Re: Final Glide Computers (was: Crusoe and Mobile Linux...)

[Ole Barnkob Kaas <admin@tet.dk>]

Hello John, on 04-Feb-00, you wrote:
> 
> Hej Ole,
> 
>   A good looking board. In particular, I see that the requirements
> on input power are quite flexible, namely 9-32 Volts DC. Sounds like
> three (or more) 6 Volt gel cells wired in series would work nicely.

And maybe a solarpanel to save powerusage from the cells. There should be a
battery monitor circuit anyway.
By the way, they are also developing a powerpc version (MPC 821 and MPC823).

> 
>> 
>> Background:
>> 
>> Some 18 month ago my gliding club installed 2 DX50FAI flight logger in our
>> Ls-4b and DG300. It's so easy to accomplish a task with those computers and
>> you allways know if you're in gliding range of the home base. But they
>> aren't wery good for club usage - eg. changing the pilot name with the push
>> buttons is a real pain in teh butt. Using a pc is easier, but the sw really
>> suck and is ms-dos only.
> 
>   I think your comment about ergonomics are right on target. The wrong
> type of input mechanism can seriously detract fron the overall usability
> of a unit. Just curious about what your ideas are in the way of input
> hardware, touch screen perhaps?
> 

Maybe. It depends on the required precision to operate - and dont neglect the
mecanical feedback you get when pressing a real button. An alternative would
be soft menu buttons - like the HP-48 calculator. Also, nothing beats a turn
nob - just a single one would be great. Anyway, with open source and modular
design you can just write a driver for whatever bizarre hw you have or
design.

>> Possible features to implement:
>> 
>> - Logging (obvious)
>> - Computing glidepath and other stuff (obvious)
>> - Pilot recognition by keycard (can be made very simple)
>> - Using packet radio to send position back to ground (150 km radius)
>> - Displaying glider position on screen on ground (with packet radio)
>                                                    ^^^^^^^^^^^^^^^^^
>   This is certainly one area where Linux is sure cutting edge compared
> to stupid 'doze, with AX.25 right in the kernel, etc. (I assume you mean
> amateur packet radio?) I guess the only downside, is that you would need
> a seperate amateur transceiver on board for this.
> 

Maybe it's possible to build a tranciever of some digital synthesizer and
stuff - I don't know. You don't need much transmitting power - actually a
normal tranciever might have to be dampened. The height of our antenna is way
beyond what radioamateurs ever dream of :-).
Another problem might be to get a license - well at least here.


>> - A logger in all the club gliders => eliminating writing start-list on
>> ground - Collect all tracks in sql database and make statistics - Display
>> an exact copy of the logbook on screen showing *exactly* where you should
>> write the data.
>> - Your own cool app that will double /your/ glide range ;-) - More...
>> 
>> While I'm waiting for the main hw to be made available, I try to get
>> someone interested in this project and maybe find someone to finance it
>> :-). As soon as the the project starts, there will be a CSV server
>> available (found one for free :-).
> 
>   Well, count me in as interested. I'm sure there are other glider
> pilots on the list who might also have an interest. If you want to
> start a mailing list for the project, Rik Riel (our host) can probably
> help.
> 

Well, if things fail here, I'll return on this issue.

>> First it might be nice with a discussion of how to implement all these
>> features. Should it all be made in C or should it only be the basic and the
>> rest made in eg python? What other tools should be used?
> 
>   I'm also a big Python fan, mostly because it allows you to get
> prototypes up and running fast. You can quickly discover your design
> mistakes and always rewrite key parts in C that need to run fast.

Yep, it's amazing what you can do in python... I've just been using Sketch - a
vector drawing tool made in python - works real good and fast.

> 
>   Personally, I haven't got that far yet. I have still been in a stage of
> reviewing what is available as far as hardware goes. If you haven't
> finalized your own opinion, there are some other options you may want to
> consider like PC-104 modules (see http://www.emjembedded.com/ and others),
> or the ucSimm module which runs ucLinux (see http://www.uclinux.org/simm/)
> 

I don't see any reason why it shouldn't run on other single board computers.
Of cource modifications will have to be made, but hopefully the design should
be modular enough to do that. The idea is that you can put together a flight
computer of mostly standard componets - and extend it with sofisticated hw if
you like.

>   I believe one design goal that should be embraced is portability. This
> may be difficult because of the specialized nature of some of the hardware
> involved. However, this is a real technology growth area, and the future
> may have some interesting surprises. The company where Linus works,
> has already given us a peek at what things may lie ahead...
> 

I agree (as I mentioned earlier). I've focused on the LIAB because the
developer lives just around the corner. Because its GPL. Because it can take
dirty power. Because the expected final price was much lower than it is now
:-( Anyway, I still think it is a good starting point. Also others in NJLUG
(the local LUG) are working on projects with LIAB - an mp3 player and a small
webbrowser with touch display. I find the latter very interesting because of
the display. 

 
          Regards/
                /Ole Kaas

-
Archives of linux-aviation: http://mail.nl.linux.org/lists/linux-aviation/
To unsubscribe: send the command "unsubscribe linux-aviation" in the body
of a mail message to <Majordomo@mail.nl.linux.org>.