Re: Ada Stuff and some confusion

["Marin David Condic" <marin.condic.auntie.spam@pacemicro.com>]

Well, there's still RTEMS which might serve as a kernel, but it isn't
(AFAIK) a full-up RTOS. (More of an RTK to put under your compiler.) I've
not had any experience using it, so I can't really say much.

The trick is that you'd want something a bit more than an RTK - something
that could load programs (in ELF format, maybe?) and execute them as
independent processes. That, and support for lots of the I/O, devices,
etc. - almost a "real" OS, but not quite.

What I'd want is something that provided "basic" OS services, alongside of
which I'd pile the "specialized" libraries of routines to handle the
hardware I've got. That bundle (presuming it had the right capabilities and
was open sourced) could fit into a number of places and would offer the end
user a few things that would make choosing Ada very attractive:

1) Since it isn't a PC environment, you don't have to compete against the
truckloads of development tools & libraries already out there. (There is
competition, but much easier to match.) Ada wouldn't suffer from the
problems it has in other areas where - even if it is a superior language -
the leverage of existing stuff makes Ada a poor choice.

2) Since the OS is in Ada & the libraries are in Ada, it would be natural to
want to build the apps in Ada - although not required. (Extra layers on top
would make it possible to use almost anything - like Java & HTML.)

3) Since its in Ada, there would be (presumably) reliability benefits,
speedier development & less debugging - faster time to market. (Remember,
you're not competing against MSVC++, etc. Everyone else has essentially the
same facilities - very little!) Get out all the standard arguments as to why
Ada is better - throw in some benefits of OOA/OOD/OOP (because your
competition is plain vanilla C!) and I think you end up making a good
business case for it.

4) The adopters of the OS have the edge over the competitors when it comes
to putting apps on top of it. The competition is a bunch of C guys who won't
understand Ada, won't like Ada and won't know how to develop rapidly in Ada.
If you adopt the OS and Ada, you get there with your apps quicker.

5) The end user market is pretty huge (potentially) so it would be
attractive to want to develop for it. Ada could gain quite a few converts in
the process...

Just some random thoughts. I may sideline some of my other projects to look
at this more in depth, but I think it would require a solid starting point
(the core OS - or at least the RTK) or it would just take too long to get
there. Hmmmmm......

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/


"Ted Dennison" <dennison@telepath.com> wrote in message
news:uflz6.3475$jz.294146@www.newsranger.com...
> That's about where I am too. I'd rather see a real-time Ada OS, but I
don't
> really have time to start such a project myself (at least not for the next
6
> months or so...). I do try to keep an ear to the ground though. There's
> currently an interesting real-time microkernel project ongoing at CMU
> (http://www.cs.cmu.edu/afs/cs.cmu.edu/project/art-6/www/rtmach.html). Just
like
> GNU/HURD is based on MACH, it ought to be possible to build an (Ada coded)
RTOS
> on top of RT MACH. The sources are available there for downloading, but I
don't
> know how acceptable the licensing terms would be.
>