[edited from support email]
We experimented a bit with the responses.xls file and custom functions and principally it works BUT there seems to be a ML error while intercepting or interpreting the data.
As you can see (see attached files) we generated a website with a textarea (named message). The questionnaire has a custom item (the website) with the exact same variable name. Also we have this variable in our excel file. But it seems to us, that the data send by our website isn't reproduced the way it should. Spaces are presented as "+" and newlines as "%0%0%0%0". Because this same error occurs in the excel data file, we suppose it's happening within ML.
Generally, we think that it would be a very neat way of doing things. But participants should be able to actually read what they wrote: Do you have any idea why this happens? We tried different regional settings but that did not affect the outcome... also we formatted the cell as text/string (in excel) but that also did not change a thing.
Your help and hopefully a solution to this problem is as always very much appreciated!