paparazzi user edition?

classic Classic list List threaded Threaded
20 messages Options
Reply | Threaded
Open this post in threaded view
|

paparazzi user edition?

Marvin Wang
hi

i have been flied as a user for about 4 years,begin in 2008. At that time the project was the only one open source UAV in the world. i was so excited flying with it. and 4 years has passed,now there are more open source UAV project, they are easier to use,has better UI, config pages that you can quickly config your airframe. 

paparazzi project has a more powerful code,so i think if we can develop an user edition to make paparazzi easier to use for beginners? i think the user edition will have these features:

1. better GCS UI, that have everything needed on one screen. you can view board/airplane condition here. as well as  multifunction PFD
2. support for MAVLINK.
3. airframe/board config pages. you can config most popular Fixwing/Multimotor/Heli by graphical interface
4. different language support.

i have some picture of the openpilot gcs(attached files).

is paparazzi user edition possible in the near future? if so, what can i do to help?


Cheers
Wang

_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel

config.png (170K) Download Attachment
GCS.png (512K) Download Attachment
radio.png (150K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: paparazzi user edition?

Aka-7
Hi Wang,

on the configuration I agree, paparazzi is hard to learn and could be improved by a graphical interface for that. At least for the basic setup of standard configurations. Question is who of the current users, that have working systems is interested in putting efforts in the programming when there are more urgent tasks like having e.g. a stable toolchain for STM32 autopilots...
Probably it is possible to adapt some parts of the Open-Pilot configuration utilities for paprazzi? Both are open source so could be possible.

Regarding GCS I have to disagree and state that Paparazzi is still the most usable open system, even better than some commercial ones...
The only drawback I see is the programming in OCAML which makes improvements hard.

In general the different projects just have different scopes regarding users and preffered field of use. Paparazzi comes out of the research environment and is still widely used there and this is part of the reason why it is the way it is.

Cheers

Marc

-------- Original-Nachricht --------
> Datum: Wed, 23 May 2012 11:04:34 +0800
> Von: wang yao <[hidden email]>
> An: paparazzi-devel <[hidden email]>
> Betreff: [Paparazzi-devel] paparazzi user edition?

> hi
>
> i have been flied as a user for about 4 years,begin in 2008. At that time
> the project was the only one open source UAV in the world. i was so
> excited
> flying with it. and 4 years has passed,now there are more open source UAV
> project, they are easier to use,has better UI, config pages that you can
> quickly config your airframe.
>
> paparazzi project has a more powerful code,so i think if we can develop an
> user edition to make paparazzi easier to use for beginners? i think the
> user edition will have these features:
>
> 1. better GCS UI, that have everything needed on one screen. you can view
> board/airplane condition here. as well as  multifunction PFD
> 2. support for MAVLINK.
> 3. airframe/board config pages. you can config most popular
> Fixwing/Multimotor/Heli by graphical interface
> 4. different language support.
>
> i have some picture of the openpilot gcs(attached files).
>
> is paparazzi user edition possible in the near future? if so, what can i
> do
> to help?
>
>
> Cheers
> Wang

--
NEU: FreePhone 3-fach-Flat mit kostenlosem Smartphone!                                  
Jetzt informieren: http://mobile.1und1.de/?ac=OM.PW.PW003K20328T7073a

_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
Reply | Threaded
Open this post in threaded view
|

Re: paparazzi user edition?

Hwarm
In reply to this post by Marvin Wang
Hello Wang,
Openpilot is using a Graphic software from NOKIA witch is free for private users but not GPL.
If this software is  used commercially  it cost  licensee.
Since Openpilot use the stm32 it is possible to make a port for the new and future  Paparzzi hardware.
There is also a interest to use the same RTOS  (FeeRTOS) with Paparazzi.
Also  the Diy Drones project will switch to 32 bit hardware next year.
Paparazzi is most used in the research field and has also same private users.
Since about 3-6 month are  necessary to jump into the structures to make own changes and improvements
this time is to large for the most  private users with no interest in informatics and software development..

Regards

Heinrich
.



wang yao schrieb:
hi

i have been flied as a user for about 4 years,begin in 2008. At that time the project was the only one open source UAV in the world. i was so excited flying with it. and 4 years has passed,now there are more open source UAV project, they are easier to use,has better UI, config pages that you can quickly config your airframe. 

paparazzi project has a more powerful code,so i think if we can develop an user edition to make paparazzi easier to use for beginners? i think the user edition will have these features:

1. better GCS UI, that have everything needed on one screen. you can view board/airplane condition here. as well as  multifunction PFD
2. support for MAVLINK.
3. airframe/board config pages. you can config most popular Fixwing/Multimotor/Heli by graphical interface
4. different language support.

i have some picture of the openpilot gcs(attached files).

is paparazzi user edition possible in the near future? if so, what can i do to help?


Cheers
Wang










_______________________________________________ Paparazzi-devel mailing list [hidden email] https://lists.nongnu.org/mailman/listinfo/paparazzi-devel

_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
Reply | Threaded
Open this post in threaded view
|

Re: paparazzi user edition?

Chris Gough-2
In reply to this post by Aka-7
Editing XML is not pretty, but it's not difficult either; the
challenge is "knowing what edit to make". I don't think simply
replacing XML with a forms and icons would make it significantly
easier to managing airframe configuration, unless the GUI also
provided helpful guidance. Because paparazzi is so flexible, it would
take a lot of "hardcoded GUI logic" to replace the current RTFM system
(time and effort in the wiki).

