Hello! [Mon, 11 Apr 2005] Robert Lipe wrote: > > Downloading from garmin does not cause the high CPU load I was > > I never received a complaint about this, but I'm assuming you're giving > a nod to this change: > > ---------------------------- > revision 1.18 > date: 2004/11/01 17:34:28; author: robertl; state: Exp; lines: +1 -1 > On UNIX, there's no reason to spin on the select. Wait at least a character > time before looking again. > ----------------------------
Yeah. Sounds exactly like what I've been seeing.
> > <bounds minlat="-286276469720.843139648" minlon ="9999.000000000"
> > maxlat="46404771532.667770386" maxlon="634281368307.067993164" />
> > <wpt lat="46184670052.667800903" lon="634281368133.769653320">
> > <ele>9999999562023526247432192.000000</ele>
> > <cmt>c^P</cmt>
> > <desc>c^P</desc>
> > <sym>Waypoint</sym>
> > </wpt>
> Garmin "serial protocol" is actually a collection of about 17 marginally
> unrelated (sigh) protocols. What GPS (and what firwmare version) is
> this?
It's a Garmin eTrex (B/W) Firmware version 3.60.
This is what I get when I download it from a powerpc (Debian powerpc32,
gpsbabel 1.2.5). The data sounds correct. Contrary to my previous note
it obviously works nicely on ppc.
<bounds minlat="39.156847894" minlon ="-66.653934717" maxlat="52.795314789"
maxlon="10.804452812" />
<wpt lat="52.667796621" lon="10.753206443">
<ele>87.124512</ele>
<name>013</name>
<cmt>013</cmt>
<desc>013</desc>
<sym>Waypoint</sym>
</wpt>
> It looks like things are basically working. There's probably some
> word sizing problem somewhere in the serial code.
I did the same on an amd64 and a ppc32. And put all files into
garmin-amd64-D9-dumps.tar.gz.
> gpsbabel -D9 -i garmin -f /dev/ttyS0
attached as garmin-D9-ppc32.txt (with the stderr output interspersed)
and garmin-D9-w-amd64.txt
> gpsbabel -D9 -t -i garmin -f /dev/ttyS0
first 200 lines attached as garmin-D9-t-ppc32.txt and garmin-D9-t-amd64.txt
> gpsbabel -D9 -r -i garmin -f /dev/ttyS0
attached as garmin-D9-r-ppc32.txt and garmin-D9-r-amd64.txt
> Sadly, they're all three different protocols in Garmin-land, so it won't
> shock me if they're three different fixes... I just want to tackle them
> in increasing complexity.
Ok. I hope these dumps help you enter the next stage ;-)
Robert.
--
Ask not for whom the <CONTROL-G> tolls.
garmin-amd64-D9-dumps.tar.gz
Description: Binary data
signature.asc
Description: Digital signature
