> If I understand you correctly:
> The behaviour that I am seeing is the *expected* behavior of Jaybird
> to, kind of, isolate the bug present in Firebird.

Expected behavior for the case when param is used in assignment or
comparison. And probably a bug to certain extent in case of LIKE clause (I
suspect that server does not crash in this case, but one has to disable this
code first to know for sure (see FBWorkaroundStringField)).

The one thingh I could tell wrong is that it might lead to "Operation
cancelled" error and not to a server crash, but there must be a bug report
somewhere telling the exact conditions (and most likely a test case).

> Makes sence, since one can argue that it is better to throw an
> exception from the driver in a case that would crash the
> server....or even better.. fix the error in Firebird :-) I will
> file a report and go and check the source code myself.

Please tell us about your findings.

Roman