XCSoar / LK8000

On Monday, March 11, 2013 12:22:15 PM UTC-4, wrote:
On Monday, March 11, 2013 10:13:55 AM UTC-5, waremark wrote:

I would recommend making a fairly quick choice between the software options and putting the time into studying the chosen version. They will all do your job well. If you spend too long choosing you will not only waste time but also end up confused.



So, Max, Paolo or US pilots:



Is either of these two packages useable in a US contest? Do they depict US start, turnpoint and finish definitions, tell you if you're under the top for 2 minutes, help plan time completion for area and mat tasks, compute final glides to a finish cylinder with a minimum altitude, compute final glides around additional turnpoints (mat)? Can you input a task in less than 15 minutes? Is there a quick task A task B option (almost all in air task changes are now A to B)?



I tried xcsoar many years ago. Charlie Spratt changed the task in the air.. 15 minutes later on my second reboot, I swore it off. Is contest task entry any easier than back then?



If not, is anyone working on a "US contest" package for either program?



Please no flame wars on how stupid our rules are for using cylinders, 2 minute rules, in air task changes, etc. The rules are what they are. The question is, can this software help a pilot to deal with the rules as they are?



John Cochrane

I spent most of three seasons using XCS for US contests. On Ipaq 3950s to start, later on Dell Streak 5. Nine or ten contests IIRC. Something important was always busted in software. Start rules, AAT task optimizer, little things :-/. I wrote lots of trouble tickets. Some got acted on right away, some never. I haven't used XCS since version 6.3.

There's a timer that can be used to tell you how long you have been under max height.

When things work in XCS, they work well. Task editing is easy on the ground. Task editing in flight depends on your interface, and all of the hardware that I am familiar with that runs XCSoar is touch screen driven. I grew to hate touch screens in flight. My big problem with a touch screen is that I have to look the display for every single poke. This sounds trivial. In fact it adds up to a lot of distraction.

There are better solutions for racing. The two that I am aware of that demonstrably work (i.e. guys win with them, a lot) are SN10 and ClearNav.

The reason that these devices work better is that they take about a quarter the head down time that XCSoar does. Better user interface, easier to read displays in about equal parts. Neither attempts to do nearly as much as XCSoar can, but an awful lot of what XCSoar can do isn't helpful for racing..

The best clues to efficient soaring are outside the window. Less distraction equals better performance in my cockpit, and probably yours too.

By way of disclosure, I work part time for CNi, however this post is my personal opinion, informed by experience.


Evan Ludeman / T8

On Monday, March 11, 2013 7:11:25 PM UTC+1, Evan Ludeman wrote:
I spent most of three seasons using XCS for US contests. On Ipaq 3950s to start, later on Dell Streak 5. Nine or ten contests IIRC. Something important was always busted in software. Start rules, AAT task optimizer, little things :-/. I wrote lots of trouble tickets. Some got acted on right away, some never.

Some never? The following points to a exhaustive list of all of your tickets we havn't worked on yet:

http://bugs.xcsoar.org/query?status=...%40g mail.com

This is exactly one ticket, one that can't be any less serious.

Yes, you found a lot of problems over the years (totalling 24), and that was your contribution to a better XCSoar, it helped a lot. Thousands of eyes find thousands of problems in such a complex software (nearly 26,000 users on Android alone). We do our best to deliver high quality to all those pilots, and we need input from people like you.

But please be honest. I know you want to drop some advertisement for your employer, but don't exaggerate the severity and number of open bugs.

Good points Evan,

From my experience last year, V6.4 was very stable, all rules and
optimizations worked as necessary. As far as the heads down time goes,
I think a lot of that has to do with the pilots setup, I for one barely
touch my Streak in flight, the information I require is on my screen
when needed, automatic thermal mode, automatic final glide mode etc. All
information provided at a glance. Most of the interaction comes in the
form of a swipe across the screen to switch to a different page but only
when necessary, which is rare in my case. Your point about most of the
top guys in the US flying with CN or SN10 is well made, on the other
hand the winner of 2013 WGC Club Class was flying with XCSoar.

http://www.xcsoar.org/discover/2013/...Argentina.html

I'm by no means saying that XCSoar is perfect, but it is constantly
improving, conversely it is part of the problem, new features mean more
bugs and more complexity. While XCSoar is extremely configurable,
configuring it just right can be a daunting process.

