Thanx.  Applied with minor tweak to the include path.

Rounding is just a mess.  I don't have a great solution to that. Chasing sub centimeter location differences across OSes is pretty silly.  Maybe we need a better way to test this stuff.  Ideas welcome.

Didn't we force testo to run in US.UTF-8 pretty much for you in r4525?  Did this not help?

We (I)  bozoed the stdin/stdout initially, but Steve made a pretty good run at fixing this in r4531.  More info, please.  Are you sure you're not just running an old tree?

I'd set fire to Bushnell before I look at it in much detail.  The ONIX devices we support were all discontinued years ago.  The inventory was liquidated through Woot a few years ago which resulted in a short burst of interest in those devices until people learned WHY they left the business.   It was a mistake for me to buy one of these and add support for them.  KML and GPX matter much more.
On Fri Sep 27 2013 at 5:28:18 PM, Gerhard Olsson <gerhard.nospam@gmail.com> wrote:
Add new formats and filters to .pro, MSVS compile patch

Status:
Cygwin testo: gtm, expertgps fails due to rounding off by one (gtm failed before too)

MSVS fails for more formats. Many failed before, I feel it is more than before.
bushnell, gopal, dusky, vpl, track-discard, tomtom_asc, mtk, kml
Most are due to locale is not forced (LC_ALL handling), some are rounding

Standard input test in gpx fails after the Qt rewrite, test hangs. I have not found why.
Similar do bushnell not work at all

(no big problem if this is not fixed, but README should be clarified)