Download CrazyBot for free. Light IRC bot coded in Python with all the features. CrazyBot is a small simple Python coded IRC bot that has many different features and is by far easier to use than most irc bots. License: LGPLv2. Macbot: So I've been designing Macbot, and I can't wait to actually build this thing. The "brain" will be a mac mini that is set up to automatically log in and connect to a computer-generated, password-protected network. It will be controlled thro.
Posted: 2016-11-07
Command Line Emulator for the Bot Framework - interacting with ocrbot on Shakespeare
UPDATE: The Command Line Emulator has been replaced with a full-fledged emulator which is cross-platform (info and install here).
tl;dr: I built a simple OCR bot using the Bot Framework (BF) from Microsoft and the Cognitive Services Computer Vision API. I chose the Node.js Bot Builder SDK from BF. I tested with the BF's unix-compatible emulator (black box above). It was nice and straightforward the whole way through. All of the instructions are here, but feel free to keep reading.
There's really almost too much to say about chat bots: definitions, bot builders, history (more), warnings and cautionary tales (more), guidelines, delightful stories, sad stories, etc. For our purposes, other than a few considerations, I'll sum it up with: they've been around for decades and will be for many more.
However, let's cover some vocabulary to start.
A blog post here details the different types of chat bots, progressing from simple to ones that can hold down a conversation. These are the bot types discussed and an example to illustrate:
- Notifier - sends a one-way message e.g. ping me with today's weather forecast at the start of the day ('push' bot i.e. bot initiates)
- Reactor - replies when I send a message, but does not remember anything e.g. send me the weather forecast when I ask for it ('pull bot' i.e. I initiate), but don't remember me or what I ask for
- Responder - replies and remembers me and my message history e.g. send me today's weather forecast, use my user name on this channel, and remember what cities I choose
- Conversationalist - replies, remembers me and my message history, knows what service I'm on, if there are others there, and when I come and go e.g. send me today's weather forecast, use my user name on this channel, remember what cities I choose, format it nicely for this channel, and if the conversation is old, archive it and send as email
Bot builders lower the activation barrier for developing bots and the MS Bot Framework (BF) Bot Builder SDKs give us a wealth of methods for building dialog and user prompts, making the creation of effective waterfalls really easy. Along with the SDKs, the BF provides free emulator tools, a Windows-compatible desktop app and a Mac OS X / Linux-compatible console app (more information on emulators here).
I know you've been waiting to dive into the code, so let's begin...
There are two choices on bot builder functions for testing locally. We can use the ConsoleConnector which simply and directly allows us to run our Node.js code on the command line. Using the bot builder SDK our code is pretty concise (see more examples on the Core Concepts page for the BF here):
Interacting with ocrbot could look like:
It's simple, but we don't get to see the actual JSON that gets passed to the bot and the JSON passed back. If we want to be able to see the message and also write code that can be used for production later, the Bot Framework Emulator is the way to go. Note, this is the beginning of my server.js Node.js file - see my ocrbot github repo for the complete project and code and the included lab file for more instructions on doing this at home. We replace ConsoleConnector with ChatConnector, for a full deployment-compatible setup, as follows:
One of the major reasons I used github to host this project is that (and outlined in a later blog, TBD) it afforded me the ability to do a continuous deployment directly from the repo. Any change I push up, immediately reflects in my bot on whichever channel I'm on - it was actually pretty astonishing to see a quick typo fix show up so immediately when chatting to my bot at the same time as pushing up the change. But I'll save this for the deployment article to come.
I've always been a fan of command terminals, so even on a Windows machine I'd probably choose to download and use the BF Command Terminal Emulator (download instructions are here). Honestly, I enjoy the simplicity and control a command prompt affords. And now I can develop bots in the exact same way agnostic of OS.
So, I decided to create a bot I could give image links to and if there was text in the picture it would reply with that text. Pretty simple and straightforward since I knew about the free subscriptions to Microsoft Cognitive Services and in particular the Vision APIs. I went to the Cognitive Services main page and clicked on My Account and signed up (or I could have signed up this way with MS, github, or LinkedIn accounts). After that I had a secret key for the service. Now to splice that into my bot.
