[cap-talk] OS juncture papers

Jed Donnelley jed at nersc.gov
Thu Feb 28 18:02:36 EST 2008


Cap-talk, cap-conf,

I'll mention these papers to both lists:

ftp://reports.stanford.edu/pub/cstr/reports/cs/tr/83/945/CS-TR-83-945.pdf
"Perseus: Retrospective on a Portable Operating System"

and:

ftp://reports.stanford.edu/pub/cstr/reports/csl/tr/82/229/CSL-TR-82-229.pdf
"The SUN Workstation Architecture"

I found the first interesting because it seems to me that
the Perseus system qualifies as a 'capability' system in
the Demos/Thoth lineage.

I found the second (referenced from the first) particularly
interesting because to me it cleanly illustrated why there
would be no more systems like Perseus developed when it
discussed the software environment for the new Stanford
University Network Workstation.  That software environment
was essentially Berkeley Unix.  Here you had nearly the
last of the OS developments from that era being abandoned
not because of its architecture (though being sold as
"portable" certainly didn't help) but because it didn't
have the application stack available on Unix.  Some of the
project leaders (e.g. Forest Baskett) were pushing both
efforts, but I'm guessing having what later became the
GNU applications available was irresistible.

Reading recently about Van Jacobsen's optimized network
stack, e.g.:

http://www.lemis.com/grog/Documentation/vj/lca06vj.pdf

got me to wondering just how much OS dependence there is
in glibc.

I wonder if somebody could share some thoughts or pointers
to discussion about application support on other capability
systems and it's relationship to GNU or more OS "independent"
software development?

What is the application stack like for KeyKOS, EROS,
and CapROS (Coyotos?)?  Is there any mechanism for
sharing infrastructure like editors, compilers,
simple utilities like cat, more, ls, etc., etc.?
Is there any GNU software running on capability
systems?

--Jed  http://www.webstart.com/jed/



More information about the cap-talk mailing list