--- In firebird-support@yahoogroups.com, Helen Borrie <helebor@t...>
wrote:

> At 08:33 PM 14/06/2003 +0000, you wrote:
> >On Saturday 14 June 2003 20:33, you wrote:

> Hyperbole apart, it's "ancient" rather than "awesome". But it

wouldn't be

> nice to acknowledge someone for being ancient, would it?
>
> h.

Hmmmm...

Perhaps society puts too much emphasis on being "youthful?" After
all, you knew the answers that I didn't. And it was your comments
that made me realize what I was missing.

"Absolute Path" wasn't clicking with me (though it bloody well should
have). What irks me about our developers is that they couldn't
communicate directly what this group has communicated (quite
effectively) in an indirect manner, due mainly to my own ignorance
and poor power to raise the question in proper context.

You see, our developers told me that I needed to follow this model in
path specification:
Server_Name:<drive>:\<Path>\Server_File.GDB

What they didn't do was explain the part about the Absolute Path- the
path to the file on the Server itself. You led me there! When I
figured out what I was doing wrong it was absolutely laughable!

I was supplying the path observed on the client, through the network
connection. Of course, the share we had been creating did not have
the same drive letter, or was not the same path via UNC, as the
Absolute Path. It worked in a round about way if, and only if, the
database server file was located on the C drive of the server. If the
database server was located on a different drive, then this
(undocumented and unauthorized) path was unreliable. TeamB guys on
the Borland InterBase General discussion forum even explained that
this method could lead to corruption of the database. Excellent! The
project is way behind schedule, and we're going to corrupt our
customer's databases, too! (not if I can help it...)

I've been around the personal computer world since the introduction
of the Commodore PET-2001, and around the IBM PC in a technical
capacity since the mid-80s. I should have picked up on this a long
time ago, but I kept having a problem with the idea of a connection
path that was not visible on the network. Go figure. I did a little
reading and realized that the InterBase client could talk to the
server without that. In fact, it was designed to do that for security
(among other things).

Anyway, my understanding has been corrected, and I have this group to
thank. It wouldn't make such a big deal out of this, but our
developers can't communicate and this group does so quite
effectively. I've been fighting with this problem for no less than 6
months (which has the decidedly chilling effect of making me feel
like a moron at this point). I have asked our developers repeatedly
for assistance, giving them example after bloody example of what we
were doing and the results we observed. None of them could figure out
what was going on. We picked the wrong developers, you say? Yup, and
that's another story for another time.

Everybody take the rest of the day off... You just earned your pay
for the week! :)

Thanks a million!
-Gary