Cheers,
Luke Szczepaniak

On 03/11/2013 2:11 PM, Evan Ludeman wrote:
On Monday, March 11, 2013 12:22:15 PM UTC-4, wrote:
On Monday, March 11, 2013 10:13:55 AM UTC-5, waremark wrote:

I would recommend making a fairly quick choice between the software options and putting the time into studying the chosen version. They will all do your job well. If you spend too long choosing you will not only waste time but also end up confused.



So, Max, Paolo or US pilots:



Is either of these two packages useable in a US contest? Do they depict US start, turnpoint and finish definitions, tell you if you're under the top for 2 minutes, help plan time completion for area and mat tasks, compute final glides to a finish cylinder with a minimum altitude, compute final glides around additional turnpoints (mat)? Can you input a task in less than 15 minutes? Is there a quick task A task B option (almost all in air task changes are now A to B)?



I tried xcsoar many years ago. Charlie Spratt changed the task in the air. 15 minutes later on my second reboot, I swore it off. Is contest task entry any easier than back then?



If not, is anyone working on a "US contest" package for either program?



Please no flame wars on how stupid our rules are for using cylinders, 2 minute rules, in air task changes, etc. The rules are what they are. The question is, can this software help a pilot to deal with the rules as they are?



John Cochrane

I spent most of three seasons using XCS for US contests. On Ipaq 3950s to start, later on Dell Streak 5. Nine or ten contests IIRC. Something important was always busted in software. Start rules, AAT task optimizer, little things :-/. I wrote lots of trouble tickets. Some got acted on right away, some never. I haven't used XCS since version 6.3.

There's a timer that can be used to tell you how long you have been under max height.

When things work in XCS, they work well. Task editing is easy on the ground. Task editing in flight depends on your interface, and all of the hardware that I am familiar with that runs XCSoar is touch screen driven. I grew to hate touch screens in flight. My big problem with a touch screen is that I have to look the display for every single poke. This sounds trivial. In fact it adds up to a lot of distraction.

There are better solutions for racing. The two that I am aware of that demonstrably work (i.e. guys win with them, a lot) are SN10 and ClearNav.

The reason that these devices work better is that they take about a quarter the head down time that XCSoar does. Better user interface, easier to read displays in about equal parts. Neither attempts to do nearly as much as XCSoar can, but an awful lot of what XCSoar can do isn't helpful for racing.

The best clues to efficient soaring are outside the window. Less distraction equals better performance in my cockpit, and probably yours too.

By way of disclosure, I work part time for CNi, however this post is my personal opinion, informed by experience.


Evan Ludeman / T8

On Monday, March 11, 2013 2:33:19 PM UTC-4, Max Kellermann wrote:
On Monday, March 11, 2013 7:11:25 PM UTC+1, Evan Ludeman wrote:

I spent most of three seasons using XCS for US contests. On Ipaq 3950s to start, later on Dell Streak 5. Nine or ten contests IIRC. Something important was always busted in software. Start rules, AAT task optimizer, little things :-/. I wrote lots of trouble tickets. Some got acted on right away, some never.



Some never? The following points to a exhaustive list of all of your tickets we havn't worked on yet:



http://bugs.xcsoar.org/query?status=...%40g mail.com



This is exactly one ticket, one that can't be any less serious.



Yes, you found a lot of problems over the years (totalling 24), and that was your contribution to a better XCSoar, it helped a lot. Thousands of eyes find thousands of problems in such a complex software (nearly 26,000 users on Android alone). We do our best to deliver high quality to all those pilots, and we need input from people like you.



But please be honest. I know you want to drop some advertisement for your employer, but don't exaggerate the severity and number of open bugs.



Actually, inability to d/l waypoints to a 302 is a pretty serious defect *for me* since I use the 303 display routinely and travel to many different soaring sites (but only because the 302 has small memory and no way to manage multiple waypoint sets). Without that capability, I have to drag out a laptop, pop the canopy off to get under the panel, etc. That much of the interface was all very handy with the little Ipags that could run the Cambridge utility.

The fact remains *every* release of XCSoar I used had significant bugs that affected its utility in competition. If you've fixed some of the tickets I left behind since I bothered to look (roughly last June), thank you, perhaps it will help others.

Evan Ludeman / T8

On Monday, March 11, 2013 2:49:06 PM UTC-4, Luke Szczepaniak wrote:

From my experience last year, V6.4 was very stable, all rules and

optimizations worked as necessary.

