[00:03] <RAOF> <mwhudson> "RAOF (he/his): say are you..." <- My nvidia system is somewhat non-standard (it has both an nvidia *and* an AMD card in it), but I can do those tests if no-one else is available.
[00:03] <mwhudson> RAOF: i haven't found anyone else yet :/
[00:26] <mwhudson> hm i wonder why "mk-sbuild --arch i386 jammy" fails
[00:27] <mwhudson> something about eatmydata
[00:29] <mwhudson> aha --skip-eatmydata
[05:38] <alicef> hello, how you ask debhelper to add the ubuntu versioning on the package filename ?
[05:39] <alicef> ~20.04.1+0ubuntu1_amd64
[09:10] <RikMills> doko: thanks for the utils update!
[09:11] <RikMills> thinsg seem to be working
[09:11] <RikMills> ^ xypron
[09:43] <schopin> If I introduce a package to the Ubuntu archive before it hits Debian, will it be possible to sync them afterwards using the regular tooling?
[09:45] <JackFrost> Some people have abused -0build1 to get free autosyncs, but...
[09:46] <schopin> That answers my question, even though I was just thinking of manual syncpackage and MoM :à
[09:46] <schopin> :) *
[09:47] <JackFrost> syncpackage -f thepackage  to force sync over the ubuntu version, just ensure to use the same source tarball.
=== alan_g_ is now known as alan_g
[10:37] <cpaelzer> RikMills: doko_: I can only agree, my formerly failing riscv64 build is happy as well now - thanks
=== gpiccoli_ is now known as gpiccoli
[16:07] <ahasenack> fyi I'm working on stunnel4 blocked in migration (dep8), and turns out it's also an ftbfs nowadays with openssl 3.x
=== dbungert1 is now known as dbungert
[19:27] <ahasenack> schopin: hi (for tomorrow), do you recall a change in openssl 3 regarding treatment of ips vs hostnames in certificate common names? I'm hitting a test failure in stunnel's test code which seems to point that way
[19:27] <ahasenack> Failed to connect to 127.0.0.1:6503: [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: IP address mismatch, certificate is not valid for '127.0.0.1'. (_ssl.c:1129)
[19:27] <ahasenack> the cert is for CN=localhost
[19:29] <ahasenack> hm, actually, it's when I replace ssl.PROTOCOL_TLS (which is deprecated) with ssl.PROTOCOL_TLS_CLIENT, hmmm
[19:30] <ahasenack> in python 3.10:
[19:30] <ahasenack> Changed in version 3.10: The context now uses PROTOCOL_TLS_CLIENT or PROTOCOL_TLS_SERVER protocol instead of generic PROTOCOL_TLS.
[19:31] <ahasenack> """The protocol enables CERT_REQUIRED and check_hostname by default."""
[19:31] <ahasenack> bingo
[19:31] <ahasenack> what is this called again, rubber duck debugging? :)
[19:31] <ahasenack> (check_hostname by default)
[19:48] <bdmurray> ;-)
[20:59] <ahasenack> hmpf, just because I disabled ipv6 locally. I got this in a ppa now:
[20:59] <ahasenack> Failed to connect to ::1:6503: [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: IP address mismatch, certificate is not valid for '::1'. (_ssl.c:997)