Bug in Labview "FTP registry.vi"

[xred]

Hello,

When someone uses the Labview VI "FTP Put File.VI", the FTP connection is in the end closed by the VI "FTP Registry.VI" with parameter "function" set to "unregister".

Unfortunately, in case of an error during the FTP transmission, the VI "FTP Registry.VI" will not do anything. Especially, according to me, it will not unregister the connection and thus will not free the memory used by the connection.

In other words, I think that in the VI "FTP Close Session.VI" the error wire should not be connected to "FTP Registry.VI".

Is it correct?

[Bart De Rouck]

Dear,

After some research, I have to agree with your statement, that this is a minor LabVIEW Bug. There is altought a work around, by implementing your own error checking and freeing the memory manually by calling FTP Registry.VI with the unregister option enabled.

We will altough report back to our LabVIEW developers this inconvenience and make sure they pick it up in the next bugfix release.

With Kindest Regards
Bart De Rouck
Sales Manager NI Belgium

[Dave_Thomson]

Three years later, and I find this same bug is in LabVIEW 2010.

In my case, my custom FTP client polls the connection.  If it finds that the connection has been lost, it closes everything on this end so that the user can then reopen the connection.  Part of this process is to call the FTP Close Session.vi, which issues the QUIT command and then unregisters the session.  However, since the connection has been lost, the QUIT command throws an error and the session does not become unregistered.  The program then cannot re-open a connection to this same address.

This is simple error handling for any Close type of operation.  Errors on the input of close operations should not prevent the operation from completing and freeing any resources.  In this case it is the Unregister that needs to ignore the input error, as discussed in the original two posts.

DaveT