Microsoft gibt Roboter-Entwicklung auf: Robotics Team gefeue

[schnaggels]

Too bad

So chances for the new TXT will be very low to get (fully) working with MS RS?

http://winfuture.de/news,83726.html

Thomas

[Defiant]

No one used that anyway...time to upgrade to ROS :)

[schnaggels]

No one? Carel would complain

[vleeuwen]

Thanks Thomas,
I already read the news this morning.
Sad news because we are lossing the nice message flow concept, which brings robot programming on a very high abstract level.
See also the basic resources for this news:
http://microsoft-news.com/microsoft-kil ... tics-team/ and the remarks about the eductional quality of MS-RDS.
http://research.microsoft.com/en-us/col ... obots.aspx

Defiant has it wrong there were some user who also contact me.

The main objective with MS-RDS was to use the FT-API at .NET level.
The results of the two main technologies (CCR and DSS) in MS-RDS has been partial integrated in the newer .NET Frameworks and REST.
For that my focus was already moved to a good .NET framework support for the online mode, including events driven programming for the sensors readings .
See the FtMscLibEx with callbacks for the sensor and button changes (including C++ support).
See the FtMscLibExNet for the .NET Framework 4 support. This enables the use of C#, VB.NET or other .NET dialects for developing programs that also can interact with other MS-Windows resources and distributed programming.
This will be needed to make use of for example: the KINECT libraries, the abstract game controller libraries.
I suppose that there will come .NET Framework support for the TXC too.

ROS does not support MS-Windows
ROS also does not support a simulation engine.
MS-RDS is a lot more than only robot programming. The grafical interface (VPL) made it very suitable for use in educational settings. ROS is not offering this stuff.

Upgrading to ROS is not a real option. To work with ROS the usser needs different set of skills and more knowledge of lower level programing issues.
ROS does not replace MS-RDS as developing system. It is a complete different approache.
ROS is a nice product but for a different target group!

See you all at the FT-Convention!

[Defiant]

ROS also does not support a simulation engine.

And what do you call Gazebo?

[vleeuwen]

Looks nice.
A windows version has been announced for 2015. I hope that it will support .NET framework.

Gazebo isn't an integrated parts of ROS (http://gazebosim.org/tutorials?tut=ros_ ... onnect_ros).
There exist more engines that gave support for serious gaming and testing of robot algorithms, including a nice developing and test environment, for example: Unity 3D.

However the strength of the MS-RDS concept including VPL was the integration of the different parts in one system, based on the high abstract level of message flow concept.
And offering a small learning curve for the end users.

ROS and Gazebo is complete different. Very usable but the target group is different when I compare it with MS-RDS.
I am looking forward to a .NET framework version of ROS. No doubt about that.

[Defiant]

Gazebo isn't an integrated parts of ROS (http://gazebosim.org/tutorials?tut=ros_ ... onnect_ros).

Your link says the opposite ;)

Very usable but the target group is different when I compare it with MS-RDS.

ROS ranges from education to bleeding edge development.

[vleeuwen]

I looks like we have some nice discussion subjects for the FT-Convention.

@Defiant,
We have a different opinion about integration, the abstraction level of robot programming itself and the usability in an educational situation/setting.
VPL is targeting almost the same group of users as RoboPro. For both end users don't need knowledge of text base lower level programming.
ROS ask a lot more knowledge of lower level programming, text base programming, program developing workbenches, etc.
ROS is very common used in higher education research projects. For me as software engineer/technician a reason to look into the system.
But in this discussion I am talking more about secondary education level. Without the need of a long learning curve for text based programming.

I also am looking forward to the .NET framework based version of ROS and an implementation of the message oriented programming concept.

@Thomas,

So chances for the new TXT will be very low to get (fully) working with MS RS?

With a .NET framework support for the TXT, it is not so difficult to implement an integration into the MS-RDS 4 system.
The main problem is that MS-RDS 4 will run to its ends of live time.
The bottleneck will not be the implementation of the FtTXTService but the stability of MS-RDS 4.
So the chance is not low but will it be meaningful?