Hi Stephen, On Fr., 27. Dez. 2019 at 18:09, Stephen Finucane <step...@that.guru> wrote:
> > I was going to ask why you did patch relations rather than series > relations, but I'm guessing PaStA works on a patch level? No issues > with that, but it would be nice to reuse the relationship information > to link series somehow. I've no idea how that would happen though, so > we can probably think about that later. > Pasta works on relating patches to each other. Historically, it was intended to relate patches and to identify their evolution in out-of-tree developments. So, these developments never had the notion of proper series, but only of versions of sets of patches. Ralf Ramsauer et al. then recognized that this could be used on mailing lists as well, relating patches on mailing lists. I agree that computing relations on series would be really nice. We already recorded the task as future work here: https://github.com/lfd/PaStA/issues/34 Given the possible solution space to compute the relationship, i.e., which factors to account and how to weigh them, this task is probably a bachelor thesis or even a master thesis work, including a proper evaluation. If you know anyone interested, let us know. We will look for students interested in working on this topic in 2020. I hope we can see all of Mete's patches included in the final 2.2.0. We want to make the current heuristics from Pasta useful to others, and the integration into patchwork is an important step for that. Lukas >
_______________________________________________ Patchwork mailing list Patchwork@lists.ozlabs.org https://lists.ozlabs.org/listinfo/patchwork
