Re: A couple of TODO notes

Tom Lane wrote:
> Bruce Momjian <pgman@candle.pha.pa.us> writes:
> > Tom Lane wrote:
> >> * Allow INET subnet tests using non-constants
> >> This should say "Allow ... to be indexed" as it's otherwise a nonissue.
> 
> > New text is:
> >     * Allow INET subnet tests to use indexes
> > Is that right?
> 
> No, because we already index subnet tests against constants.  (It's pretty
> much like the situation for "column LIKE anchored-pattern" --- we
> transform to an indexable range-test.)

Uh, what should be in the TODO?  I am confused.

> >> This is either done or rendered irrelevant, depending on your point of
> >> view, by the V3 protocol.
> 
> > Do we allow SELECT data to come back as BINARY form (not C strings)
> > like we do with BINARY cursors?
> 
> Yes, you can make that happen in the V3 protocol: you can select either
> text or binary output format for each column.  Declaring a cursor BINARY
> is in fact irrelevant in V3 protocol.

How is binary selected?

--  Bruce Momjian                        |  http://candle.pha.pa.us pgman@candle.pha.pa.us               |  (610)
359-1001+  If your life is a hard drive,     |  13 Roberts Road +  Christ can be your backup.        |  Newtown Square,
Pennsylvania19073