Rich Salz
I don't think fixing the apps is that important; making the API's that downstream developers used to use hav3e no performance regression is more important.
Is the fix just to add a trailing slash at https://github.com/ietf-tools/datatracker/blob/main/ietf/settings.py#L1010 ?
It would be bad if *one part* of OpenSSL did NULL checks, and the rest of the public API's did not.
>A first stab at fixing this myself noted that data type in struct ossl_param_st should be made const, but this revealed to me all the locations at which the data...
Should the same fix be made for the buttons, at https://github.com/ietf-tools/datatracker/blob/main/ietf/templates/meeting/session_buttons_include.html#L139 ?
Q What other external engines are available? A We know of one for GOST (at xxxxxurlxxxx) and one for based on TPM (xxxxxxx). Others may be listed in our wiki,...
and another one: Look here: https://github.com/MicrochipTech/cryptoauth-openssl-engine Perhaps "We do not know all of the engines that other projects have released. Here is a list of the ones we know:"
> We would need something more generic. I have something on mind, I'll try prepare a PR when I have some free time. It is disappointing that the team does...
So if neither OPT_SELF_TEST_ONLOAD or OPT_SELF_TEST_ONINSTALL is specified, it can do the autodetect if the version string parameter has 3.0 or 3.1
I think if 3.1 requires 3.0 provider to have a 3.0 executable around, that is arguably a bug in 3.1 and you should fix it in 3.1 The project has...