On 8/11/2025 3:45 AM, david.k...@libertysurf.fr wrote:
Just to feel at ease with a round number or the compiler having crossed a REAL
milestone with a complete C standard support ?
Then which one should it be ?
If I understand your question, you see two options:
1. Bump tcc to version number to 1.0 now, arbitrarily, for no technical
reason
2. Wait for tcc to reach C11, then set version number to 1.0
The argument for option 2 is it seems "a REAL milestone". The argument
against is nobody has volunteered, or is being paid, to implement tcc
C11 support. Could take forever.
The argument for option 1 is it would help tcc escape the doldrums of
the public not taking tcc seriously because the version number says not
to. The argument against is might seem deceptive, if there is no
difference between 0.9.28rc and 1.0.
If as you suggest, we only have this binary multiple choice, and you
want to know what I would choose if up to me, I would choose the option
that embraces change. Not the option that embraces endless waiting,
hoping somebody will do something someday.
Fabrice Bellard didn't specify which version of ANSI C when he defined
the purpose of tcc as, "The Smallest ANSI C compiler". If anyone wants
to propose tcc is already good enough to be version 1.0, that would be
ok for tcc at ANSI C99.
Widening our view, reality is not constrained to only two options, of
arbitrarily declaring v1 or endlessly waiting. There's a third way.
Improve tcc to be able to compile the latest Linux kernel version,
currently Linux 6.16 released 27 July 2025.
I propose that building the latest Linux kernel with tcc, instead of
with gcc, be made the milestone to define tcc version 1.0. How does that
sound?
Is there anyone excited for building kernels, who would like to
volunteer to start this work? I am offering to assist.
>> Who is the build master to track this?
>
> Dunno.
If tcc had a build master, I suppose you would know.
Does anyone want to volunteer to be tcc build master? Would be doing
github CI/CD, including building cross-platform, tracking bug-fix
progress and defining sprint goals.
>> Who is building software with tcc on ARM Cortex-M?
>
> No me, but since there is support for ARM targets, anyone could
theoretically...
Would be interesting to find out who are tcc users.
>> I merely asked what is needed to bring tcc to version 1.0.
> Again, what the version bump for ?
Wouldn't everyone like tcc to progress beyond 0.9?
>> serious use until 1.0
> Then tell me what should we put into "serious use" ?
I would like to better understand everyone's goals before offering an
opinion what "we" should do.
Robin
_______________________________________________
Tinycc-devel mailing list
Tinycc-devel@nongnu.org
https://lists.nongnu.org/mailman/listinfo/tinycc-devel
