The latest beta have a new 't' switch to percent encode/decode a string, but $encode/$decode already had a 't' switch to force the input to be treated as text.
A different letter must be used instead, 'p' is not used and make sense for 'percent'.

After investigation, the 't' switch in $encode and $decode for 'text' is just a fluke in the help file, those functions weren't supporting a 't' switch in the first place.
So 't' can be kept as the letter.

Note: it would be great to have the new help file in the beta zip when it is updated (currently the case)

Thanks, yes, the 't' switch was never actually supported in the code because it was just the default behaviour. However, since it is documented, scripters may still have used it. For backward compatibility, I will be changing the percent encoding switch to another letter in the next beta.