That's interesting. The task optimizer -- which worked marvelously well for me in about July of 2011 -- was busted as of late Summer 2011 and I d/l'ed whatever release was current late Summer 2012 (i.e. a year later) just to see if that had been fixed... and found it hadn't. In my case, I always had to enter silly high Mc settings to get reasonable expected XC speeds. I got blown off on that... I guess it wasn't judged to be "serious" :-). The suggestion to add a dirt simple manual override for projected XC speed was likewise rejected.

I agree that figuring out a way to minimize the poking and prodding is a good thing. I liked the visual presentation of the Streak.

Evan Ludeman / T8

On Monday, March 11, 2013 7:57:17 PM UTC+1, Evan Ludeman wrote:
If you've fixed some of the tickets I left behind since I bothered to look (roughly last June), thank you, perhaps it will help others.

Not true, your memory doesn't serve you well.

The latest ticket we touched was #2034, fixed on 2012/05/03, reported by you on 2012/04/24. That was just 3 days after you wrote your last ticket, and 6 days before your last XCSoar mailing list post.

On Monday, March 11, 2013 3:24:58 PM UTC-4, Max Kellermann wrote:
On Monday, March 11, 2013 7:57:17 PM UTC+1, Evan Ludeman wrote:

If you've fixed some of the tickets I left behind since I bothered to look (roughly last June), thank you, perhaps it will help others.



Not true, your memory doesn't serve you well.



The latest ticket we touched was #2034, fixed on 2012/05/03, reported by you on 2012/04/24. That was just 3 days after you wrote your last ticket, and 6 days before your last XCSoar mailing list post.

What I am thinking of is consistently bad wind information at high altitude in wave and that damned task optimizer, mentioned in my reply to Luke. I'll leave you to look up the details if you like (and you do seem to like) but these things were not "fixed" while I was around, even if someone closed the tickets.

T8

I suppose you are talking about the AAT task optimizer? I've flown two competitions with it last year, and it worked as well as always before... It would be great though if you could point us to the first version that behaved badly for you, because otherwise we can only do wild guesses too.

For me the connectivity of Android based systems is important. I
like to connect with FLARM, variometer and logger. (I am also
dreaming of GPS-NAV communication with XCSoar).
After using Android device going back to Windows CE is not easy.
Mobile communication sends my position into web server, and
maybe I can track my friends online while flying.
User interface is IMHO the problem of all these touch devices.
Simple buttons or rotary switch would be better in turbulence.

For me the most important feature is the moving map and airspace
limits and FLARM traffic view. Then maybe the final glide
calculation. I fly with voice variometer.

On Monday, March 11, 2013 4:10:31 PM UTC-4, Tobias Bieniek wrote:
I suppose you are talking about the AAT task optimizer? I've flown two competitions with it last year, and it worked as well as always before... It would be great though if you could point us to the first version that behaved badly for you, because otherwise we can only do wild guesses too.

The task calculator, yes.

There was a problem early on in 6.3 that did get sorted out. That problem caused my XC speed predicted vs Mc setting to run very high. I wrote a ticket, it got worked on, many thanks.

Later on, I noticed the inverse: entering for instance Mc 3.0 (knots) would yield an anomalously low predicted XC speed, low 40 mph range. This would have been with what ever version of XCS was current in April/May 2012, 6.3..something. I pulled out my Dell Streak to see if I could reproduce the problem quickly and I cannot, so I withdraw the shot about not being fixed a year later. The version I have loaded is 6.4.5. Mc 3.0 yields 51 mph which is about right if cruising is done in still air for dry ASW-20A, which is the polar I have loaded. Of course we always try to cruise in better than still air, so it's a good thing to be able to tell the nav device that "yes, the next 40 miles are going to be *really* good" due to ridge, cloud street, convergence or what have you. Having to set the Mc value, observe the speed calculated, deciding if it's reasonable, adjusting again maybe and *then* looking at the turn areas to see how far I need to go is a gigantic pain in the ass and an unreasonable waste of time in the cockpit. It's much easier for me to simply estimate speed (based on experience and what I see ahead) and work from this. I find that I can estimate speed quite well for days that go well. On the days that don't go well, I don't care about speed, I just care about getting around and the calculator is of no real import. Anyhow, my $0.02. Sorry I can't shed any illumination on whatever the issue was, perhaps it really is all fixed now.

T8