I'm currently starting to experiment with using formal logic
programming for preflight checking (with http://pyke.sourceforge.com/,
like Prolog but python) based on rules and observations about the
system (flight plan, range safety plan, standard operating procedures,
airframe manual, etc). My vision is for some sort of "conversational
AI" that only asks the operator exactly what it needs to know to
accomplish some particular goal, such as prove that the necessary
preconditions have been met to authorise takeoff.

I chose preflight checking because it seemed the simplest/easiest
place to start, and because complying with a big list of rules makes
intuitive sense. But configuration management another areas where the
logic programming approach might be useful. Especially given
Paparazzi's flexability and potential compexity. Anyone interested in
collaborating on this? throwing some students at it? :)

Chris Gough


On Wed, May 23, 2012 at 4:58 PM, Marc Schwarzbach
<[hidden email]> wrote:

> Hi Wang,
>
> on the configuration I agree, paparazzi is hard to learn and could be improved by a graphical interface for that. At least for the basic setup of standard configurations. Question is who of the current users, that have working systems is interested in putting efforts in the programming when there are more urgent tasks like having e.g. a stable toolchain for STM32 autopilots...
> Probably it is possible to adapt some parts of the Open-Pilot configuration utilities for paprazzi? Both are open source so could be possible.
>
> Regarding GCS I have to disagree and state that Paparazzi is still the most usable open system, even better than some commercial ones...
> The only drawback I see is the programming in OCAML which makes improvements hard.
>
> In general the different projects just have different scopes regarding users and preffered field of use. Paparazzi comes out of the research environment and is still widely used there and this is part of the reason why it is the way it is.
>
> Cheers
>
> Marc
>
> -------- Original-Nachricht --------
>> Datum: Wed, 23 May 2012 11:04:34 +0800
>> Von: wang yao <[hidden email]>
>> An: paparazzi-devel <[hidden email]>
>> Betreff: [Paparazzi-devel] paparazzi user edition?
>
>> hi
>>
>> i have been flied as a user for about 4 years,begin in 2008. At that time
>> the project was the only one open source UAV in the world. i was so
>> excited
>> flying with it. and 4 years has passed,now there are more open source UAV
>> project, they are easier to use,has better UI, config pages that you can
>> quickly config your airframe.
>>
>> paparazzi project has a more powerful code,so i think if we can develop an
>> user edition to make paparazzi easier to use for beginners? i think the
>> user edition will have these features:
>>
>> 1. better GCS UI, that have everything needed on one screen. you can view
>> board/airplane condition here. as well as  multifunction PFD
>> 2. support for MAVLINK.
>> 3. airframe/board config pages. you can config most popular
>> Fixwing/Multimotor/Heli by graphical interface
>> 4. different language support.
>>
>> i have some picture of the openpilot gcs(attached files).
>>
>> is paparazzi user edition possible in the near future? if so, what can i
>> do
>> to help?
>>
>>
>> Cheers
>> Wang
>
> --
> NEU: FreePhone 3-fach-Flat mit kostenlosem Smartphone!
> Jetzt informieren: http://mobile.1und1.de/?ac=OM.PW.PW003K20328T7073a
>
> _______________________________________________
> Paparazzi-devel mailing list
> [hidden email]
> https://lists.nongnu.org/mailman/listinfo/paparazzi-devel



--
.

_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
Reply | Threaded
Open this post in threaded view
|

Re: paparazzi user edition?

Tilman Baumann-3
In reply to this post by Marvin Wang
I think a lower hanging fruit and much more effective for beginners would be some ready to fly setups.

Clear straight forward hardware basis. Some cheap common airframe like from hobbyking that everyone can buy.
Straight forward wiring diagram and RC integration.
Parts that can be easily acquired everywhere (outside US).
And a pre-tuned airframe config that 'just works'™. Also with good annotations.
Perhaps a way to generate that from some 'meta' file which describes which functional blocks are used. (Telemetry type, IMU type)

Same for the GCS layout.

And perhaps some way of creating basic flight plans. Perhaps online with google maps...

I don't think the systems complexity is really a problem. Except for the 'get you going first' stage.
I'm not sure if it matters that it does not run on Windows. These days...


Ideally someone would start a kit business like diydrones. Where you can get functional units like pre-bound telemetry sets or pre-calibrated IMU.
All with the right wiring harness and so on...
There are quite a few drones or FPV shops. Has any of them ever voiced any interest?


On 23/05/12 04:04, wang yao wrote:
hi

i have been flied as a user for about 4 years,begin in 2008. At that time the project was the only one open source UAV in the world. i was so excited flying with it. and 4 years has passed,now there are more open source UAV project, they are easier to use,has better UI, config pages that you can quickly config your airframe. 

paparazzi project has a more powerful code,so i think if we can develop an user edition to make paparazzi easier to use for beginners? i think the user edition will have these features:

1. better GCS UI, that have everything needed on one screen. you can view board/airplane condition here. as well as  multifunction PFD
2. support for MAVLINK.
3. airframe/board config pages. you can config most popular Fixwing/Multimotor/Heli by graphical interface
4. different language support.

i have some picture of the openpilot gcs(attached files).

is paparazzi user edition possible in the near future? if so, what can i do to help?


Cheers
Wang


_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel


_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
Reply | Threaded
Open this post in threaded view
|

Re: paparazzi user edition?

