--- In ib-support@y..., "Woody" <woody.tmw@i...> wrote:

> I have seen too many times in the past where convenience

caused

> future bugs to appear. I tend to think that this "bug" should be

fixed to

> always provide an error if trying to compare two totally different

data

> types.

Woody, experience and professional logic says to me the same. And
personally I never use different types in one expression without
explicit casting, even if tool allow it. But don't forget about
thousands of existing applications, which will be chopped off
possibility to migrate to FB. Some wrong behavior is like a heavy
trunk - difficult to carry but pitifully to throw out. I, for example,
at some point stopped to follow up IBX because it's progress way is
too careless about such a things, I can't revisit all my modules after
each release of the new version. Perhaps the best is leave it as it is
now but emphasise it in documentation. Maybe different behavior in
dialects, as it was made with ambiguity.

Best regards, Alexander V.Nevsky.