Hello, On Mon 25 Nov 2019 at 04:02PM -05, Daniel Kahn Gillmor wrote:
> On Mon 2019-11-25 07:59:45 -0700, Sean Whitton wrote: >> I'm a bit apprehensive about firing up a copy of gpg-agent. The dgit >> test suite has had a lot of issues over the past few years with properly >> setting up and tearing down an agent for testing. It seems to work for >> now, though. > > If mailscripts runs into problems with gpg-agent, i want to know. > > I saw all the trouble dgit had, but never managed to replicate it > myself, and dgit itself was too convoluted for me to follow what was > going on with the test suite :/ the reproducers that Ian managed to come > up with were something like 10000 independent processes talking to a > single gpg-agent at once, iirc. I don't think that mailscripts will run > into that particular setup (not that it should fail there either, sigh) > > If we can replicate a failure with gpg-agent on the mailscripts test > suite, that should be simple enough that it will be easier to convince > upstream to deal with. Okay, cool. I believe that we are unlikely to see the problem until and unless the mailscripts test suite sets up the temporary GNUPGHOME more than once. > btw, i've been filing other bug reports with some of these dependent > projects about spurious error messages, warnings, failures in obscure > modes, or other noise as i work through building out this test suite. > So having this stuff recorded and automated in the mailscripts test > suite is turning out to be a good thing for the ecosystem generally, > because those bug reports now have a chance of being fixed. Nice :) -- Sean Whitton
signature.asc
Description: PGP signature