Christophe De Wagter
In reply to this post by Chris Gough-2
Even for developers it's not easy to know what to edit in xml. I personally very often find myself doing:

ls ./conf/autopilot/*.makefile

in order to know the current names of subsytems and 

cat ./sw/airborne/modules/digital_cam/*.h | grep -n2 define 

to know the name of the define and

cat ./conf/modules/d*.xml 
or
cat ./conf/boards/lisa-m*.makefile

to know which parameters can be configured.

This is the kind of stuff that could easily go into a (autogenerated) editor.

Question I have is: which programming language to write this into. OCAML would be an obvious choice, with its built-in xml + loads of libs and would probably only be a few pages of code. But suppose we for once would not use OCAML, then what would be a better choice?

-Christophe 



On Wed, May 23, 2012 at 10:01 AM, Chris Gough <[hidden email]> wrote:
Editing XML is not pretty, but it's not difficult either; the
challenge is "knowing what edit to make". I don't think simply
replacing XML with a forms and icons would make it significantly
easier to managing airframe configuration, unless the GUI also
provided helpful guidance. Because paparazzi is so flexible, it would
take a lot of "hardcoded GUI logic" to replace the current RTFM system
(time and effort in the wiki).

I'm currently starting to experiment with using formal logic
programming for preflight checking (with http://pyke.sourceforge.com/,
like Prolog but python) based on rules and observations about the
system (flight plan, range safety plan, standard operating procedures,
airframe manual, etc). My vision is for some sort of "conversational
AI" that only asks the operator exactly what it needs to know to
accomplish some particular goal, such as prove that the necessary
preconditions have been met to authorise takeoff.

I chose preflight checking because it seemed the simplest/easiest
place to start, and because complying with a big list of rules makes
intuitive sense. But configuration management another areas where the
logic programming approach might be useful. Especially given
Paparazzi's flexability and potential compexity. Anyone interested in
collaborating on this? throwing some students at it? :)

Chris Gough


On Wed, May 23, 2012 at 4:58 PM, Marc Schwarzbach
<[hidden email]> wrote:
> Hi Wang,
>
> on the configuration I agree, paparazzi is hard to learn and could be improved by a graphical interface for that. At least for the basic setup of standard configurations. Question is who of the current users, that have working systems is interested in putting efforts in the programming when there are more urgent tasks like having e.g. a stable toolchain for STM32 autopilots...
> Probably it is possible to adapt some parts of the Open-Pilot configuration utilities for paprazzi? Both are open source so could be possible.
>
> Regarding GCS I have to disagree and state that Paparazzi is still the most usable open system, even better than some commercial ones...
> The only drawback I see is the programming in OCAML which makes improvements hard.
>
> In general the different projects just have different scopes regarding users and preffered field of use. Paparazzi comes out of the research environment and is still widely used there and this is part of the reason why it is the way it is.
>
> Cheers
>
> Marc
>
> -------- Original-Nachricht --------
>> Datum: Wed, 23 May 2012 11:04:34 +0800
>> Von: wang yao <[hidden email]>
>> An: paparazzi-devel <[hidden email]>
>> Betreff: [Paparazzi-devel] paparazzi user edition?
>
>> hi
>>
>> i have been flied as a user for about 4 years,begin in 2008. At that time
>> the project was the only one open source UAV in the world. i was so
>> excited
>> flying with it. and 4 years has passed,now there are more open source UAV
>> project, they are easier to use,has better UI, config pages that you can
>> quickly config your airframe.
>>
>> paparazzi project has a more powerful code,so i think if we can develop an
>> user edition to make paparazzi easier to use for beginners? i think the
>> user edition will have these features:
>>
>> 1. better GCS UI, that have everything needed on one screen. you can view
>> board/airplane condition here. as well as  multifunction PFD
>> 2. support for MAVLINK.
>> 3. airframe/board config pages. you can config most popular
>> Fixwing/Multimotor/Heli by graphical interface
>> 4. different language support.
>>
>> i have some picture of the openpilot gcs(attached files).
>>
>> is paparazzi user edition possible in the near future? if so, what can i
>> do
>> to help?
>>
>>
>> Cheers
>> Wang
>
> --
> NEU: FreePhone 3-fach-Flat mit kostenlosem Smartphone!
> Jetzt informieren: http://mobile.1und1.de/?ac=OM.PW.PW003K20328T7073a
>
> _______________________________________________
> Paparazzi-devel mailing list
> [hidden email]
> https://lists.nongnu.org/mailman/listinfo/paparazzi-devel



--
.

_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel


_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
Reply | Threaded
Open this post in threaded view
|

Re: paparazzi user edition?

Andy Brown-4
In reply to this post by Tilman Baumann-3
As someone still getting their head around the entire system and tool chain (for the last several months actually) my observations are as follows:

- it's quite hard to get things going on non-Linux systems and once you do it's difficult to get Linux and Mac based systems to talk (hard coded subnet issue)

- I don't care for ivy bus very much at all. It's quite brittle and only works on a single subnet.

- storing all the config files within the main repository makes keeping things updated pretty difficult - id love to see most of the XML get its own repo and the reference with a git submodule.

- The config files are quite opaque and there is not much documentation that is useful for someone not familiar with flying or telemetry or gps - (I'm just a lowly iOS programmer and network engineer type :)

- fixed number of waypoints are a bummer and it's all pretty brittle.

So what are you gunna do about it random guy from the internet you ask? Here is what I'm working on. Feedback welcome:

- tool chain is hard: I'm working on prebuilt amazon machine images running paparazzi and some better automated scripts to keep them up to date. I'll let the list know when I get the first version done so the list can try it out. My ultimate goal is to be able to configure and build from the remote system and just download the compiled result to be transferred to  the micro controller.

