
Hi,

I'm trying to explain the long silence.

I've been very busy with GNUnet. Furthermore I have to increase my
working hours, and quiet recently (2? months ago) I've started
exploring NetBSD further to the point of contributing. In addition
to that I still have university going on and I'm looking into a
research topic for gnunet. All in all, I don't really have the
time to give updates.

The gist of this project is, without further explanations and
hoping that I don't repeat myself:
The 2nd prototype based on heavy modification of Guix and GuixSD
was disregarded many months ago because it is just not worth the
outcome.
For now I have ideal goals planned out in smaller steps.
I am going to focus on a plugin-based all-purpose package manager
(with some more features and theories) build around GNUnet.
This draws some inspiration from Nix and Guix, but it won't
base on either one of them.

This will require work on GNUnet to fulfill everything my goals
set out to be (in particular, it requires my work on onion routing).

While the PM can exist independent from the fulfillment of all
goals, it will very likely take years because of my other
work.

Lastly: there won't be an OS. If I ever dig down that hole again (I'm
still interested in and reading into OS developments and researc)
it won't be Unix, it won't be for immediate use, and it will most
likely be for research or simply myself.
An OS requires a dedication of time I don't have. If this will
happen against everything I stated here and on paper, I will
have proper financial backing. But I very much doubt this will
happen, so for now a more advances package manager without a
central infrastructure is more practical. Even if the world would
not be interested in it, it still has its Proof of Concept
application and it is something I personally want to use.

So for now it is quiet around infotropique and plant. It will
remain quiet as I'm thinking about the best ways to implement
my ideas. As long as infotropique.org and plant.pm don't
disappear without a reason, the project is still going on.
Just at a slower pace.

As I see some parts of the GNU community kicking around this idea
every now and then: You must be very clueless if you assume that
you can just take the kernel of an BSD Operating System and it
will just work. There's (1) social implications: a good part of
the systems I looked into are okay'ish with GNU. When you go
ahead and assume ahead that this will work without sending
any message because it could be fun (I have been there, despite
knowing this implication and I stopped as I reflected on it),
you are not thinking this through, as a BSD kernel is not Linux
and it is not Hurd, end so forth. (2) you don't get the kernel
just like this. as this is a whole OS and its userland, you will
have to do massive untangling and legal analysis. That's the
point where you will realize why Gentoo stopped its BSD part,
even though it built from upstream sources.
The idea of building a modular OS is interesting, but a BSD
Operating System is more than its kernel or its software. It
has people involved in it which have gathered years, decades
of experience about the specific parts they contribute to.
The most ignorant move one can make is to fork parts of it.
If you think it through with all consequences beyond the
technical gain, you would be no better than corporations
taking the source just because they can. And corporations
(occasionally) contribute back at least. GNU and BSD have a
history, simply ripping out pieces and forking from the GNU
side could have grave implications. Wisdom lost, etc.