From: | "Clark, Joel" <jclark(at)lendingtree(dot)com> |
---|---|
To: | "'Randy Jonasz'" <rjonasz(at)click2net(dot)com> |
Cc: | pgsql-interfaces(at)postgresql(dot)org |
Subject: | RE: Re: FW: New PostgreSQL Project |
Date: | 2000-10-05 13:23:03 |
Message-ID: | 69F195289743D411B428009027E293C401290E26@CLTEXCH1 |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-interfaces |
> this is great! Should we take the next few weeks in
> preparing what we'd
> like to see in a C++ API? Plus I need a little time to become more
> familiar with current code base. As a note I was wondering
> what would be
> the advantages/disadvantages of *not* using the c interface
> but writing
> the connection/parse routines directly in the C++ classes?
> i.e. get away
> from having just wrapper classes. On the plus side, I think
> we gain in
> flexibility in design and implementation. On the negative
> side, we may be
> re-inventing the wheel, although I am partial to proceeding
> in a complete
> rewrite.
The only reason I would be hesitant to write our own parsing/network
routines is that changes to the PG protocol would have to be duplicated in
our code. If we wrap libpq, it will be version independent (as long as the
libpq interface doesn't change). Plus the fact that we would be re-writing
code that is already quite stable (and useful, for that matter).
Joel
From | Date | Subject | |
---|---|---|---|
Next Message | Erny | 2000-10-05 17:18:21 | Driver or Postgres doesn't report refer. integr. errors |
Previous Message | Gabriel Lopez | 2000-10-05 11:36:31 | Re: wrong documentation and others ..... |