- ground station difficult to manage: Continue work iPad ground station based on the iGCS project from github. I have made progress and can dynamically bind to ivy bus telemetry from a ground station based as long as I have a copy of the xml config files on the iPad (writing my own parser for these was as fun as it sounds). I'm going to publish my work on Github as soon as I can get my fork out it's svn repo and up. It's by no means complete but I'm happy with the start. Eventually I want this on android tablets as well.

- brittle system: Finally I just want to try and understand the guts of the system better so I can start to help make it a bit more user friendly, less brittle and more robust. (I'm also going to need to learn Ocaml)

I hope I can help and would be interested in collaborating on making the entire system easier to use - we should not have to fight our tools and I would love to help build on the awesome work everyone has done!

- Andy Brown
IRC/twitter: magicbeef 
Team evaForge
San Francisco, California

Future work will be posted to github.com/evaforge

Feel free to contact me off list as well.






On May 23, 2012, at 2:43 AM, Tilman Baumann <[hidden email]> wrote:

I think a lower hanging fruit and much more effective for beginners would be some ready to fly setups.

Clear straight forward hardware basis. Some cheap common airframe like from hobbyking that everyone can buy.
Straight forward wiring diagram and RC integration.
Parts that can be easily acquired everywhere (outside US).
And a pre-tuned airframe config that 'just works'™. Also with good annotations.
Perhaps a way to generate that from some 'meta' file which describes which functional blocks are used. (Telemetry type, IMU type)

Same for the GCS layout.

And perhaps some way of creating basic flight plans. Perhaps online with google maps...

I don't think the systems complexity is really a problem. Except for the 'get you going first' stage.
I'm not sure if it matters that it does not run on Windows. These days...


Ideally someone would start a kit business like diydrones. Where you can get functional units like pre-bound telemetry sets or pre-calibrated IMU.
All with the right wiring harness and so on...
There are quite a few drones or FPV shops. Has any of them ever voiced any interest?


On 23/05/12 04:04, wang yao wrote:
hi

i have been flied as a user for about 4 years,begin in 2008. At that time the project was the only one open source UAV in the world. i was so excited flying with it. and 4 years has passed,now there are more open source UAV project, they are easier to use,has better UI, config pages that you can quickly config your airframe. 

paparazzi project has a more powerful code,so i think if we can develop an user edition to make paparazzi easier to use for beginners? i think the user edition will have these features:

1. better GCS UI, that have everything needed on one screen. you can view board/airplane condition here. as well as  multifunction PFD
2. support for MAVLINK.
3. airframe/board config pages. you can config most popular Fixwing/Multimotor/Heli by graphical interface
4. different language support.

i have some picture of the openpilot gcs(attached files).

is paparazzi user edition possible in the near future? if so, what can i do to help?


Cheers
Wang


_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel

_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel

_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
Reply | Threaded
Open this post in threaded view
|

Re: paparazzi user edition?

agressiva
In reply to this post by Christophe De Wagter
Paparazzi is almost perfect to me.

The major problem is:
1- lack of documentation about modules (EX: photogrametry, cam etc..)
2- the way waypoints work in GCS.
    all other autopilot have full dinamic waypoint creation.
    PPZ use hardcoded waypoint creation so the way it would be made easy is a way to load an archive with waypoint position´s.
    When this archieve is load the HARDCODED waypoint are re-alocated.
    In this way i can use a pre-formated flightplan with 200waypoint and the waypoint loaded file re-arrange all waypoints.
    I an not a good programer but i will work in this solution with a friend that is a good C programmer.




Date: Wed, 23 May 2012 16:06:04 +0200
From: [hidden email]
To: [hidden email]
Subject: Re: [Paparazzi-devel] paparazzi user edition?

Even for developers it's not easy to know what to edit in xml. I personally very often find myself doing:

ls ./conf/autopilot/*.makefile

in order to know the current names of subsytems and 

cat ./sw/airborne/modules/digital_cam/*.h | grep -n2 define 

to know the name of the define and

cat ./conf/modules/d*.xml 
or
cat ./conf/boards/lisa-m*.makefile

to know which parameters can be configured.

This is the kind of stuff that could easily go into a (autogenerated) editor.

Question I have is: which programming language to write this into. OCAML would be an obvious choice, with its built-in xml + loads of libs and would probably only be a few pages of code. But suppose we for once would not use OCAML, then what would be a better choice?

-Christophe 



On Wed, May 23, 2012 at 10:01 AM, Chris Gough <[hidden email]> wrote:
Editing XML is not pretty, but it's not difficult either; the
challenge is "knowing what edit to make". I don't think simply
replacing XML with a forms and icons would make it significantly
easier to managing airframe configuration, unless the GUI also
provided helpful guidance. Because paparazzi is so flexible, it would
take a lot of "hardcoded GUI logic" to replace the current RTFM system
(time and effort in the wiki).

I'm currently starting to experiment with using formal logic
programming for preflight checking (with http://pyke.sourceforge.com/,
like Prolog but python) based on rules and observations about the
system (flight plan, range safety plan, standard operating procedures,
airframe manual, etc). My vision is for some sort of "conversational
AI" that only asks the operator exactly what it needs to know to
accomplish some particular goal, such as prove that the necessary
preconditions have been met to authorise takeoff.

