Tales of soaring software + device crashes, wrong or misleadinginformation presented, and user errors?

It would be useful to hear your tales of soaring software crashing and/or providing misleading or inaccurate information. Please be specific.

It would also be good to hear about cases of "user error" where you made a mistake in either setting up the device, misinterpreting the information displayed, or you found yourself confused or mislead by the way the information was presented. Of particular interest would be pitfalls that you stumbled upon in flight and how you resolved the difficulty.

I'd also like to hear about any cases of two devices running in parallel and presenting conflicting or diverging information (for example different indications of wind direction/speed or final glide arrival height).

On Thursday, May 30, 2013 9:03:40 AM UTC-7, son_of_flubber wrote:
It would be useful to hear your tales of soaring software crashing and/or providing misleading or inaccurate information. Please be specific. It would also be good to hear about cases of "user error" where you made a mistake in either setting up the device, misinterpreting the information displayed, or you found yourself confused or mislead by the way the information was presented. Of particular interest would be pitfalls that you stumbled upon in flight and how you resolved the difficulty. I'd also like to hear about any cases of two devices running in parallel and presenting conflicting or diverging information (for example different indications of wind direction/speed or final glide arrival height).

My gps source is a cambridge 302A. I run XCSoar 5.2.4 on an old iPAQ 3850. It is stone-axe reliable, and has the brightest in-sunlight display I've ever seen. I also have a Dell Streak, but the display is really nowhere near as good as the iPAQ. However.... the iPAQ won't run later versions of XCSoar gracefully.

Software Crashes and bugs:

I use XCSoar and I am a strong proponent of it, but I have had my share of crashes or bugs. Because of that, I'm glad to have backup loggers for OLC or Contest purposes. None of the crashes have required anything more than restarting XCSoar to recover from them. The biggest tragedy is the loss of a contiguous IGC file.

- Internal GPS on the Dell Streak isn't perfect. I've had XCSoar lose position info because of this. Not XCSoar's fault obviously. Connecting to a more reliable GPS like Flarm as the primary GPS source works great. When the Internal GPS does get wonky, you may just notice that your moving map isn't moving or updating.

Cambridge L-Nav/GPS-Nav: Remarkably stable. Only a few issues that have ever required a restart in the air. Wind data in straight flight isn't as trustworthy as I would like and switching between screens will show different arrival heights. That's me not remembering what those screens show for calcs. Not a software problem.


*** Software Limitations that produce less than reliable information: (Really Human failings to account for those limitations)

Bogus Wind:
GPS based software like See You Mobile , LK8000 or XCSoar all only understand the wind if you are circling consistently, unless you've got air data hooked up to it from an internal flight computer like a 302. Super common rookie move is to climb in ridge, convergence or wave where it can't get a sense of actual winds. Then trust the glide computer as you head off into a headwind "above glide."

Murphy's Law:
Glide computers work on theory. In theory your glider should get 40:1. In theory lift and sink will balance out. In theory the winds will be consistent. Trusting all of those theories to correlate and work out will eventually put you in a position where you will either land out, or at least need to change plans, stop and thermal, do something.

No Final Glide Mode:
Some flight computers don't take winds into consideration in final glide when you are looking to cover max distance over the ground. This means that flying into a headwind, the speed director may not tell you to fly faster than best glide.


Mountains in the way:

Older pure GPS, no map/terrain systems don't know about that mountain in front of you. If things behind the mountain are disappearing, you're not getting over it and better have an option.

*** The Good:

Wind Calculations:
XCSoar with it's GPS based winds is usually within a few degrees and a knot or two of my L-Nav that has airspeed data as well. Close enough for government work if I'm circling and winds are generally consistent.

Wind Input: XCSoar will let me manually update the wind so that I can tell it that I know I'm circling in3 knots , but I'm about to leave the convergence and fly into a 10 knot headwind for 30 miles.

Safety Margins:
Arrival Heights of 1000agl can help with Murphy's Law.

Either using a high MC value like 4.0 and then flying MC0 or using the Polar Degradation feature in XCSoar, you can further buffer against counting on published glide performance.


