The report of the first week can be found here.
The report of the second week can be found here.
In this week I attended to the RustFest in Paris and I spent some time to discuss ideas with the amazing rust community. As part of this I contributed to some project in the community.
I come up with a new design for cargo-ebuild. The big new deal is to drop cargo as dependency this involves increase significantly compilation time and binary size beyond that depend on cargo-metadata output is more stable in time than cargo version. This also reduced the complexity of the code and boilerplate (particularly for more features to come). In order to get the work done I contributed to ¢argo_metadata.
In this period the work was a bit slower because, as agreed, I spend some
time on my university exams. Besides that the discussion around
cargo-ebuild has starved a bit because different point of view of the
The work on
cargo-ebuild has been completely rewritten with a new design.
Another approach has been adopted on
cargo-ebuild in agreement with the
maintainer in order to integrate more features.
The work on multi-target support in gentoo has started with a first draft submitted to the gentoo-rust.
Keep working on gentoo to get full support in rust and cargo eclass. The work involved also rust and cargo ebuild and slotting system.
Because the work on portage is almost finished I spend some time to cleanup
the code and add the support into
eselect-rust. Now features-as-use-flags
is working correctly with the patched version of
cargo-ebuild and cargo
- add support to cargo binaries
The work done until now has been submitted to gentoo tree and in developers mailing list. Thanks to reviews and discussions some fix has been integrated in the patch.
Final discussions and fix around the gentoo main patch. I also summarized my experiences in a write-up about gentoo eclass.