I chose preflight checking because it seemed the simplest/easiest
place to start, and because complying with a big list of rules makes
intuitive sense. But configuration management another areas where the
logic programming approach might be useful. Especially given
Paparazzi's flexability and potential compexity. Anyone interested in
collaborating on this? throwing some students at it? :)

Chris Gough


On Wed, May 23, 2012 at 4:58 PM, Marc Schwarzbach
<[hidden email]> wrote:
> Hi Wang,
>
> on the configuration I agree, paparazzi is hard to learn and could be improved by a graphical interface for that. At least for the basic setup of standard configurations. Question is who of the current users, that have working systems is interested in putting efforts in the programming when there are more urgent tasks like having e.g. a stable toolchain for STM32 autopilots...
> Probably it is possible to adapt some parts of the Open-Pilot configuration utilities for paprazzi? Both are open source so could be possible.
>
> Regarding GCS I have to disagree and state that Paparazzi is still the most usable open system, even better than some commercial ones...
> The only drawback I see is the programming in OCAML which makes improvements hard.
>
> In general the different projects just have different scopes regarding users and preffered field of use. Paparazzi comes out of the research environment and is still widely used there and this is part of the reason why it is the way it is.
>
> Cheers
>
> Marc
>
> -------- Original-Nachricht --------
>> Datum: Wed, 23 May 2012 11:04:34 +0800
>> Von: wang yao <[hidden email]>
>> An: paparazzi-devel <[hidden email]>
>> Betreff: [Paparazzi-devel] paparazzi user edition?
>
>> hi
>>
>> i have been flied as a user for about 4 years,begin in 2008. At that time
>> the project was the only one open source UAV in the world. i was so
>> excited
>> flying with it. and 4 years has passed,now there are more open source UAV
>> project, they are easier to use,has better UI, config pages that you can
>> quickly config your airframe.
>>
>> paparazzi project has a more powerful code,so i think if we can develop an
>> user edition to make paparazzi easier to use for beginners? i think the
>> user edition will have these features:
>>
>> 1. better GCS UI, that have everything needed on one screen. you can view
>> board/airplane condition here. as well as  multifunction PFD
>> 2. support for MAVLINK.
>> 3. airframe/board config pages. you can config most popular
>> Fixwing/Multimotor/Heli by graphical interface
>> 4. different language support.
>>
>> i have some picture of the openpilot gcs(attached files).
>>
>> is paparazzi user edition possible in the near future? if so, what can i
>> do
>> to help?
>>
>>
>> Cheers
>> Wang
>
> --
> NEU: FreePhone 3-fach-Flat mit kostenlosem Smartphone!
> Jetzt informieren: http://mobile.1und1.de/?ac=OM.PW.PW003K20328T7073a
>
> _______________________________________________
> Paparazzi-devel mailing list
> [hidden email]
> https://lists.nongnu.org/mailman/listinfo/paparazzi-devel



--
.

_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel


_______________________________________________ Paparazzi-devel mailing list [hidden email] https://lists.nongnu.org/mailman/listinfo/paparazzi-devel

_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
Reply | Threaded
Open this post in threaded view
|

Re: paparazzi user edition?

Chris Gough-2
In reply to this post by Andy Brown-4
Hi Andy

> - tool chain is hard: I'm working on prebuilt amazon machine images running
> paparazzi and some better automated scripts to keep them up to date. I'll
> let the list know when I get the first version done so the list can try it
> out. My ultimate goal is to be able to configure and build from the remote
> system and just download the compiled result to be transferred to  the micro
> controller.

please elaborate :)

I've thought about a web-based tool for configuring and compiling
firmware, to all-but eliminate tool-chain issues for nonstandard
operating systems. It would require a tool for flashing firmware, web
access, and a way of managing files (I still don't know how to do that
on my stupid phone).

I imagined it might run on something like a raspberry-pi in the ground
segment (or airframe) as well as an online service. That way phones,
tablets and workstations could all use it.

Chris Gough

_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
Reply | Threaded
Open this post in threaded view
|

Re: paparazzi user edition?

Gareth Roberts-4
Hi all,

> please elaborate :)
>
> I've thought about a web-based tool for configuring and compiling
> firmware

+1, I've got some source code kicking around somewhere.  The trick is:
- trusting the remote compiler
- securing the remote against malicious attack code

> And a pre-tuned airframe config that 'just works'™. Also with good  
> annotations.

I was thinking about a seperate git repo with the conf files in, plus a  
simple json API.
This could be parsed clientside and turned into a nice list.
Choose airframe -> choose autopilot -> choose imu or IR -> choose radio ->  
choose telemetry
That would get you 90% of the way, and could be done by simply adding  
annotations to the xml files.
As an example:




On Wed, 23 May 2012 16:04:02 +0100, Chris Gough  
<[hidden email]> wrote:

> Hi Andy
>
>> - tool chain is hard: I'm working on prebuilt amazon machine images  
>> running
>> paparazzi and some better automated scripts to keep them up to date.  
>> I'll
>> let the list know when I get the first version done so the list can try  
>> it
>> out. My ultimate goal is to be able to configure and build from the  
>> remote
>> system and just download the compiled result to be transferred to  the  
>> micro
>> controller.
>
> please elaborate :)
>
> I've thought about a web-based tool for configuring and compiling
> firmware, to all-but eliminate tool-chain issues for nonstandard
> operating systems. It would require a tool for flashing firmware, web
> access, and a way of managing files (I still don't know how to do that
> on my stupid phone).
>
> I imagined it might run on something like a raspberry-pi in the ground
> segment (or airframe) as well as an online service. That way phones,
> tablets and workstations could all use it.
>
> Chris Gough
>
> _______________________________________________
> Paparazzi-devel mailing list
> [hidden email]
> https://lists.nongnu.org/mailman/listinfo/paparazzi-devel

_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
Reply | Threaded
Open this post in threaded view
|

Re: paparazzi user edition?

