[pacman-dev] -Sp is broken?

[Xavier]

On Mon, Nov 19, 2007 at 06:45:52AM -0600, Dan McGee wrote:
> The best solution would be to factor print URIs operation out of
> needing a transaction at all. This may be helped by me actually
> merging your upgrade patch first.
> 

That patch alone doesn't help much.
But if the idea is alright, maybe it could be extended to sync_prepare, after
find_replacements and sync_sysupgrade :
skip the QUESTIONs when the trans argument is NULL.

The questions in sync_prepare are LOCAL_UPTODATE and CONFLICT_PKG.
If I'm not mistaken, the first one will never be skipped in case of -Su
anyway, and the second one shouldn't come often. So that would not be a
problem.

But I'm looking at it now, and sync_prepare looks much more dependent on
the trans argument than the others.
It also uses trans->packages, trans->targets, trans->flags, EVENT(trans,...).

So that doesn't seem easy. Using a null trans argument looks quite hackish
anyway..

If the current code really doesn't allow to implement features like -Qu and
-Sup correctly, then I think these options should be removed, and the
problematic code rewritten, so that they can be added properly later.