On 2021-12-18, Mattia Rizzolo wrote: > Thank you for the patch, that I can surely apply, but I wonder if you had a > chance to have a look at the mass archive rebuild we had that applied this > flag to everything.
That was a related but different flag, CMAKE_SKIP_RPATH=ON, and the rebuild tests just tested weather they FTBFS, not the reproducibility status. http://qa-logs.debian.net/2021/10/25/dcsr/ http://qa-logs.debian.net/2021/10/25/res.dcsr.txt That shows ~1k failures ... which is neither small nor huge. In my limited experience, either can be used to make a package reproducible, though I have a vague memory that possibly CMAKE_BUILD_RPATH_USE_ORIGIN=ON might be less likely to cause FTBFS, though I can't recall the details... > Just so to avoid spreading more boilerplate around... Yeah, I happen to have hit numerous packages today which become fully reproducible with CMAKE_BUILD_RPATH_USE_ORIGIN=ON passed to configure... just the luck of the packages I saw today! To avoid boilerplate, switching to debhelper 14 would fix them (which passes both CMAKE_SKIP_RPATH=ON and CMAKE_BUILD_RPATH_USE_ORIGIN=ON) ... but debhelper 14 is currently experiemental which involves a different set of hoops to jump through to enable it. live well, vagrant
signature.asc
Description: PGP signature
