Re: Tuple and changes for m68k with -malign-int
On Wed, 21 May 2025, John Klos wrote:
>
> But at the same time I've seen you say you don't need and arent
> interested in these fancy new languages and packages, so I can't see you
> volunteering to do it.
>
> So if you're not going to volunteer yourself, then it's not really your
> place to suggest to others that they should be doing it instead of
> taking the easier route, which is to make more things work by default.
>
I won't volunteer if the plan is to break the ABI so you can convert your
Quadra from C to Rust because you're afraid of NULL pointers. Same goes
for running bytecode in an Openjdk VM running in a QEMU VM.
I am actively disinterested in that kind of trade-off because the ABI
churn would impact both users and upstream developers, many of whom will
receive zero benefit.
Of course, the ABI we have isn't perfect and I'm not against people
designing a better one, that will be stable for another 30 years. Though
that is an ambitious undertaking for an under-resourced project.
For example, I doubt that we would ever have gained a TLS implementation
if had been left up to volunteers. There just aren't enough of us.
Reply to: