This reverts commit 988f5a5910.
The release process for many OCaml packages and in extension mirage
related packages usually entails creating a release in the respective
own repository so a release tarball becomes available and then opening a
PR against ocaml/opam-repository to finalize the release. During this
new issues can be discovered which push the release back.
This happened for mirage-tcpip 6.1.0 several times:
https://github.com/ocaml/opam-repository/pull/18357
Prompting in total 3 different 6.1.0 releases with different hashes
respectively (the hash for ocamlPackages.tcpip.src shouldn't be
reproducible anymore, but we probably have cached the tarball already).
Ultimately the PR to opam-repository was closed to investigate some
failures on opam-repository's CI and the release postponed:
https://github.com/ocaml/opam-repository/pull/18357#issuecomment-808434285
I jumped the gun with the release and updated tcpip in nixpkgs before
tcpip was “properly” released in opam. I usually watch the github
repository of package I maintain for releases and can react pretty
quickly to a release as a result. Most of the time I also check
opam-repository's PRs nowadays for extra context or information, but
when everything seems fine and tests succeed I deem the update alright
to PR to nixpkgs. Being faster than opam was achievable in these cases
and actually seems kind of tantalizing.
In the light of this experience however, we should wait for the opam
PR getting merged at least for some packages that exhibit this behavior
of rereleasing the same version number multiple times to get the release
just right (afaik the 6.1.0 tag pointed to three different revisions for
tcpip). To me this is questionable upstream behavior we just have to deal
with in some way.
Contains vendored in patches from https://github.com/mirage/ocaml-freestanding/pull/93
to make building within nixpkgs simpler. These serve as a makeshift
solution until ocaml-freestanding is released with its overhauled dune
based build system which will also support cross compiling. Upstream has
indicated that they are interested in making it work well with nix as
well.
1.7.8 changed the behavior of the minimal build type (which we are
keeping as the default because opam-repository does it as well): It now
excludes the Base64 module which is prone to namespacing problems.
Since google-drive-ocamlfuse still uses the Base64 module, we need to
override it to use extlib without the minimal build type. 1.7.9 (?)
should make this obsolete as it is planned to split the Base64 module
into a separate package.
Co-authored-by: sternenseemann <sternenseemann@systemli.org>
Enabling tests is not feasible as the test suite takes > 45min (haven't
even run it to completion yet) which is not a good idea in ocamlPackages
where you often have to build from source. Additionally it would require
fetching the git repository since the test suite is not contained in the
release tarball.
Co-authored-by: locallycompact <dan.firth@homotopic.tech>
https://github.com/ocamllabs/ocaml-ctypes/blob/0.18.0/CHANGES.md#ctypes-0180
* ocamlPackages.async_ssl: fix compatibility with ctypes 0.18.0 by using
ctypes.foreign instead of ctypes.foreign.threaded since the distinction
between threaded and unthreaded has been removed in this release.
* libbap: link with -thread so linking ctypes.foreign doesn't fail
https://github.com/BinaryAnalysisPlatform/bap-bindings/issues/18
* ocaml-ng.ocamlPackages_4_07.sodium: patch lib_gen/_tags to also add
the `package(bigarray)` directive since `ctypes.stubs` no longer
propgates that, leading to module not found error.
* ocaml-ng.ocamlPackages_4_{04,05,06,07}.async_ssl: mark as broken: due to
the bigarray-compat dependency, we need dune 2 for ctypes which breaks
compilation of the legacy async_ssl 0.11 version since we can't
upgrade to dune 2 for it since that version doesn't support the legacy
jbuild files.
A bit of a pitfall of // is that it doesn't merge recursively which
often leads to unintended deletion in meta sets: If meta is in args it
is also present in the set right of the // operator which means the
right value is used to replace the left value completely. This throws
away anything extra we've set in the meta set in args.
This is fixed by this comment, allowing the descriptions and broken =
true; set in janestreet/old.nix to propagate to the output meta sets.
