[Xastir] HSP/AUX/Timings etc.

[Curt Mills, WE7U]

On Tue, 15 Jul 2003, Stephen West-Fisher wrote:

> First, more of a comment about how I fixed it, and what is next. I did a
> very bad programmer thing (I've been known to get harsh with programmers
> who work for me and do this) and just copied another section of code,
> made a few little modifications, and it worked.

Guess how most of Xastir was extended to add new features?  ;-)


> Looking at the code, it looks like the HSP stuff is an extention of the
> normal serial GPS stuff. However, it also appears a better use of code
> would be to treat the HSP code as an extention of the AUX code. In both
> HSP and AUX all data goes into the same buffer and is parsed from there.
> I'm going to give this a try when I get a chance.

That sounds reasonable.  I think HSP was added after serial, and
then AUX came later.


> Secondly, I'm still having the odd Segfault, but it may be my crappy
> code modification above. Origionaly I thought it was happening on the
> GPS strings, but this morning I had one on a digi of my beacon.

Were you doing map interrupts at the time?  Tigermap stuff?  What
was different at the time you had the segfaults?  Do they happen
even when you're doing nothing, and not tracking anyone (as in maps
are static)?

Gerry seems to think that interrupting tigermap draws might cause
them to happen.  I haven't seen any segfaults here no matter what I
try.  If I could reproduce them at will, I could solve them.


> Third, the timing issue I had brought up earlier. I'm now not so sure it
> is a problem, but Curt's idea of a timeout when reading the GPS may be
> needed in the HSP code. Last night I was getting very frustrated because
> the DTR was getting stuck low no matter what I did. When I decided to
> give up for the night, I noticed the GPS had lost sync. So my theory for
> today is that when Xastir gets stuck with DTR low, the GPS has lost
> signal.

That is correct.  I looked through the code a few days back and
that's what it does.  It was intended to be a "feature" that you
switch back to TNC mode as quickly as possible, meaning as soon as
you parse good GPS strings the switch occurs.  As it stands, if the
GPS becomes disconnected or loses lock, you stay in GPS mode.  That
wasn't the intention.  I'll look into changing that.

I'll put in a bug report on that very thing.

-- 
Curt Mills, WE7U                    hacker_NO_SPAM_ at tc.fluke.com
Senior Methods Engineer/SysAdmin
"Lotto:    A tax on people who are bad at math!"
"Windows:  Microsoft's tax on computer illiterates!" -- WE7U
"The world DOES revolve around me:  I picked the coordinate system!"