This identifier can be used to abort the execution of the command
accociated with that identifier. Note that it is generally not
necessary to obtain a new id for each command, because commands are
executed synchronously. So the id for a command is valid again after a
command previously associated with this id has finished. In fact you
only should get one identifier per thread and assign it to every
command executed by that thread.
Also, after a call to XCommandProcessor::abort, an
identifier should not be used any longer (and instead be released by a
call to XCommandProcessor2::releaseCommandIdentifier),
because it may well abort all further calls to
XCommandProcessor::execute.
To avoid ever-increasing resource consumption, the identifier
should be released via
XCommandProcessor2::releaseCommandIdentifier
when it is no longer used.