Re: hurd init, translator upgrade
> I know that RMS doesn't want binaries to be stripped, but currently this is
> common praxis in Debian and will reduce the load of my ppp link. But
> stripping can't be done without thinking. Is the following okay:
I think all the issues for stripping are the same as on Linux or elsewhere.
> 1) All binaries and shared libraries --strip-unneeded.
Should be fine. I've used --strip-all for binaries and had no problem.
There shouldn't be any difference between the two in fact, because strip
never affects the dynamic symbol table and that's all that should matter.
> 2) All static libraries --strip-unneeded.
I've only used --strip-debug myself, but I presume --strip-unneeded is fine
too if it is appropriate on Linux.
> 3) The dynamic linker will NOT be stripped at all.
I always stripped the dynamic linker (--strip-all); I know of no reason not
to, just like any other shared library.
> (at least this is linux praxis). Will translators suffer from stripping? I
> think not, at least I haven't encountered problems so far.
Translators are just normal executables, no special rules apply to them
aside from who usually runs them.