주의사항
해당 문서는 중요 변경 작업 중에 있습니다. 일부 문서에 텍스트가 빠져있거나, 불어로 플레이스홀더(placeholder) 자리차지만 되어 있을 수 있습니다.
The current version of Reeborg’s World still has room for improvement. Feel free to report any suggestions for improvements at https://github.com/aroberge/reeborg/issues
input() typically cannot be used successfully with programs in Reeborg’s World. The reason for this is rather technical.
Javascript is single-threaded with no equivalent to Python’s time.sleep() function. In order to implement control over the speed of the animation (including pausing) as the robot performes a task, the strategy used in Reeborg’s World is to run the entire program at once, recording each new world state with additional information (including whether or not a pause has been requested, and if the desired animation speed has been changed). After the program execution has been completed, these “recording frames” are then played back at the requested speed using Javascript setTimeout.
input() is implemented using Javascript prompt() which is a modal widget that interrupts the main Javascript thread. Any call to input() occurs during the program execution, thus before the animation has started. So, if one wanted to prompt the user for a choice, say after the Nth step, using input(), the user would have no way to know what the state of the world would be at that point since only the initial state of the world would be shown at that point.
Skulpt, an incomplete version of Python 2, apparently uses something called suspensions to allow pausing the program’s interpretation. (https://github.com/skulpt/skulpt/search?utf8=%E2%9C%93&q=suspension) If Brython were to implement a similar approach, using input() effectively in robot programs could become a possibility.
When highlighting is enabled/requested, the program run is not simply that written by the user but includes some additional information so that the line about to be executed can be highlighted in the code editor. This additional information is inserted by a simple program (https://github.com/aroberge/reeborg/blob/master/src/libraries/brython/Lib/site-packages/highlight.py) which only perform a superficial analysis of the user’s program. This information is used during the playback (https://github.com/aroberge/reeborg/blob/master/src/js/recorder.js#L109).
In the absence of a robust highlighting procedure, it may very well happen that requesting code highlighting will result in the insertion of lines of code that will raise Exceptions when the Python program is executed.
This highlighting is only available for Python programs.
Given the design decision that has been made to not save any information from users on the site and not requiring any login, coupled with the absense of a clear and well-defined tutorial, no gamification is currently possible. http://en.wikipedia.org/wiki/Gamification
Given enough, well-designed tasks readily available, it would be possible to implement a per browser based approach (using either cookies or local storage) for gamification. Alternatively, one could have a way to save the current “accomplishment status” (encoded in some way as to help prevent “cheating”)) in a file, and load such file so that the status could be transported to update the status in another browser.
Ideally, there should be a simple single-file (zipped) containing a all relevant files that could be used to make a copy of the entire site on a school’s website: this could be required by some schools which limit access to a small number of external sites. However, I have not received any such request yet.
Currently, if a Python program fails with some types of errors like SyntaxError, IndentationError, or NameError, some limited analysis is performed in an attempt to identify some common mistakes (e.g. missing colons, or missing parentheses). This could be expanded to cover more types of errors and identify more potential problems.
It has been suggested that if a program runs successfully, but a given task has not been accomplished successfully, that a program be analyzed to look at common errors (e.g. writing if object_here:, which would be treated as being always True instead of if object_here():) and report if any such errors are identified.