> Maybe IBX tries to prepare the update queries and traps the status error
> codes, who knows...
> I remember a previous discussion on this one. The unanswered requirement
> still applies, I think. One wants a user to be able to only insert,
> another to be able to only edit, all without modifying the application
> and without adding code and data structures to track permissions, just
> grant/revoke. Is this such an uncommon situation?

I'm perfectly aware of what the request is. I recall discussing this at
length previously.

The reason I do not want to add this is mostly because of priorities. I have
more important and useful things going right now.

Secondly, there are some negative aspects of a feature like this. There are
a lot of things that can cause the request-live logic to return false. I
always appreciate being able to tell which condition it is so that I can
make it work more quickly. If I make this transparent instead of allowing
the error message to show then I have just added one more thing that will
potentially confuse me. It may be a permissions problem but I may pull my
hair out thinking it is some other reason that my dataset won't allow edits.

I don't think it is any harm to the user to see a simple error message
telling them that they do not have sufficient permissions for certain
operations. Perhaps there is a weird aspect of this if it pops up at an
awkward time. This is will have to look into more.

FWIW,
Jason Wharton
CPS - Mesa AZ
http://www.ibobjects.com

----- Original Message -----
From: "Nando Dessena" <nandod@...>
To: <IBObjects@yahoogroups.com>
Sent: Sunday, March 18, 2001 11:02 AM
Subject: Re: [IBO] LiveMode - read this Jason.


> Jason,
>
> > I don't get how this works. I agree that I could do some metadata
queries to
> > find out what roles people are using and therefore the permissions but I
> > think it is a bit of a can of worms to try and automate this all.
>
> --
> ____
> _/\/ando
>
>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>