upgrades fail after leopard update

[Ryan Schmidt]

On Jun 12, 2008, at 20:53, Alan Batie wrote:

> Cameron Simpson wrote:
>
>> Breaking cycles doesn't turn it into a tree, but fortunately  
>> that's ok;
>> an acyclic digraph is fine - you can traverse it like a tree and just
>> need to avoid rebuilding a package you rebuilt on another path. Which
>> macports seems to do just fine anyway.
>
> If it looks like a tree... ;-)
> You would probably redundantly check a port, but the following  
> times around, it would match as current, so nothing would be done.

We don't have circular dependencies in MacPorts. If we did,  
attempting to install any port involved in that cycle would cause  
port to go into an infinite loop. Believe me, we've (inadvertently,  
of course) had this situation before.


>> As remarked, it would be good to distinguish chosen variants from
>> auto-variants in the port install state information.
>
> I'm not familiar with the variant options (I usually just say "port  
> install something" and if it doesn't work, curse ;-) and manually  
> install it), but yes, I can see where someone may choose to force a  
> non-standard version and that would need to be tracked.

If a port does not work as expected, your first step should not be to  
abandon the port and install the software manually; instead, please  
report a bug in our issue tracker so the bug can be fixed so the port  
works as expected.