Jim et al,

> Leyne, Sean wrote:
>
> >
> >
> >>Probably I would introduce new data type BINARY(n) of maximum size
> >>32000 bytes content of which would be stored on the database page
> >>(that would be VARCHAR type without all character information, like
> >>charset, collation, etc.) whose content is sent over the wire

together

> >>with the content of the rest of the fields. People that need objects
> >>without upper bound should use BLOBs that in this case would be

always

> >>stored separately.
> >>
> >>
> This is a truly dreadful idea. Absolutely retrograde.
>
> Computers should know how to best organize their innards. Introducing

a

> crock like this is simultaneously admitting failure while make success
> impossible.

Just to keep the record straight... *I* did not propose the BINARY data
type. (Jim's quoting/cliping would suggest that I did)

My posting read:

I don't have a problem with a new datatype, but I'm not so sure this
will solve anything -- in fact depending on the page size and the BINARY
size I could see a case where disk I/O would *significantly* increase
(degrading performance).

The big negative for a new type is the inevitable user confusion -- what
is the difference between BLOB and BINARY???


Sean