Thanks for the reply.
There are probably more than 3 options here. I know of one commercial
development that is using x86 embedded CPU with DOS to make a hand-held
Automotive Test instrument . They are using C & C++.
Since Ada is not available that was not even a factor. This is an
upgrade/conversion from an earlier CPU that is no longer available.
See my comments below and others can chime in.
> From: Robert C. Leif, Ph.D.
> Sent: Wednesday, September 22, 1999 11:10
> To: Dale Jr, William; [log in to unmask]
> Subject: RE: DOS site
> To: William Dale Jr. et al.
> From: Bob Leif
> There are 3 reasonable possibilities for real-time programs which are
> on an operating system: DOS (perhaps DR-DOS),
There is a free DOS - see [ http://www.freedos.org ]
> Windows CE, and Linux with an
> Ada core. The only real disadvantage of DOS is the short(8.3) file name.
If you're not using a disk then that is not a problem. Even a Ram disk or
Flash can usually deal with short names. I'm talking embedded Single Board
Computer (SBC) -- may or may not need "Real-Time"
> Otherwise, it works very well. Windows CE could be very inexpensive
> alternative. It supports long file names, runs on Intel and other
> and can be used for WebTv. In fact, it is better suited for most home
> applications than the other Microsoft operating systems.
Wince (ugh!) is Not Real-Time and is a CLOSED OS. Ada for the Palm Pilot !
That is an idea!
> Linux with an Ada
> core could be sold or distributed by an Ada compiler vendor. They could be
> traumatized by making a real profit.
Linux and Ada are a good fit here as well. Linux has a Real-Time Kernel
option and word is ACT is working with other folk to integrate that with the
Ada Tasking Model. The FSU thread model may do all that most folk need.
> The other trick is make the front end XML or XHTML. XHTML is HTML that
> conforms to XML. This would be what an Ada software engineer would normal
Not too familiar with these but bindings and libraries for this are
available. Embedded systems may not need to deal with this at this level.
> -----Original Message-----
> From: Team Ada: Ada Advocacy Issues (83 & 95)
> [mailto:[log in to unmask]]On Behalf Of Dale Jr, William
> Sent: Wednesday, September 22, 1999 9:48 AM
> To: [log in to unmask]
> Subject: Re: DOS site
> I think the DOS port is important.
> There are Free or at least Cheap DOS clones and I 'd like to see a SBC
> running DOS w/ Gnat Ada 95 as a very interesting real-time system. DOS is
> the next best thing to NO Operating system - which is were I think Ada
> really shines! If the DOS / Gnat Ada 95 supported Ada tasking it could
> awesome! I don't think it does in previous versions.
> So sitting in a shrink-wrapped box is:
> * x86 SBC with lots of memory, ports, ethernet, USB what have you
> * Ada Gnat Ada 95 development tools that run on Windoze and/or Linux
> Sounds like a nice prototyping/target embedded development environment.
> Unfortunately I'm not real experienced at such ports. If there are
> others who find this an interesting idea let me know and we can kick it
> around and see if there is a need.
> Bill Dale
> mailto:[log in to unmask]
> mailto:[log in to unmask]
> > ----------
> > From: Victor Scamerda
> > Reply To: [log in to unmask]
> > Sent: Wednesday, September 22, 1999 01:16
> > To: [log in to unmask]
> > Subject: Re: DOS site
> > Hello
> > >>> Michael Feldman <[log in to unmask]> 09/22 2:02 AM >>>
> > >What I need to continue the site is some evidence that DOS-port
> > >development will continue. I need to hear from the developers
> > >who will do it.
> > >
> > >The ALT (Ada for Linux Team) has been extremely active, as anyone
> > >can see who looks at their website (www.gnuada.org). On the DOS
> > >side, 3.12 is nearly here and there is still (apparently) no usable
> > >3.11 for DOS.
> > I thought, the recent discussion on the DOS port did show that people on
> > this list who care about it were agree to skip 3.11p version alltogether
> > and jump to the 3.12. At least nobody argued about that. Am I wrong on
> > this subject?
> > >It would be nice if a DOS port emerged very shortly after the
> > >public distributions were released. One of the ALT folks mentioned
> > >(here or on chat) that ACT is cooperating with them - I assume this
> > >means they get the final sources a bit early. In theory, I guess
> > >this could also happen with DOS, but I rather doubt ACT will give
> > >any help to a group that cannot seem to get the previous version out!
> > >
> > >Asking for early 3.12 sources would, I conjecture, be answered
> > >with "where is the 3.11 port?"
> > Don't forget that Linux is supported OS and DOS is not. Apparently Linux
> > has much larger active developer base, than DOS and as the sutuation
> > tools infrastructure to build DOS and Linux applications also is not
> > expecting the same results from both of them is a bit unfair. I remember
> > when Jerry van Dijk asked if somebody on this list would host a CVS
> > for DOS development, nobody ever answered. Some of the former DOS
> > developers are quickly moving towards Linux or Windows. So, these are
> > clear signs that a DOS niche is narrowing and not asking too much would
> > a more realistic approach.
> > Despite all these facts I will continue to support DOS in the future as
> > long as possible, and when GNAT 3.12p arrive I hope to release the DOS
> > port within a week or two after that.
> > If anyone disagree with my views, please speak up now or it will look
> > nobody really cares about DOS port's future.
> > Victor Shkamerda
> > !
> > !
> > !
> > !