Ok... that is what I just said.
Yes... a project should always follow it's roadmap.
And yes... Musl is not compatible with projects that are not interested in changing their roadmap to it.
I believe Musl is only useful for those interested in using it. And like any project... it has pro's and cons.
This thread is posted using Musl with a Porteus derivative that I have used for years with no problems.
It works for me but for others mileage my vary. So as in all things... I only speak for myself.
Musl's merit ultimately is determined by whether it produces the desired result for the person using it.
When I compile I turn Musl off because it is not compatible with GCC or Tiny C Compiler
In my viewpoint (and not @ the expense of any other)...
Musl's optimal application is @ runtime in a "GNU" (GCC) userland.
Alpine Linux is problematic because it's Musl compilation truncates it's repo.
But this also explains its dominance in the virtual server space w/ Docker containers.
I was in communication w/ Dockers Australian devs years before that project became ubiquitous.
And what I could see early on was a trend toward sandboxes (like... Alpine/Docker) running any OS.
That approach guarantees a limitless repo.
Thus, addressing what Tovalds implied in various interviews... that a system is only as good as the size of it's repo.
And this remains true no matter how good (or bad?) the system is.
So consequentially... proliferation (Ubuntu?) is King.
This why Adelie and Void Linux made the compromise of a
glibc userland running on
Musl instead of compiled with it.
Then you get the best of both worlds.
My post signature indicates that my interest is to share
info to "consider"...
with those who choose
to explore it's merits (or lack thereof).
That pretty much sums up my primary focus.
I am more or less... posting a tutorial on how to do something. In my
understanding I don't see myself as making a request for how to do something.
I personally see hardening as a... tactic.
In the wrong hands... the result will be incompatibility (and vulnerability).
In the right hands... the result will be compatibility (and reduced attack vector).
If hardening is done right... it operates like a daemon in the background.
Which means the only one that should be aware of what is "under the hood" is the... mechanic.
So with that said the primary goal should be that "hardening" be...
non-intrusive to the end user but very punitive to the perpetrator.
I change port 22 to make it appear closed (aka... Security through obscurity
It might not be "better" than the alternative but it could be an alternative in a toolbox of tactics.
So I really don't see Musl as being "better" than glibc... I see it as a tactic.
For me, Musl is just simply... the "right tool for the right job"
(aka... Unix philosophy: Do One Thing and Do It Well
With that said... if Musl is a hammer that does not mean everything is a nail.
Which is why I don't use it to compile... because it won't on a "GNUee" (GCC) system.
I also believe that... Porteus has been an ideal base for the derivative (I made long ago)
that I created for "my" use case
as an alternative to the default.
More or less from my perspective... I don't really see myself as targeting an audience.
I imagine that what I'm doing is... sharing my findings with those that might
find them useful to one degree or another (if @ all... I haven't taken a poll).
When I do look for help (for the last 20yrs) what works best for me, has been
mostly FAQs... blogs and tutorials (or developer lecture videos) and of course
embracing the contributions of the giants who's shoulders we all stand on.
That makes for a long learning curve but... I think it's more proactive than (what I see as) being a beggar seeking to be worthy of pity.
I say this because the years I wasted "panhandling" could have been better spent... analyzing the rules so I could break them (hacking).
those gatekeepers who would dismissively patronize me with
hubris or stroke their snarky egos (while amusing themselves) with my ignorance.
This is what I have ultimately concluded in my post Windoze years spent in the GNU+Linux ecosystem across untold distros.
But in fairness... the exception to that critique have been prolific devs like Gilbert Ashley (src2pkg) and Bill Spitzak ( FLTK
I also forgot to mention "cemi" (curaga) @...
I owe them a debt I can never repay for their mentoring and the selfless
donation of their valuable time in explaining their solutions to me.