Re: Unified BSD?

[Martin <martin.kelly4000%gmail.com@localhost>]

I see your point. But again the problem arises of if you have one big fat sign "BSD" which BSD are you developing for?

Without having at least a universal package management system if you don't want each individual BSD developments to be porting drivers across. 

Which i believe is already in place in the form of "pkgsrc", but this would require a standardization of pkgsrc and how its implemented within each BSD. As i know NetBSD, FreeBSD and DragonflyBSD all apply pkgsrc differently.

The biggest fault i see with this idea is updates. Pkgsrc is generally just one large blob of binaries or source that have been deemed stable and compatible with that current release of whichever BSD in question (generally archived by date). That means in its current form with each BSD on a different release schedule there would be incompatibilities for those BSDs that don't update as frequently (namely NetBSD, OpenBSD & DragonflyBSD). Which means a change of how source and binaries are archived not to mention pooled together.

On Wed, Nov 14, 2012 at 7:45 PM, Magnus Eriksson <magetoo%fastmail.fm@localhost> wrote:

On Tue, Nov 13, 2012, at 14:18, Martin wrote:
> My point is about the possibility of creating a new BSD project (with
> separate developers) that aims for 100% compatibility with at least
> FreeBSD, NetBSD, OpenBSD and maybe DragonflyBSD.

On Tue, Nov 13, 2012, at 22:43, matthew sporleder wrote:
> If you are interested in generating linux-like "buzz" advocate
> hardware manufacturers and industry types to fund (with money)
> development of drivers.

Not a developer, but here's something I've been thinking about: Are
there perhaps some *parts* of the major BSDs (kernel interfaces, file
formats) that could benefit from being unified / standardized?  Maybe at
least a subset of syscalls and libraries that could be agreed on and
declared stable forever so that simple binaries can run?  That is
something that's already being done for Linux compatibility - except for
the bit about stability.  But why should I have to keep Linux binaries
around for handling weird archive formats?

I think matthew is basically right; but if there was only one single
target to develop for, with a big fat sign on it saying simply "BSD",
I'd bet that arguing for getting things written - graphics drivers or
userland tools for managing ones RAID setup, or whatever would end up
being feasible - would be easier.

In my daydreams (slightly less unrealistic than the mail that started
the discussion) I'm sending an email to a developer that says "Hey, you
can get four done in one shot, and it's also a standard.  And did I
mention that Apple and the Minix project have been using lots of code
from the BSD projects?  Want to bet they'll adopt this too?".

Yeah, I need to get out more, but you get the point.


Magnus