I agree. Plocate/mlocate is useful technology, but it's not important
enough to justify maintaining two or three different implementations.
If we can make plocate cover all bases, and it's faster, I'd just make
it *the* implementation. Let's get the agreement of mlocate maintainer
first though. Michal, wdyt?
Hi everyone,
I agree that we should transition to the newer (faster and well maintained) implementation and in the long run we should have only one locate implementation in the distribution. However, I think we should have some transition period (e.g. two Fedora release cycles), during which both packages will be present. Introducing alternatives scaffolding seems like an overkill for something that is going away in a year, hence plain Conflits: tag would be enough. As for maintenance, I think that having an older (mlocate) package available for a year or so while we iron out potential compat issues or bugs shouldn't add too much work.
Zbigniew, please take plocate through the review process and once it is included in the distro I am willing to take up (co)maintenance of the package.
Cheers,
Michal