RagingGoblin

Many thanks, this should work indeed. I will try this out tommorrow. The reason for storing the errors inside the shift register is that I want to do a lot more and some processes will go wrong if there is any failure (others will work allright). Some of my subvis will be inside a case structure (error does nothing, no error the process starts). Others will run anyway regardless if there are failures or not. It works fine!!

This is the code I am working on. A state machine with 2 states. Both of them do exactly the same: 1 Delete all files from a certain directory 2 Run an FTP script to fetch new files into this directory 3 Checks whether the filenames meet certain string patterns. If something goes wrong, the third process will generate a false and the same case should be run again. In the end I have to do this for 6 machines. Because the code is exactly the same I want to put this inside a loop, but then if number 3 goes wrong I have to FTP all over again. That would be most inconvenient. Download File:post-4151-1153228842.zip

Don't know if this helps because I do not really understand the issue. But I looks to me that you can simply put this code inside a state machine. (many discussions on this are on this forum). http://zone.ni.com/devzone/conceptd.nsf/we...6256c4e00600196 An event structure would do the trick too I think.

I want to do an ftp session to a remote machine, getting a number of files, move on to the next machine... The code for doing this will be the same for every machine, so naturally I will put this inside a loop because the only thing changed is the ftp script I call. If anything goes wrong I want to restart this ftp session and try again. But I don

Put everything into the second box of a sequence structure and a while loop into the first. Your run button can stop the first box and you will move to you app. I found some other solutions on: http://forums.ni.com/ni/board/message?boar...ssage.id=122684