Script errors

PGvdVeen shared this question 4 years ago
Answered

I like to use scripts. In the past, GG was forgiving when a script line calls a variable which is not present. (Which happens quickly when you copy and paste scripts)


But now, GG 5.0233, a script call to a not_longer_present variable results in a serious problem.

If the script is an update script, executed f.e. by a corner[] command, GG reports a script error continuously. Sometimes I even have to kill the entire process to stop GG.

I believe that the response on an error should not be worse than the error itself....

I prefer that GG just ignores calls to nonexisting variables.

Comments (4)

photo
1

I was forced to repair many applets with scripts calling nonexiting variables.


Still many more to do. One single repait in GG would do the job.

Please implement that GG ignores in scripts calls to nonexisting variables.

photo
1

Sorry about the extra work, but it is better (for most users) that we have fixed this bug

photo
1

Sorry, but you have not fixed a bug but just implemented one.


When an update script contains an error by calling a non-existing object, it generates a continous stream of error messages because it is an update script. This makes it difficult correct the error.

An error message should never generate a bigger problem than the error itself.

This is basic requirement for any system.

When GG would just ignores these calls there will be a lack of functionality, not an addition of problems.

photo
1

after starting to use UpdateConstructie[] (dutch) instead of redefining all used random variables I diminished the number of script lines to do in the script and in the mean time the risk of using lines with non-existing variables when copying existing scripts. It won't solve all the problems of editing your files, but it can help.

© 2020 International GeoGebra Institute