[Rpm-maint] [rpm-software-management/rpm] Add rpmtsAddEraseElement2 to API, extend rpmtsAddEraseElement with key data (#1360)

[Panu Matilainen]

> From my point of view as a user of the rpm library, the "key" is a pointer to generic user data

Assuming something does not make it so. 

The key is opaque to rpm of course, but like I said, it's passed around with very specific semantics. Arbitrary user data would be something that has no semantics to rpm.

> But we have prepared transaction with all elements explicitly added. 

Yes, dnf thinks it's in control but it's not, as I've explained many times now. A large part of those added erasures are simply discarded as duplicates as rpm has already implicitly added them. See commit a144c29831a39ed303d6ea8d2ae91e1c36d64c84, this was causing a [severe bug](https://bugzilla.redhat.com/show_bug.cgi?id=1620275) due to dnf not setting the relation information for upgraded packages (it couldn't because that is not exported to Python for another set of complicated reasons)

Sorry but NAK. An incremental addition to a mess only makes a bigger mess. This needs a different solution.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/pull/1360#issuecomment-694803356
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.rpm.org/pipermail/rpm-maint/attachments/20200918/9309f16d/attachment.html>