[Microkernel-devroom] Devrooms for 2012

[Jakub Jermar]

On 15.11.2011 22:30, Norman Feske wrote:
> I like your suggestions and I agree with Martin that we should always
> try to ignite the interest of visitors that come by to look what is
> going on. If I image stumbling into a room of people working on crazy
> stuff that I'm not familiar with, I'd like to see two things: A way to
> quickly find out whether this room is of interest to me and worth to
> stay, and a hook to get somehow involved with the scene.
> 
> I don't know exactly how to achieve both but one thing that I would find
> attractive is some toy to play with, e.g., a running microkernel-based
> OS, an interactive demo - something that is fun to explore. Maybe we
> could dedicate one corner of the room to get people hooked?

At this point, it would be essential to hear from people on this list
who have already attended one of the previous FOSDEMS and have a devroom
experience, otherwise, I am afraid, we may start with wrong assumptions.

If there is any public present - no problem. On the other hand, I was
under the impression (and the original plan was) that this is going to
be primarily the developers' meeting. I hoped to be able to do without
the mandatory 'What is a microkernel' introductory talk and rather focus
on the more technical stuff (which can also be of interest to any
public, if present).

> For example, I could imagine to give an informal talk about our
> experiences with using C++ as implementation language, in particular
> about the troubles we experienced and the benefits gained from this
> approach - things like replacing an IDL compiler by C++ templates.

Sounds interesting.

> I would love to hear HelenOS' experiences with implementing a custom
> TCP/IP stack. What was the reason for this decision and has it paid off
> in hindsight?

We could talk about this too, perhaps wrap it together with some similar
topics.

> Also, I see that there is activity going on to get HelenOS self-hosting
> - a goal that Genode also pursues. However, I think the approaches taken
> are pretty different and I would appreciate getting insights into the
> respective developments.

Sure, why not.

> An ever-lasting topic is user-level device drivers. With Genode, we went
> a pretty long-winded route including the use of the DDE-approach, custom
> drivers, and device-driver OSes. Maybe these experiences are worth sharing?

Definitely. We can communicate a rather different experience with
developing our own device driver framework.

Jakub