Hi Timothy,
On Mon, 6 Dec 2021 at 21:02, Timothy Sample <samp...@ngyro.com> wrote: > <https://ngyro.com/pog-reports/2021-12-06/> Thanks! Really cool! As mentioned in [1], upstream of one disappeared: https://code.divoplade.fr/mkdir-p.git and the package had been removed by 0d2400ceca8d0a0358abaf4cd699e54ddad0e885 two months ago. Hopefully, we still have it in Berlin. guix time-machine --commit=3275c9e1f5 -- build -S guile-mkdir-p However, it returns /gnu/store/dz88dc1k8q7161f3j1m668hi8zna4qcx-guile-mkdir-p-1.0.1.tar.xz So, it needs some work to push it to SWH and get the same checksum 01k20rjcv6p0spmw8ls776aar6bfw0jxw46d2n12w0cb2p79xjv8, IIUC. 1: <https://lists.gnu.org/archive/html/guix-devel/2021-12/msg00032.html> We have to investigate to these 3 ones: R "https://github.com/dalanicolai/djvu3"; R "https://github.com/halostatue/minitar"; R "https://github.com/stardiviner/org-contacts.el"; because the Web front-end of SWH says 'successed' when archiving and for instance, https://archive.softwareheritage.org/browse/origin/directory/?origin_url=https://github.com/dalanicolai/djvu3 but PoG asks for the commit 37b675be1d4d436cdd0c3b5d3f13e88b59a7bf18 which is not in, if I read correctly. Well, the only reference of djvu3 in the Guix repo is commit b61ee34c9d7b679cb772e6e8ff0c0876ccf087b1 which indeed refers to commit 37b675b. However, if upstream is cloned: git clone https://github.com/dalanicolai/djvu3 git -C djvu3 log --oneline | grep 37b returns nothing. Which probably means an upstream in-place rewrite. Arf! Berlin also has it. However, again it requires some work to push to SWH with the expected commit. I have not investigated the two others. Similar cases I guess. Last, maybe I miss something but I do not see 'python-scikit-learn' or 'sway' when this bug#51726 is still pending. Somehow, I have the same question I tried to ask there [3]. 2: <https://issues.guix.gnu.org/51726> 3: <https://lists.gnu.org/archive/html/guix-devel/2021-12/msg00064.html> Cheers, simon