[openssl-project] A proposal for an updated OpenSSL version scheme (v2)

[Viktor Dukhovni]

> On Sep 21, 2018, at 9:28 PM, Tim Hudson <tjh at cryptsoft.com> wrote:
> 
> That parsing of history I think is at best a stretch and not supportable and also not what our users think.

This isn't a mathematical theorem or even a legal debate.
We should do what makes the most sense going forward.

Richar proposes, and I agree that it makes more sense to leave room for
two nibbles of major number, and that for better or worse applications
expect to see the same top 3 nibbles in a given stable ABI.

The proposal to lose many of the low bits in the ordinal, and lose
the status bits, ... is just not appealing, and there is no API/ABI
reason to retain the verbatim mapping from notional major/minor/micro/...
into the ordinal, this is an ad-hoc encoding with monitonicity as the
the only constraint.

I don't agree that anything about semantic versioning, or our past
practice constrains the encoding beyond that sole constraint.  So
I'm afraid there are no arguments that will convince me that what
appears to be a less flexible encoding is somehow a forced choice.

We'll have to disagree on this one, and see where the consensus
falls.

-- 
	Viktor.