I have not found any problems with name space collision using the technique
Chipp described. Locking messages and loading a stack with the same name
into memory seems fine - you just remove it from memory using the file name
(long stack name). I tested it quite a bit and routinely run through 40 or
50 stacks searching for stuff (they often have man duplicates and so
duplicate names and sometimes they are corrupted).

I do it regularly but not that often to claim it is totally bug free - but
would very much like to see if anyone can find an example where it does not
work?

On 13/07/07, Trevor DeVore <[EMAIL PROTECTED]> wrote:

On Jul 13, 2007, at 1:29 AM, Chipp Walters wrote:

> So, what to do if there already is a stack with the same name open?
> Just curious, would this work?
>
> lock messages
> if there is a stack tStackPath then
>  put true into tExists
>  delete stack tStackPath
> else
>  put false into tExists
> end if
> unlock messages
>
> I would think this would work even with namespace collisions, as the
> delete stack explicitly names the stacks filepath.

Chipp,

I haven't tried the above code but using "if there is a stack" does
load a stack into memory so I would think the collision would happen
with that call. What Jerry and I did for Galaxy was write a function
to extract the stack name from disk before trying to open it. That
was the only reliable method we found to circumvent the problem.
But not even that works if there is a substack that will cause a name
collision.

--
Trevor DeVore
Blue Mango Learning Systems
www.bluemangolearning.com    -    www.screensteps.com
[EMAIL PROTECTED]


_______________________________________________
metacard mailing list
[email protected]
http://lists.runrev.com/mailman/listinfo/metacard

_______________________________________________
metacard mailing list
[email protected]
http://lists.runrev.com/mailman/listinfo/metacard

Reply via email to