fllnxtnavigation Discussions Rss Feed

New Post: FLL NXT Navigation Tutorial

weststar42 — Fri, 02 Jan 2009 01:27:42 GMT

Please see the first tutorial video for an overview of how to use FLL NXT Navigation.

New Post: FLL NXT Navigation Approach to Managing Software Repository

weststar42 — Tue, 30 Dec 2008 02:37:48 GMT

FLL NXT Navigation’s Approach to Version Control

FLL NXT Navigation (FNN) uses Codeplex’s Subversion based repository to store all source code, release files, and documentation (see https://fllnxtnavigation.svn.codeplex.com/svn).

FNN uses a single repository for the entire project that follows the standard pattern of /trunk, /branches, and /tags folders with the addition of /reference and /releases folders to track anything that is released on the web site (i.e. executable zip or installation files, videos, or podcasts).

The /release folder allows FNN to have separate internal tags for marking any desired milestones and for specific releases like those found on the Release tab of the FNN site on Codeplex (http://www.codeplex.com/fllnxtnavigation/Release/ProjectReleases.aspx?ReleaseId=21107).

FNN Use of Branches and Tags

Most (if not all) of the development will be done on the main trunk
Tags are used for releases and any internal milestone that bears remembering
Releases are for anything placed in the Codeplex /release folder
Branches are used for significant changes to the code base
Expect a branch for each First Lego League competition season (2008, 2009, …)
Branches are also used for trying out potential solutions, new approaches, or new developers, without tying up the main trunk for everyone else

FNN, Subversion, and Microsoft Visual C++ 2008 Express Edition

FNN was created and is currently maintained in Microsoft Visual C++ 2008 Express Edition. This entirely free and very capable integrated development environment (IDE) does not support a direct plug-in with Subversion (or any revision control system that we could find). However, by using TortiseSVN and using a strategy of managing the entire Visual C++ solution’s folder there is a straightforward to get the advantages of the free IDE and the very capable Subversion.

The details below are geared toward Windows users.

Consider watching the intro to Subversion videos by Jeremy Wosborne (http://www.youtube.com/watch?v=-h6cj6O94Gs). Jeremy walks you through how to use TortiseSVN on a local machine that will be similar to how to use TortiseSVN to develop with FNN.

Download and install TortiseSVN for Windows

Watch the FNN introductions at: [url to be provided]

For Developers Who Want To Experiment with FNN at Home and Outside of the Main Trunk:

Key Steps for Interacting with FNN’s Repository:

Check out the entire Visual C++ solution folder from the repository (https://fllnxtnavigation.svn.codeplex.com/svn) to a local working directory using the TortiseSVN. There are a number of ways to do this. I like to use TortiseSVN Repo-Browser (from the right mouse context menu) to navigate to the desired file in the trunk and then use the checkout feature.
Navigate to the local working directory and open up the solution file (i.e. FLL_NXT_Navigation.sln)
Use Visual C++ to modify code, resources, windows, recompile, etc.
Have fun and post some messages on how it is working for you.

For Developers Who Want To Experiment with FNN at Outside of the Main Trunk And Want to Contribute Back to the Open Source Baseline:

Key Steps for Interacting with FNN’s Repository:

Use TortiseSVN to copy the latest version to a /branch/yourusername folder using the TortiseSVN Repo-Browser\copy to … command.
Check out the entire Visual C++ solution folder from the new branch folder to a local working directory using the TortiseSVN. There are a number of ways to do this. I like to use TortiseSVN Repo-Browser (from the right mouse context menu) to navigate to the desired file in the trunk and then use the checkout feature.
Navigate to the local working directory and open up the solution file (i.e. FLL_NXT_Navigation.sln)
Use Visual C++ to modify code, resources, windows, recompile, etc.
Close the solution in Visual C++ (to avoid file locking conflicts)
Navigate to the parent of the local working directory, notices that TortiseSVN displays a green check if nothing was changed in either a file or the entire folder, or a red mark if a file was changed somewhere in the folder.
Assuming that the local working directory has a red mark on its folder icon, right mouse click on the folder icon and select the TortiseSVN “SVN Commit …” menu item.
Fill in the message log prompt (always fill in the message log prompt!) and press ok, making sure the SVN includes all the files, sometimes it auto-ignores the exeutables and compile related files.
Notify the development team if your changes are candidates for merging into the main trunk.

New Post: Running the FLL NXT Navigation 12-27-2008 Release

weststar42 — Sat, 27 Dec 2008 20:31:27 GMT

There are probably two main ways to get FLL NXT Navigation (FNN) working:

Not recently tested but it worked before on multiple Vista machines
1) Download and install the Microsoft Visual C++ 2008 Redistributable Package (x86)
3) Download current FNN release
3) Extract all the zip files or only the Release directory and files
4) Navigate to the Release directory
5) Click on and run FLL_NXT_Navigation_12-27-08.exe (requires the NXT_picture.jpg file to be in the same folder)

Recently tested on a brand new Windows XP installation
1) Download Visual C++ 2008 Express Edition (not required to run)
2) Download current FNN release
3) Extract all the zip files
4) Navigate to the Release directory
5) Click on and run FLL_NXT_Navigation_12-27-08.exe (requires the NXT_picture.jpg file to be in the same folder)

Future versions will try to eliminate the dependency on the redistributable package. Suggestions on how to configure Visual C++ to do this are welcome.

New Post: What is FLL NXT Navigation?

weststar42 — Sat, 27 Dec 2008 19:52:26 GMT

FLL NXT Navigation (FNN) is a Visual C++ program created by SilentSTRKR to improve how programs are written for the First Lego League international robotics competition. The FLL competiton uses the NXT Lego Mindstorms robot kits. The program was created and is maintained in Microsoft's completely free Visual C++ Express Edition. SilentSTRKR is a Middle-School student, who over the years got tired of trying to figure out and fine tune, mostly by trial and error, the navigation portions of the roboty programs. After taking a C++ course, he realized there could be a better way.

FNN displays a scaled image of the competition map on which you can use the mouse pointer to click and enter the desired robot path (via way points). FNN knows the map's scale in pixels per inch and the distance the robot's wheels move in inches per rotation (10 inches/rotation for the big Lego wheels) to calculate the number of rotations to enter into the NXT programming motor move blocks. FNN can also calculated the number of wheel rotations for a given desired turn angle when the robot uses a zero radius turn (one wheel moves forward while the other wheel moves equally backward) and for a given wheel base.

When FNN is calibrated for a specific robot, using FNN significantly decreases the time necessary to develop new and complex navigation programs for the FLL robots. SilentSTRKR's robot team was able to get very accurate and complex programs (more than four legs and three turns) drawn up, programmed in the NXT programming language, and successfully tested on the robot in less than five minutes. Previous years, it would have taken over two hours and up to five hours to achieve the same accurancy without using FNN.

Typically, the robotics team manual creates a NXT program using motor move blocks programmed with the values calcuated by FNN. This can be done very fast and has the advantage of easily adding motor blocks for an arm motor, adjusting speeds, customizing braking and coasting.

FNN also has the ability to download a navigation file to the NXT controller that can be read by the read_test_motor.rbt program for NXT. Read_test_motor.rbt can read any number of line segments and turns and provides for rapid navigation prototyping. Steps for this are: 1) create robot path using FNN, 2) save navigation file, 3) download navigation file to NXT, 4) run read_test_motor.rbt in the NXT, 5) step back and be amazed at how well the robot performs.