Hi and thank you very much for your response!
The reason I'm trying to avoid manipulating the string received is that
the system I'm working on is a little more complex (that's of course
very relative, but I'm pretty much a beginner, so it doesn't take much

).
My program is spread across several vi's and it looks something like:
serial write -> do actions -> serial read -> do more actions
with the read and write in different vi's. Now, I know I could just pass the written message down the line, but I'm worried about this approach because it might cause problems when synchronisation between sending a message and receiving the answer is not done right, which a user of my app other than me might do. I am documenting everything, but not everything will be read, I'm afraid.
In the meantime, I did just what you predicted: serial write > wait 5ms > flush buffer > wait some more (do actions) > read and it seems to work. I'm more confident about this approach, because I think it's less confusing seen from the outside. Now all I have to do is make sure it works in all cases. Do you think there are problems that might arise from this?
Sorry I didn't post a sequence diagram, but I'm not at work right now. Oh, by the way! Have a great week-end!