Re: [Gpsbabel-code] Course data field value changed during kml output

[Bortel L. <bo...@fr...>]

Hi Robert,

thank you very much for your reply and for addressing my concerns.
I used multiple format output just to demonstrate the issue: actually I would like to produce kml output with unchanged/uncorrected data. I got all data from a Queclink GL200 personal tracker and I would like to keep it as it is, either good or wrong. It seems for me that it is not possible to achieve this with gpsbabel 1.4.4.

Cleverness is good, even better if it can be controlled: would it be possible to make track_recompute() optional (default active)?
Or even better: compute course/speed/distance only if it is not present in the original data? (I do not know if it is possible to detect if course/speed/distance info is present or not.)
What is your opinion?

Thanks again.
Best Regards,
Laszlo

  Robert Lipe <rob...@gp...>, ezt írta: 
  
  As the author of that cleverness, I agree that it's an oddity.  The
original justification for track_recompute was to have more data in the
balloons such as instantaneous speed and course but, more importantly, to
allow the orientation of the icons to more clearly represent direction of
motion.  Other formats don't do that, but they also don't need those
fields.   It's very rare that we modify the source data in a format writer,
but this is one of the times we do.   We do have other formats that need to
do this.

We could have a KML-internal representation of speed and course but that's
kind of awkward, too.   That would be possible to do, but it's mostly busy
work. Yes, it would clean up an architectural blemish that writing KML
changes outputs that appear after that on the same invocation.  Very few of
our users write multiple formats in one shot.

Requiring every user of KML (and it's one of our most common formats) to
use a filter to get sensible output just makes things complicated for the
users.  While blemishes annoy me, I'd need a better reason than this to
torment our users. If changing the order of the writes troubles you so
deeply, use the stack filter to push before the KML write and pop afterward