Gareth Roberts-4
In reply to this post by Christophe De Wagter
Would it be worth adding some descriptions to the xml files currently in  
the repo?
For example:

   <description>
        <model name="Mini Swift" manufacturer="HobbyKing" type="fixed-wing" />
        <image  
src="http://www.hobbyking.com/hobbyking/store/catalog/Mini-Swift.jpg" />
        <author name="Dr Jones" email="[hidden email]" />
        <notes>
        <![CDATA[
      Made with EPO, this small glider is both agile and very resliant to  
crashes.
    ]]>
        </notes>
   </description>

This could be enforced in the DTD, and a seperate conf file git repo  
parsed out by a simple API server in the future to provide a wizard...
Choose airframe -> Choose autopilot -> Choose IR/IMU sensor -> Choose  
radio -> Choose telemetry
and you get a 90% there xml.

A major problem I forsee is the current coupling between airframe  
(servo/gains etc) data & autopilot data in the airframe file.
Perhaps a seperate autopilot.xml would be required.

Cheers,
Gareth

On Wed, 23 May 2012 15:06:04 +0100, Christophe De Wagter  
<[hidden email]> wrote:

> Even for developers it's not easy to know what to edit in xml. I  
> personally
> very often find myself doing:
>
> ls ./conf/autopilot/*.makefile
>
> in order to know the current names of subsytems and
>
> cat ./sw/airborne/modules/digital_cam/*.h | grep -n2 define
>
> to know the name of the define and
>
> cat ./conf/modules/d*.xml
> or
> cat ./conf/boards/lisa-m*.makefile
>
> to know which parameters can be configured.
>
> This is the kind of stuff that could easily go into a (autogenerated)
> editor.
>
> Question I have is: which programming language to write this into. OCAML
> would be an obvious choice, with its built-in xml + loads of libs and  
> would
> probably only be a few pages of code. But suppose we for once would not  
> use
> OCAML, then what would be a better choice?
>
> -Christophe
>
>
>
> On Wed, May 23, 2012 at 10:01 AM, Chris Gough  
> <[hidden email]
>> wrote:
>
>> Editing XML is not pretty, but it's not difficult either; the
>> challenge is "knowing what edit to make". I don't think simply
>> replacing XML with a forms and icons would make it significantly
>> easier to managing airframe configuration, unless the GUI also
>> provided helpful guidance. Because paparazzi is so flexible, it would
>> take a lot of "hardcoded GUI logic" to replace the current RTFM system
>> (time and effort in the wiki).
>>
>> I'm currently starting to experiment with using formal logic
>> programming for preflight checking (with http://pyke.sourceforge.com/,
>> like Prolog but python) based on rules and observations about the
>> system (flight plan, range safety plan, standard operating procedures,
>> airframe manual, etc). My vision is for some sort of "conversational
>> AI" that only asks the operator exactly what it needs to know to
>> accomplish some particular goal, such as prove that the necessary
>> preconditions have been met to authorise takeoff.
>>
>> I chose preflight checking because it seemed the simplest/easiest
>> place to start, and because complying with a big list of rules makes
>> intuitive sense. But configuration management another areas where the
>> logic programming approach might be useful. Especially given
>> Paparazzi's flexability and potential compexity. Anyone interested in
>> collaborating on this? throwing some students at it? :)
>>
>> Chris Gough
>>
>>
>> On Wed, May 23, 2012 at 4:58 PM, Marc Schwarzbach
>> <[hidden email]> wrote:
>> > Hi Wang,
>> >
>> > on the configuration I agree, paparazzi is hard to learn and could be
>> improved by a graphical interface for that. At least for the basic  
>> setup of
>> standard configurations. Question is who of the current users, that have
>> working systems is interested in putting efforts in the programming when
>> there are more urgent tasks like having e.g. a stable toolchain for  
>> STM32
>> autopilots...
>> > Probably it is possible to adapt some parts of the Open-Pilot
>> configuration utilities for paprazzi? Both are open source so could be
>> possible.
>> >
>> > Regarding GCS I have to disagree and state that Paparazzi is still the
>> most usable open system, even better than some commercial ones...
>> > The only drawback I see is the programming in OCAML which makes
>> improvements hard.
>> >
>> > In general the different projects just have different scopes regarding
>> users and preffered field of use. Paparazzi comes out of the research
>> environment and is still widely used there and this is part of the  
>> reason
>> why it is the way it is.
>> >
>> > Cheers
>> >
>> > Marc
>> >
>> > -------- Original-Nachricht --------
>> >> Datum: Wed, 23 May 2012 11:04:34 +0800
>> >> Von: wang yao <[hidden email]>
>> >> An: paparazzi-devel <[hidden email]>
>> >> Betreff: [Paparazzi-devel] paparazzi user edition?
>> >
>> >> hi
>> >>
>> >> i have been flied as a user for about 4 years,begin in 2008. At that
>> time
>> >> the project was the only one open source UAV in the world. i was so
>> >> excited
>> >> flying with it. and 4 years has passed,now there are more open source
>> UAV
>> >> project, they are easier to use,has better UI, config pages that you  
>> can
>> >> quickly config your airframe.
>> >>
>> >> paparazzi project has a more powerful code,so i think if we can  
>> develop
>> an
>> >> user edition to make paparazzi easier to use for beginners? i think  
>> the
>> >> user edition will have these features:
>> >>
>> >> 1. better GCS UI, that have everything needed on one screen. you can
>> view
>> >> board/airplane condition here. as well as  multifunction PFD
>> >> 2. support for MAVLINK.
>> >> 3. airframe/board config pages. you can config most popular
>> >> Fixwing/Multimotor/Heli by graphical interface
>> >> 4. different language support.
>> >>
>> >> i have some picture of the openpilot gcs(attached files).
>> >>
>> >> is paparazzi user edition possible in the near future? if so, what  
>> can i
>> >> do
>> >> to help?
>> >>
>> >>
>> >> Cheers
>> >> Wang
>> >
>> > --
>> > NEU: FreePhone 3-fach-Flat mit kostenlosem Smartphone!
>> > Jetzt informieren: http://mobile.1und1.de/?ac=OM.PW.PW003K20328T7073a
>> >
>> > _______________________________________________
>> > Paparazzi-devel mailing list
>> > [hidden email]
>> > https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
>>
>>
>>
>> --
>> .
>>
>> _______________________________________________
>> Paparazzi-devel mailing list
>> [hidden email]
>> https://lists.nongnu.org/mailman/listinfo/paparazzi-devel

