🚀 go-pugleaf

Ralph,
Thanks for you response.
I succeeded.  See below for details.
JB

"Ralph Wade Phillips" <ralphp@techie.com> wrote in message
news:9iolor$k3k2g$1@ID-81734.news.dfncis.de...
> Howdy!
>
> "John Becich" <jbecich@glendale.net> wrote in message
> news:%HN37.1912$23.229253@newsread2.prod.itd.earthlink.net...
> > I'm reading the COM1 port of my computer with turbo C++ code.  I'm using
a
> > simple loop in my code to repeatedly read the COM port and printf the
> result
> > to the computer screen.
>
>         No, you're reading a single port that's the data register.
There's
> a difference.

Agreed.
>
> >
> > The problem is that my program cannot tell the difference between fresh
> data
> > on that COM port, and stale data.  I get a repeated print out to the
> screen.
>
>         That's because you're not checking the status lines off of the
UART.
> If we check out some of National's application notes on, say, the
NS16550AN
> UART (which the I/O of your machine is based on), we find out that AN-660
> (at http://www.national.com/an/AN/AN-660.pdf ) which is sample code for a
> NS32GX32 processor (but the algorythms are the same for any processor)
that
> at  BASE+3, bit 0, is a flag that states if there's new data in.
>
>         Why complicate matters?  Just ask the UART if there's anything
> available.

I did that by asking 0x3fd, bit 0, if it was a "1."  When it came up "1"
then I knew fresh data was waiting, and I fetched it, once.  This solved the
problem.
Thus the appropriate UART (FIFO) flag is at base address + 5.

>
> >
> > Isn't there some kind of flag that goes up when new data arrives
incoming
> to
> > the port from a serial communications line; and down after my computer
has
> > read that data?  Might that be DSR, DTR, or similar?
>
>         None of these.  Again, checking out the specs on the chip that the
> Super I/O is based on (another appnote is AN-628 at
> http://www.national.com/an/AN/AN-628.pdf ) is HIGHLY recommended.
>
>         RwP
>
>
>