sbcl-1.5.8 released, SBCL20 news

Dear all,

I have released sbcl-1.5.8, and uploaded source and a binary for
x86-64/linux to SourceForge.  The binary is explicitly *not* built with
support for older Linuxes (those with glibc earlier than 2.28): while
previously the only cost to supporting earlier releases was a slightly
suboptimal memcpy(), glibc 2.28 fixes a bug in fcntl().  SBCL is still
supported on older Linux platforms, but users on those platforms should
build using a suitably older toolchain and installation.  Other
significant changes in sbcl-1.5.8 are appended to this message.

I am also pleased to announce that as well as the contribution by Robert
Smith from Rigetti Computing, the SBCL20 workshop (8-10 December,
Vienna, Austria) will feature a talk by Doug Katzman provisionally
titled "SBCL: playing nice with Unix, at last", where he will discuss
signals, memory management, and the work he has been doing to make SBCL
look like a 'normal' Unix executable.  Registration is still open, but
we will dealing with local organization issues where numbers will begin
to matter: please sign up at <https://forms.gle/FfHF8W8ptyPfMJir8> if
you haven't already to let us know that you are coming.

Best wishes,

Christophe

changes in sbcl-1.5.8 relative to sbcl-1.5.7:
  * platform support:
    ** support for Mac OS X Catalina
    ** improvements in interoperability with C code compiled with memory
       sanitization options
    ** libsbcl.so links to zlib when sb-core-compression is enabled.
       (lp#1845763, thanks to Juan M. Bello-Rivas)
    ** workarounds for BSD issues around mmap() and pthread_attr_setstack().
       (lp#1845936)
    ** support SSE for bzero if available on OpenBSD
  * optimizations:
    ** improved type understanding and translations for division operators
       (including TRUNCATE, GCD, LCM).
    ** sequential comparisons and branches can elide intermediate comparisons.
       (lp#1847284)
    ** convert EQUAL and EQUALP to EQL if either of the arguments is a type
       for which the structural equality predicate is identical to the simpler
       equality check.  (lp#1848583)
    ** internal operators implementing string comparisons produce
       simpler-to-consume values.  (lp#1848776)
  * bug fix: inspecting adjustable arrays and vectors with fill pointers works
    better.  (lp#1846191, reported by Yves Pagani)