_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
Reply | Threaded
Open this post in threaded view
|

Re: paparazzi user edition?

Andy Brown-4
In reply to this post by Chris Gough-2
So I have spent most of my time fighting to get things working on Mac.
Our team uses Mac, windows and Linux etc but we have wasted a ton of
time on this generally.

I have also wasted a ton of time with our streaming video server
software so last weekend I configured an AMI I can spin up on amazon
and I want to do the same for PPZ.

I think the system would look like this.

- base ubuntu 12.04 system with paparazzi installed and the tool chain working.

- Web front end (or just API) to upload my own XML config files and
kick off the build.

- process to archive and version the xml bundle in a db or on S3 so we
can roll back/diff the config.

- after we have the config files place them approbropraitly in the PPZ
directories or somewhere appropriate.

- kick off a build using the uploaded setting and archive the result
and version it so we know what configs go to what builds.

- using and API and web interface I want to be able to download the
the bin file (I think this is how it works) and tranfer it to the
micro controller - manually first at first.

-  I want to be able to point my ground station (iPad or desktop app)
at an endpoint and have it download the correct config files from the
server and configure itself.

- stretch goal. Gumstick talks to the server, downloads the micro
controller boots from the Gumstick and I don't have to plug anything
in at all.

That's  loosely what I'm going to try to build. Starting with a base
PPZ image I can boot up at will on amazon cloud.

- Andy





On May 23, 2012, at 8:04 AM, Chris Gough <[hidden email]> wrote:

> Hi Andy
>
>> - tool chain is hard: I'm working on prebuilt amazon machine images running
>> paparazzi and some better automated scripts to keep them up to date. I'll
>> let the list know when I get the first version done so the list can try it
>> out. My ultimate goal is to be able to configure and build from the remote
>> system and just download the compiled result to be transferred to  the micro
>> controller.
>
> please elaborate :)
>
> I've thought about a web-based tool for configuring and compiling
> firmware, to all-but eliminate tool-chain issues for nonstandard
> operating systems. It would require a tool for flashing firmware, web
> access, and a way of managing files (I still don't know how to do that
> on my stupid phone).
>
> I imagined it might run on something like a raspberry-pi in the ground
> segment (or airframe) as well as an online service. That way phones,
> tablets and workstations could all use it.
>
> Chris Gough
>
> _______________________________________________
> Paparazzi-devel mailing list
> [hidden email]
> https://lists.nongnu.org/mailman/listinfo/paparazzi-devel

_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
Reply | Threaded
Open this post in threaded view
|

Re: paparazzi user edition?

Tilman Baumann-3
In reply to this post by Chris Gough-2
On 23/05/12 16:04, Chris Gough wrote:
>
> I imagined it might run on something like a raspberry-pi in the ground
> segment (or airframe) as well as an online service. That way phones,
> tablets and workstations could all use it.

Interesting idea actually. I mean something like raspberry as ground
segment.
Because you could run the whole paparazzi suite on that...
Plug in a screen or export via vlc to a windows box and you are good to
go...

I don't need 'yet another unfinished project' but that might actually be
worth doing.
Add all ppz packages to one of the pupular arm distros. like Debian,
Fedora and Openembedded (OpenWRT, Pokey, Ångström and the like are all
Openembedded)
There are not too many hardware platforms to consider, so integration
could be quite tight.
Raspberry-Pi, Beagleboard and Pandaboard are the three big ones I think.

_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
Reply | Threaded
Open this post in threaded view
|

Re: paparazzi user edition?

Chris Gough-2
In reply to this post by Gareth Roberts-4
>> I've thought about a web-based tool for configuring and compiling
>> firmware
>
> +1, I've got some source code kicking around somewhere.  The trick is:
> - trusting the remote compiler

If web tool stored "config files" virtually, in a RDBMS, and rendered
XML from templates as required (much like web pages), there should be
plenty of scope for input sanitisation etc.

> - securing the remote against malicious attack code

For a few reasons, I had imagined a worker process (such as celeryd)
pulling compilation jobs from a message queue (such as RabbitMQ),
rendering the config out before building and then publishing it. So,
assuming careful sanitisation upstream (into the config DB), no real
opportunity to send arbitrary code directly to the native toolchain.
Not foolproof, but not a gaping hole either.

Chris Gough

_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
Reply | Threaded
Open this post in threaded view
|

Re: paparazzi user edition?

