Larry,
I get around the table unload problem by having a table with the table name,
a data sort sequence and an unload table sequence which I run a cursor
through to get the reload to work. It only takes a little maintenance
whenever a new table is added. There are also columns that indicate what
tables depend the current table and what the current table depends on to
visually help the sequencing.
I get around the views by running the reload twice - as I re-discovered in
January having forgotten why I was getting error messages but the views
created.
Both methods are a little klutzy but they work and I can reconstruct my
databases very quickly. It would, however, be nicer if RBTI could find a way
to make it all automatic. Us "oldies" are aware of the way it works but new
users would probably wonder why their new database may not be complete.
Regards,
Alastair.
PS: I hope you took my double exclamation marks in my reply yesterday as
intended - I whizzed it off rather too quickly. They were meant to signify
surprise rather than a shout. Apologies if you (or anyone) was offended.
----- Original Message -----
From: "Lawrence Lustig" <[EMAIL PROTECTED]>
To: "RBASE-L Mailing List" <[email protected]>
Sent: Thursday, January 24, 2008 7:06 PM
Subject: [RBASE-L] - RE: UNLOAD Command in R:BASE (was Re: How do I get
ridof a DBFTable??)
The biggest enhancement(s) I would like to see in UNLOAD:
1. Transitive closure for PK/FK dependencies when unloading table
structures. I believe it's currently possible for a table with an FK into
another table to be unloaded _before_ the table with the corresponding PK.
Trying to load this structure will fail. R:Base ought to ensure that the
order of tables is reloadable.
2. Transitive closure for view dependencies when unloading views. I
believe it's currently possible for view B that depends on view A to be
unloaded before view A is unloaded. This can happen if you define A, then
B, then drop A and redefine it. Obviously, the resulting file will fail
to create view B when it is run.
3. Currently, stored procedures are not unloaded in any way. I'd like to
see the SYS_PROCEDURES and the other table for procedures (the name
escapes me at the moment) unloaded as well.
--
Larry
--
No virus found in this incoming message.
Checked by AVG Free Edition.
Version: 7.5.516 / Virus Database: 269.19.10/1240 - Release Date:
23/01/2008 17:47