So, I borrowed much of my code from Samuele Resca's blog post (excellent blog, btw). I placed these helper methods in with the server.js code above:
Then I set up a config file as shown in the repository containing my Computer Vision API key (as well as placeholders for the future app id and app password I get after the deployment step - in a followup article TBD). So, don't worry about the app id and app password for now.
If you are doing this at home, replace the process.env.VISION_API_KEY with a string holding your key (this is a separate file I called configuration.js), see instructions in this lab file.
Then back in my server.js file with the main code I set up my vision OCR functions to read the image text (basically call out the the Computer Vision service with an image url) and then process the text:
Finally, we tie it all together in a user dialog:
We could have added a waterfall here to say confirm they want to process this image or perhaps added a built-in prompt that allows uploading an image attachment. Perhaps in future releases.
Now we test.
Mac Os Download
On Windows one would double click on the BFEmulator.exe directly and on Mac we use the Mono framework software to run BFEmulator.exe from the command line. Pretty easy, peasy.
Let's take this image:
The emulator experience will look a little something like:
Command Line Emulator for the Bot Framework - interacting with ocrbot locally on stats
In conclusion, to get started now, check out the github repo and start playing with the code and emulator. For more instructions check out this lab. Chat later!
Maybe you’ve always wanted to make C-3PO or R2-D2, or wondered whether you could create a Roomba Robotic Floorvac yourself. Whatever your motivation, the Lego MindStorms Robotics Invention System has everything you need in order to build robots. So you put down $200, take the kit home, and build something brilliant—only to discover that the software is Windows only.
Can’t you use a Mac to program robots? Of course you can! Several excellent Lego robot-programming tools exist for the Mac, many of them free. I’ll show you how to get started with one of the most popular tools, MacNQC. (For others, see “Programming Choices.”)
About Lego Robots
The Lego MindStorms Robotics Invention System (RIS) kit contains more than 700 Lego bricks, including motors, sensors, gears, wheels, and a robot brain called the RCX. You can write programs for the RCX on your computer and then download the programs to your robot via an infrared link. The kit includes an infrared USB tower that you plug into your computer so it can talk to the RCX.
So what can you do with all this stuff? Whatever you want. Imaginative builders from around the world have built robot bugs, pets, tanks, plant waterers, lawnmowers, chess players, and just about everything else. You won’t be making C-3PO or the Jetsons’ Rosie, but you’re going to have fun and learn a lot. In fact, many teachers now use Lego robots to teach programming to beginners of all ages.
Meet MacNQC
Although there’s Windows-only software in the regular MindStorms box, the clever and dedicated Lego robotics community, composed of enthusiastic fans worldwide, has created a wide variety of alternative RCX programming environments.
NQC (Not Quite C), originally developed by Dave Baum, compiles text source files and can download the compiled programs to the RCX. In its purest form, NQC can be a little uncomfortable for people who don’t live on a command line. Fortunately, there’s MacNQC, an excellent GUI version of NQC that runs in OS 9 and OS X. At this writing, the current version of MacNQC X (the version for OS X) was 3.0 r2.
MacNQC Quick Start
The first thing you need to do to start programming robots with MacNQC is install firmware on the RCX. The RCX is a very tiny computer. Like a Mac without system software, or Frankenstein without a brain, the RCX can’t do much by itself. The firmware serves as a kind of operating system, enabling you to download and run your own programs.
MacNQC does not come with its own firmware; you will use the firmware that Lego supplies with the RIS kit. The easiest way to get the firmware file is to download the RCX 2.0 Beta SDK. Click on the link at the bottom of the page to download the SDK. (You’ll need to click through one or two more pages to get to it.) Once you’ve downloaded and unzipped the SDK, copy the firm0328.lgo file into the Firmware directory of your MacNQC folder.
If you don’t have an Internet connection handy, you can get the firmware file from the RIS disk, but you’ll need to install the RIS software on a Windows computer first. Then you can copy the firmware file from Windows to your Mac and install it on the RCX using MacNQC. On the Windows computer, you can find the firmware file here: Program FilesLEGO MINDSTORMSRIS 2.0scriptFirmware firm0328.lgo. Copy it to your Mac and place it in the Firmware directory of your MacNQC folder.