Marvin Wang
I think configure wizard  maybe is a good idea.



_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
Reply | Threaded
Open this post in threaded view
|

Re: paparazzi user edition?

flixr
Administrator
Hi all,

just some quick links on what has been discussend/done in that direction so far:
mavlink support
auto-settings related to modules/subsystems
doc node for modules

Cheers, Felix

On Thu, May 24, 2012 at 4:02 PM, wang yao <[hidden email]> wrote:
I think configure wizard  maybe is a good idea.



_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel



_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
Reply | Threaded
Open this post in threaded view
|

Re: paparazzi user edition?

Stephen Dwyer
Hello,

Little more information:
New issue for configuration wizard:
https://github.com/paparazzi/paparazzi/issues/200

Regarding flight plan editing, you can always edit the waypoints
graphically using the existing gcs
(http://paparazzi.enac.fr/wiki/Flight_Plan_Editor). As far as other
flight plan parameters, this is not easy to do because of the awesome
flexibility and customization. Flight plans are far more than a set of
waypoints to fly.

This is the same sort of method other autopilots use (for example,
Micropilot has waypoint files and flightplan files, which by the way
are at least as challenging and less flexible than Paparazzi ones).

Thanks,
-Stephen Dwyer

On Thu, May 24, 2012 at 9:01 AM, Felix Ruess <[hidden email]> wrote:

> Hi all,
>
> just some quick links on what has been discussend/done in that direction so
> far:
> mavlink support
> auto-settings related to modules/subsystems
> doc node for modules
>
> Cheers, Felix
>
> On Thu, May 24, 2012 at 4:02 PM, wang yao <[hidden email]> wrote:
>>
>> I think configure wizard  maybe is a good idea.
>>
>>
>>
>> _______________________________________________
>> Paparazzi-devel mailing list
>> [hidden email]
>> https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
>>
>
>
> _______________________________________________
> Paparazzi-devel mailing list
> [hidden email]
> https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
>

_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
Reply | Threaded
Open this post in threaded view
|

Re: paparazzi user edition?

Tilman Baumann-3
On 24/05/12 17:19, Stephen Dwyer wrote:
> Hello,
>
> Little more information:
> New issue for configuration wizard:
> https://github.com/paparazzi/paparazzi/issues/200
>
> Regarding flight plan editing, you can always edit the waypoints
> graphically using the existing gcs

Not practically when they are on the other side of the globe.
Unless I missed some trick.

Can waypoints be defined as relative to a home waypoint? Ideally be set
to the GPS coordinates in pre-launch.

The places would still be all over, but at least they would be near your
actual flying location.
And for simple 'return to home and loiter' plans no input would be required.
> (http://paparazzi.enac.fr/wiki/Flight_Plan_Editor). As far as other
> flight plan parameters, this is not easy to do because of the awesome
> flexibility and customization. Flight plans are far more than a set of
> waypoints to fly.
Sure. And I think nobody is for removing complexity.
But hiding it for beginners might be a good idea.
Some good practice flight plans with useful options to base individual
customisations on might be the way to go.

I don't know. I think flight plans are a relatively minor problem. But
you get still quickly overwhelmed when using any of the (sadly less than
perfectly annotated) example flightplans from the repo.

_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
Reply | Threaded
Open this post in threaded view
|

Re: paparazzi user edition?

flixr
Administrator
Hi,

Waypoints are most often specified as relative coordinates with x/y: http://paparazzi.enac.fr/wiki/Flight_Plans#Waypoints
Then when you go through the geo init in the flight plan, all your (relative) waypoints will be relocated to the actual location.
So you don't have to pick them from the other side of the globe.

Cheers, Felix

On Thu, May 24, 2012 at 6:34 PM, Tilman Baumann <[hidden email]> wrote:
On 24/05/12 17:19, Stephen Dwyer wrote:
Hello,

Little more information:
New issue for configuration wizard:
https://github.com/paparazzi/paparazzi/issues/200

Regarding flight plan editing, you can always edit the waypoints
graphically using the existing gcs

Not practically when they are on the other side of the globe.
Unless I missed some trick.

Can waypoints be defined as relative to a home waypoint? Ideally be set to the GPS coordinates in pre-launch.

The places would still be all over, but at least they would be near your actual flying location.
And for simple 'return to home and loiter' plans no input would be required.

(http://paparazzi.enac.fr/wiki/Flight_Plan_Editor). As far as other
flight plan parameters, this is not easy to do because of the awesome
flexibility and customization. Flight plans are far more than a set of
waypoints to fly.
Sure. And I think nobody is for removing complexity.
But hiding it for beginners might be a good idea.
Some good practice flight plans with useful options to base individual customisations on might be the way to go.

I don't know. I think flight plans are a relatively minor problem. But you get still quickly overwhelmed when using any of the (sadly less than perfectly annotated) example flightplans from the repo.


_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel


_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
Reply | Threaded
Open this post in threaded view
|

Re: paparazzi user edition?

Tilman Baumann-3
On 24/05/12 17:49, Felix Ruess wrote:
> Hi,
>
> Waypoints are most often specified as relative coordinates with x/y:
> http://paparazzi.enac.fr/wiki/Flight_Plans#Waypoints
> Then when you go through the geo init in the flight plan, all your
> (relative) waypoints will be relocated to the actual location.
> So you don't have to pick them from the other side of the globe.

Cool. I guess I have to research more before I complain.
Will try that soon...


_______________________________________________
Paparazzi-devel mailing list
[hidden email]
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel