From: Bob Leif
To: Chris Sparks et al.
Referring to the statement, "Granted it has parts written in "C"." Knowledge
of C should NOT be a requirement to program in Ada. I have been looking for
an inexpensive real-time environment for an Ada program. From my
perspective, I require that all real-time software be in Ada including
permitting my program to lock-out all normal interrupts. In the future, if I
were to write true multitasking software, I want to use the Ada tools
available in the Ada Annexes. As I have repeatedly stated, the GUI should be
based on XHMTL or XML. The file system, if it had a good Ada binding could
be Windows CE, Palm, Linux, etc. It also could be in Ada based on an
I might note that a very important requirement for a commercial operating
system, is ease of restoration after a crash. In case any of you are using
Windows 98, make a Bootlog file while your system is working. I just had a
near fatal crash of my system and found a corrupted file by comparing the
before and after Bootlogs. An operating system with well thought out
exceptions that facilitated recovery after the inevitable catastrophes,
would be extremely competitive.
From: Team Ada: Ada Advocacy Issues (83 & 95)
[mailto:[log in to unmask]]On Behalf Of Chris Sparks
Sent: Friday, June 09, 2000 6:56 AM
To: [log in to unmask]
Subject: Re: Ada RTOS.
Robin Reagan wrote:
> With all the recent talk of an Ada "killer app" I was thinking... (I know
> this would not be the popular "killer app" but maybe a micro "killer app"
> for the RT community :).
There is nothing wrong with doing killer apps, however, we need to set the
ground work first and have all of our tools in place (plug and play so to
speak). We need to first determine which standards we will be supporting.
After all the infrastructure is done then we can attack killer app projects.
If we do it too soon it might discourage us.
> Ada shines in the real-time/embedded areas (among others) this is where
> roots are! How about an RTOS (written in Ada of course) designed for
> embedded systems (The popular term is "appliance" I think). I'm sure this
> has come up before... but why hasn't it been done (or has it?).
Correct me if I am wrong, however, wouldn't RTEMS be an area of interest
for you. Granted it has parts written in "C", however, maybe we can upgrade
it so to speak.
> After having worked with vxWorks for the past several months, I can't help
Mea Culpa! :-) Just a little humor here. I have had nothing but bad luck
> but think that an open source (BSD license perhaps?) modern RTOS written
> Ada should do well if it can be shown to be very reliable, easy to
> configure, and flexible.
This is fine, however, which application market will you be targeting once