At this point, you’re ready to plug in the infrared tower and start up MacNQC. Turn on the RCX and place it near the infrared tower. Choose RCX: Download Firmware from the MacNQC menu. MacNQC will find the firmware file you just copied and install it on the RCX. Be patient; this takes a couple of minutes. With the firmware installed, you’re ready to start playing.
Choose RCX: Motor And Sensor Panel from MacNQC’s menu. This window allows you to control motors (outputs A, B, and C) and view sensor readings (inputs 1, 2, and 3). For example, you can see the current reading of a light or temperature sensor. To turn on a motor, click on one of the green arrows for forward or reverse, and then click on the Start button. To view a sensor value, first choose the sensor type from the pop-up menu and then click on Refresh to get the sensor values from the RCX.
Ready Your Robots!
Programming in NQC is simple. Let’s start with a rudimentary example: a program that makes a robot move forward for one second and then stop. If you haven’t built a robot yet, make a basic one, such as my own RoboTag or Trusty.
I’ll assume you have a robot that uses motors A and C to move. All the program has to do is turn on the motors, wait one second, then turn them off.
In MacNQC, create a new file and enter this:
The commands in this program are all straightforward. You just need to remember that the Wait command accepts a time measured in hundredths of a second, so
means “wait for one second.”You can save the file if you want, but it’s not necessary for testing. Make sure that your RCX is turned on and that the infrared tower is pointed at it; then click on the toolbar button that looks like 1s and 0s. MacNQC will download your program to the RCX. (Your robot will chime when it has received instructions.) To try the program out, press the Run button on the RCX.
If you’d rather just check your program for errors without downloading it, click on the red check-mark button. MacNQC will tell you about any errors it finds in a separate error window. Double-click on an error to go to that spot in your program.
The next example moves the robot forward until it bumps into something. You’ll need a robot with a bumper on the front, connected to input 1. (The bumper should press a touch sensor on input 1 when the robot bumps into something.) Also, motor A should control the left side of the robot; motor C, the right side. RoboTag will work for this project. After a bump, the robot will back up, turn right, and then drive forward again.
Crazybots Mac Os 11
The
line <>A<> (as seen in the screenshot) tells the RCX that a touch sensor is attached to input 1. Next, the program moves the robot forward. If the bumper ever touches anything (that’s when equals ), the robot backs up for one second, turns for one second, and then starts moving forward again.When you’re ready to do more, check out the comprehensive documentation for the NQC language, and information about MacNQC, available from the Help menu.
Welcome to a New World
By now you should be pretty pleased with yourself. You have successfully used your Mac to program Lego robots, and it didn’t cost you anything. (You can watch RoboTag navigate obstacles using the second program outlined in this article.)
You are now part of an amazing global community whose nexus is LUGNET, the Lego Users Group Network. You’ll find other enthusiastic, brilliant people who like to build Lego robots. Browse or search the archives for project ideas or information about troubleshooting, advanced programming, and building techniques. Or you can join (for free) and participate in the discussions.
Crazy Bots Mac Os 11
Now go forth and program robots with your Mac. Have fun!
[ Jonathan Knudsen was the author of one of the first Lego MindStorms books in 1999. By day, he writes about Java technology. ]
Less Geeky You can’t use the Windows-only software that comes in the MindStorms box, but you can use Lego’s Robolab software for OS 9. (It also runs in Classic.) In this visual environment, snap together little program blocks, such as
Crazybots Mac Os X
or . When you’re finished, click on a button to download the program to your robot for testing. This is a good choice for younger robot builders or those with little computer experience. Buy it separately at the Pitsco Lego Dacta store for $69.Crazy Bots Mac Os Catalina
Also, don’t miss the cool kits sold at the Pitsco site, such as the $114 Robolab Intelligent House Building Set, which you can use to create a working model of an automated home.
Geekier If you’re familiar with Java, try the free leJOS, an environment based on the Java programming language.
Use MacNQC’s Motor and Sensor Panel to control robots directly, or create programs in a color-coded text editor. See RoboTag navigate obstacles using the second program outlined in this article.