[okl4-developer] open source OKL4

[Jim Howard]

You call OKL4 open source? You call this "Open Kernel"?!?

This is so weak and ridiculous I hardly know where to begin....


--Jim Howard <ji...@finestplanet.com>


_______________________________________________
Developer mailing list
Deve...@okl4.org
https://lists.okl4.org/mailman/listinfo/developer

[Daniel Potts]

Hi Jim,

I can tell you're struggling for air. Perhaps this isn't the right forum for you. Are you after our open source release? Is the website download not working for you?

Daniel

[Jim Howard]

You mean 3.0, circa 2008? That's the last open source release I see.


--Jim Howard <ji...@finestplanet.com>

[Daniel Potts]

Hi Jim,

Wow. Time flies. You're right that the 3.x series kernel is the last that was released on an open license, and that was the last time we called OKL4 open source.

For now the 4.x kernels have not been released open source. If there was more interest from the community we would explore that. Unfortunately we received very little community participation in 1-3.x to justify that. It was very one sided. Plenty of takers and very few contributors. We'd be interested in any ideas on how do do it better!

Stay tuned ...

Daniel

[Jim Howard]

3.0 may be the last time you called OKL4 open source, but it's still
called Open Kernel Labs, and it doesn't look very open to me.

There are always far more takers than contributors for any open source
project. It's just a question of scale. Linux has the scale -- in more
ways than one. There's always room for more junk in the Linux yard, and
that attracts contributors like flies.

So about "very little community," hey, it's a very little kernel,
right? The squeaky wheel gets the grease, and the smaller the wheel the
less there is to squeak.

[Jim Howard]

And by the way, what became of that slogan: "Be open, be safe."

[Daniel Potts]

Jim,

This is a developer support mailing list. Thanks for your support and contributions.

Daniel

[Jim Howard]

I'm referring to the logo on the title page of the OKL4 Microkernel
Reference Manual that accompanies 3.0.

I just copied this from the current page one gets when one clicks on
"Why Choose OKLabs" at <www.ok-labs.com>:


Open for business with open source

Mobile developers rely on OKL4 to implement trusted and secure mobile
computing. General Dynamics Broadband believes that transparency
engenders trust and so embraces open source software and community-based
development.

OKL4 is a prime example of successful commercial open-source software.
Product source code is freely accessible to developers, decision makers
and enthusiasts for inspection, evaluation and documentation. General
Dynamics Broadband also underwrites and supports a vibrant open-source
community around the OKL4 platform.

--Jim Howard <ji...@finestplanet.com>



Daniel Potts wrote:

[Ph.T]

I think the problem is OKL4 is military-grade isolation,

and the NSA doesn't want us to be protected like that .

--
Americium Dream Documents
"(real opportunity starts with real documentation)

[Daniel Potts]

Hi Ph. T,

I love the theory!  OKL4 is used for many Information Assurance programs - those that make it harder to hack (irrespective of who you are).  They are also used in the same way to improve safety, reliability, etc across all sorts of connected and embedded devices.

I'm not sure what the point of this thread is, other than to troll OKL.  However, I am keen to hear some constructive ideas (as well as conspiracy theories - love them!), so I'd like to hear about what people are actually wanting to do with OKL4.  What do you need/want?  What do you plan to build?  Product or academic?

As mentioned previously, one of the major factors with lack of updates under a liberal license for OKL4, is lack of community involvement such as contributions of new SoC ports, bug fixes, demo systems, products, etc.  I don't want to discount that many members of our community have done some amazing work in the above categories, but I suppose this open source program has not gained enough momentum to be self sustaining.  We also had some people take the code and fork it off.  Again, while perhaps within their rights, that isn't particularly helpful in the spirit of community.

We have kept OKL4 3.x available, because our main past community involvement was people wanting to look into a modern kernel, and those were largely academic.

With OKL4 4.x and later, we would like to explore making it more widely available, and to make a case for that I am keen to hear more about what you'd like to use it for.

Thanks,

Daniel

[Tim Newsham]

"I am keen to hear more about what you'd like to use it for."

My perspective is probably a bit unusual, but I am a security consultant
and am seeing systems based on OKL4 deployed in real world projects.
Having an open platform helps me to better understand the security
architecture of the product and point out flaws (where they exist). To be
honest, I do end up spending a bit of time reverse engineering these
details because of closed source releases or closed source patches
added to open source releases. This is just wasted energy (but sometimes
very fun and rewarding in itself :). Keeping details of the system closed
does provide from short term benefit (IMHO) but in the medium and long
term, vulnerabilities will be found by well funded security researchers (not
always the people you want to find vulnerabilities) and will probably go
unnoticed and unfixed by the larger community much longer.

my two cents...

Tim Newsham | www.thenewsh.com/~newsham | @newshtwit | thenewsh.blogspot.com

[Ph.T]

I see that the open source security community

is rallying around Xen, (see Qubes OS)

as it supports unmodified kernels .

. however,

as has been pointed out by okL4 staff,

the Xen TCB is much larger than okL4,

making verification impractical,

in contrast to okL4 Verified .

http://www.ok-labs.com/releases/release/open-kernel-labs-provides-okl4-verified-for-download-and-prototyping

[Norman Feske]

Hi Daniel,

> With OKL4 4.x and later, we would like to explore making it more widely
> available, and to make a case for that I am keen to hear more about what
> you'd like to use it for.

it is good to read that you are re-considering the Open-Source route.

If OKL4 4.x became available under an Open-Source license, we would
certainly consider it as an optional base platform for the Genode OS
Framework. Btw, we still support and sporadically use OKL4 v2.1 (the
last version with support for x86).

That said, personally I'd be even more thrilled to see seL4 released as
Open-Source software because the design of this kernel seems to be a
particularly good match for the dynamic workloads that we address with
Genode.

Best regards
Norman

--
Dr.-Ing. Norman Feske
Genode Labs

http://www.genode-labs.com · http://genode.org

Genode Labs GmbH · Amtsgericht Dresden · HRB 28424 · Sitz Dresden
Geschäftsführer: Dr.-Ing. Norman Feske, Christian Helmuth

[Jim Howard]

Daniel Potts wrote:
> I'm not sure what the point of this thread is, other than to troll OKL.

There does seem to be an element of that, doesn't there? But I did come
to your site in all innocence, to see if there was something to develop for.

The point of the thread has been to point out the rather large
discrepancy between what your site says and what your site does. This
discrepancy is the reason I really have no idea what I'm doing here,
other than to point it out.

You are quite rightly free to approach your business as you see fit.
Unlike some people I respect that, and I know there are reasons that
influence the decisions. But notice that you did say "troll OKL." The
"OKL" part means "Open Kernel Labs," and I can only repeat, the kernel
is not open, for whatever reasons you may have, and for myself, that
makes the issue of whether I have something to develop for moot.

I might recommend reading the statement I copied from your site in my
preceding post. It states rather well some other reasons for open
source besides "community" and "contributions."

As for forks: Forks are normally considered very much in the spirit of
open source community. Many projects would rather be forked than have
people pushing them to make changes they don't want to make. And if
forks are open source (as they must be under GPL, which OKL4 was not),
you get contributions from them whether the forkers intend them or not.

And didn't OKL4 start out as a fork of L4?

Thank you, I believe I've beaten this dead horse thoroughly into the
ground by now.



--Jim Howard <ji...@finestplanet.com>