tinycc-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Tinycc-devel] [OT] Re: 486-tcc-linux-musl - principle issues with x86


From: gz8cx4
Subject: [Tinycc-devel] [OT] Re: 486-tcc-linux-musl - principle issues with x86
Date: Mon, 2 Dec 2024 15:24:22 +0100

On Mon, Dec 02, 2024 at 10:42:21AM +0000, Michael Ackermann via Tinycc-devel 
wrote:
> Several messages were flagged off-topic.

Hello Michael,

Yes, I marked my messages with [OT] and suggested leaving the
"bootstrapping" discussion off-list.

This one will be my last message on this matter here.

> it is bootstrappable.org who solved those problems, not me.

You happen to praise a project who misrepresents its goals and
achievements[1] moreover in the area of my interest and competence.

That's why I find this not only offtopic here but regrettably also
disturbing.

Would you be so friendly to limit the discussion to the (highly
appreciated) thorough testing of tinycc?

Kind regards,
/tccm

[1] my comments (on the off-topic matter) are below

> On 2024-11-26 10:47, gz8cx4@xxxx wrote:
> > bootstrappable.org ... insists on relying on
> > a binary seed, occasionally pretending that its hex-codes is a source :)

> bootstrappable.org initial
> seed is <256 bytes in size with commented hex/opcodes.

> They don't pretend to providing source,
> they actually do.

It is a commented binary seed, not a "source".
Symbolical opcodes and comments are not programming statements
there, they are only "descriptions of beliefs".

Can you mentally emulate a modern CPU and reliably interpret the meaning
of all the _bits_ represented by the hex text?

But more importantly, a seed has also to trust other, much larger binaries
to be used to put it into RAM (possibly modifying the seed behind the
scenes during this operation?) and to let the CPU execute them (or
possibly other codes instead?).

In other words, this kind of bootstrapping just _shifts_ the provenance
problem out of the scope of the activity, at the same time proclaiming it
"solved".

Of course, the provenance _can_ be confirmed by "multiple and diverse
bootstrapping" - but then the result will be equivalent to a much less
expensive multiple diverse cross-building. Binary seeds and hex codes
serve actually no purpose.

> aggi



reply via email to

[Prev in Thread] Current Thread [Next in Thread]