> -----Original Message-----
> From: Frank Ingermann [mailto:frank@...]
> Sent: Martes 16 de Abril de 2002 4:49
> To: IBDI@yahoogroups.com
>
> Otoh there are operations (like re-ordering the fields inside a
> table) that
> afaikt can only be done by directly writing the RDB$'s and for that far
> i'm just glad to have a tool that lets me do that.

Maybe in IB5. But in IB6 the simple alter table/alter column allows one to
change a field's position.

> And with all due respect, i
> don't think
> it is very constructive for the whole community when the only reaction to
> this is to focus on one single function that is not implemented the way
> it is supposed to.

Sir, I work for Firebird, not for IBExpert, IBAdmin, IBWorkbench, QuickDesk,
etc. It's not about a single function. It may be the only one that fails,
granted. It's about a mindset of taking shortcuts and go straight to system
tables when there are alternatives and IBE is only an example. The "most
expert tool" should use normal DDL while possible, recognizing the engine's
version (due to missing functionality in IB5) to see if direct manipulation
is really needed. I'm not saying that it's trivial.

What happens when a tool takes shortcuts: if properly done, nothing.
Otherwise, the engine is charged with possible bug reports that don't exist
until the user says after 10 replies for example "no, I didn't use DDL, but
a wonderful tool X that offered to change the index's ASC/DESC ordering on
the fly" (I don't know about a real tool that offers this functionality).
And unfortunately, this is waste of time for volunteers trying to figure
what's going on, until you realize the engine's methods were bypassed. That
happened to Craig Stuntz and me and has happened another times. It's not a
gratuitous attack on IBExpert and my observation about general metadata
modification is not focused directly on IBExpert either.

> bottom line: instead of picking on bugs (that, as discussed
> extensively in
> the last days, every software has... but they are usually fixed very fast
> in IBE) i think it is appropriate to say a big THANK YOU to the authors
> of IBE for their decision.

As I wrote, I have no issues with IBE or another tool. However, releasing a
free lightweight version of a commercial tool is a marketing decision above
anything. Altruism doesn't happen too often or for nothing in commercial
endeavours. If it's useful for all people, welcome. Maybe it will cure the
woes of several people. Maybe it will become the de-facto free tool, who
knows. But if something causes a lot of phantom bug reports (whatever tool
is used, WHATEVER), I'm going to speak again.

And remember that regardless how good a tool may be, not all people will
like it, starting with the interface, following with the license, etc.

C.