David Woodhouse

> You know, if I understand things correctly, the real issue here is that OpenSSL's `apps/apps.c:setup_engine()` calls `ENGINE_set_default()`. … Does that make sense? Completely. > one solution could be to...

> the ENGINE stuff was never written to support the use of several engines at the same time. It's really a one-engine-per-process deal... So... within the same process _(which might...

> It's kinda hackish, but I can see the benefit there. Yeah. And it's fairly close to what we're already doing. I won't say I like it more... but I...

And you'll do that mask, or some other better solution for OpenSSL 1._$NEXT_. Although actually, my plan of record is that I won't _care_ about all this engine nonsense in...

> You're thinking that crypto/pkcs11 would become an alternative to the ENGINE API, right? Note that we already have an alternative to the ENGINE API. Applications can link directly to...

And once it's physically present, _then_ we can look at integrating it properly, and interfaces to make OpenSSL Do The Right Thing, automatically accepting a PKCS#11 URI where it would...

Looks good to me.

> • that your planned hack is impossible for the 1.1.0 series. The result of that, of course, is that I have to re-think my "planned hack", and at _least_...

FBReader on Android can now read AZW3 files. But it doesn't seem to be able to download them from Calibre's OPDS service, as it can other types of file. I...

Apologies for the delayed response. I've updated to 2023.2 on the master branch, will look at backporting just the libpython3.11 fix to stable, once I've got local builds of 23.05...