User Errors I've seen or done:

Wrong Polar: Jumping in the club 1-26 with your PDA still configured for an ASW-20. Makes for bold moves.

Wrong Arrival Height: Thinking your arrival height is 1000agl. Not realizing it is for 0AGL or 500AGL.


The main thing is to not blindly trust or depend on a flight computer. Have a notion of what seems correct and use that in your head to validate what the flight computer says.

On Thursday, May 30, 2013 9:14:58 PM UTC+2, wrote:
the iPAQ won't run later versions of XCSoar gracefully.

I havn't seen your ticket on the XCSoar bug tracker about this problem.

On Thursday, May 30, 2013 3:11:41 PM UTC-7, Max Kellermann wrote:
On Thursday, May 30, 2013 9:14:58 PM UTC+2, wrote: the iPAQ won't run later versions of XCSoar gracefully. I havn't seen your ticket on the XCSoar bug tracker about this problem.

Pobably because I never opened a ticket. I brought it up on the XCSoar forum, but the consensus was that the old 3850 just didn't have the processing power. The later versions of XCSoar run great on the Dell Streak, but mine has a problem with the pins on the connector, and the cable doesn't fully connect. No connection, no charge, short flights.

On Friday, May 31, 2013 12:57:36 AM UTC+2, Uncle Fuzzy wrote:
I brought it up on the XCSoar forum, but the consensus was that the old 3850 just didn't have the processing power.

Which is bull****.

At 06:26 31 May 2013, Max Kellermann wrote:
On Friday, May 31, 2013 12:57:36 AM UTC+2, Uncle Fuzzy wrote:
I brought it up on the XCSoar forum, but the consensus was that the old
3850 just didn't have the processing power.

Which is bull****.


Agreed, I've never had an problem running any previous versions of xcsoar,
and the latest version 6.6.1 is just fine too.

I've never had it crash either, ( usage ~ 100 hrs a season since it went
open source ), but I always make sure I use the 'last stable' version not
the betas.

The menus, zooming, in fact all operations work fine and quickly for me,
although I don't have the topology or overlay maps turned on, I do have the
flarm radar turned on.

I think xcsoar is by far the best soaring software money can buy.

A big thanks to Max and the team for keeping XCsoar going for those of us
with older hardware, I often don't think they get the recognition they
deserve.

On Friday, May 31, 2013 2:26:38 AM UTC-4, Max Kellermann wrote:
On Friday, May 31, 2013 12:57:36 AM UTC+2, Uncle Fuzzy wrote:

I brought it up on the XCSoar forum, but the consensus was that the old 3850 just didn't have the processing power.



Which is bull****.

A guess: Uncle is using ppc 2002. 5.2.4 worked fine on 2002. 6.x, not so much. My experience with Ipaq 3850/3950, anyway. I had better luck with ppc 2003.

T8

T8

My gps source is a cambridge 302A. I run XCSoar 5.2.4 on an old iPAQ 3850. It is stone-axe reliable, and has the brightest in-sunlight display I've ever seen. I also have a Dell Streak, but the display is really nowhere near as good as the iPAQ. However.... the iPAQ won't run later versions of XCSoar gracefully.

I've had exactly the same experience as you except I think I can see the Dell Streak screen much more clearly in polarized sunglasses than the iPAQ 3850. In the middle of the contest I upgraded to XCSoar 6.6 Beta2, and only during flight I noticed much slower responses. For example, just before arriving at a turnpoint the XCSoar would freeze for 30 seconds or more, and unintentionally I would make the turn later than planned. Then on a final glide the info boxes would update once every 30 seconds or so, which made me quite uncomfortable too. The same day I reverted back to an older version (I believe it was 6.5.2) which works fine. I did not post a ticket because I did not see anyone else complaining about it.

Wiktor

I upgraded to XCSoar 6.6 Beta2
I reverted back to an older version (I believe it was 6.5.2) which works fine.

Speaks for itself... Stable versions are called "stable" for a reason, while a beta version is introducing new stuff that might still fail every now and then.