Re: BOOLEAN types, was: Quick feature question

On 24-May-06, at 4:26 PM, Mike Rylander wrote:

> On 5/24/06, Toby Thain <toby@xxxxxxxxxxxxx> wrote:
> > On 24-May-06, at 3:59 PM, Mike Rylander wrote:
> >
> > > On 5/24/06, Toby Thain <toby@xxxxxxxxxxxxx> wrote:
> > >>
> > >> On 24-May-06, at 1:39 PM, Mike Rylander wrote:
> > >>
> > > [snip]
> >
> > Hi Mike,
> >
> >
> > >
> > >> > I'd be happy if there was an attr that the driver could set on
> > >> fields ... It would be up to the application to look for the  
> > boolean
> > >> > attribute on fields, but at least it wouldn't break any existing
> > >> > applications, and would allow for future apps to make use of the
> > >> > information when it is available.
> > >>
> > >> You're right, an attribute rather than a new type would avoid
> > >> breaking existing applications and give you (PostgreSQL users) the
> > >> extra information you need. It would be meaningless for most  
> > drivers
> > >> though :)
> > > ...  Would there be any objection to me
> > > creating a patch to implement somthing close to the attr plan I  
> > laid
> > > out above?
> > >...
> >
> >
> > I don't have any objection, but it may make more sense as a flag (not
> > an attribute), necessitating a "dbi_result_field_is_bool", because
> > attributes are defined per-type (and for backwards compatibility and
> > portability, we can't assume its type). Maybe Markus has an opinion
> > here.
> Ahh... ok.  I haven't looked at flags on fields.  I assume that
> dbi_result_field_is_null() investigates these flags, and doesn't
> simply look at the NULLness of the C value?
That's right.

> > ...
> > >
> > >>
> > >> Markus and I have considered a new family of functions that  
> > retrieve
> > >> values less strictly. Perhaps this could be the first in that  
> > family.
> > >>
> > >
> > > I wouldn't fight that, but I probably won't use them for this  
> > project.
> > > I need as much information about the data type as I can get. :)
> >
> > The flag gives you the information you need. The function gives me
> > the portability I need. :-)
> I certainly does.  I'll look in the archives for the
> dbi_result_coerce_* discussion.
It's quite recent & brief. I proposed something like this on 4 May.  
Markus is open to the idea, but I haven't offered much more than the  
sketch above in terms of a proposal.
Now that the TEXT typing bug is fixed in the MySQL driver, my current  
needs are satisfied (as long as I use CHAR(1) instead of BOOLEAN :)
--Toby




-------------------------------------------------------
All the advantages of Linux Managed Hosting--Without the Cost and Risk!
Fully trained technicians. The highest number of Red Hat certifications in
the hosting industry. Fanatical Support. Click to learn more
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=107521&bid=248729&dat=121642