> > Does anyone else have work that they're trying to finish up before
> > we go beta?
> >
>
> Well, I did have a question that got lost in the details of the recent
bytea
> discussion. Specifically it was this: is there a good reason that byteaout
> octal escapes all non-printable characters?
>
> ISTM that if you are using bytea, it's because you need to store and
> manipulate binary data, and therefore it ought to be the client's
> responsibility to do any required escaping of the returned results. In the
> work I'm doing with bytea presently, I either have to unescape it on the
> client side, or use a binary cursor (and some app environments, like PHP,
> don't currently give me the option of a binary cursor). Not that big of a
> deal, but it doesn't seem right, and as you've seen by recent discussions
it
> confuses people.
>
> The only reason I can think of for this behavior is that it makes the
> results displayable in psql -- but again, I'd expect psql to deal with
that,
> not the backend.
>
I guess add pg_dump to that list -- any others? Do you think there is a lot
of existing code?
-- Joe
