Ported U*ix programs to Plan 9?

Ported U*ix programs to Plan 9?

[rsalz]

Er, rather than porting existing stuff (GNU, etc) shouldn't
you be writing NEW programs with the plan9 model?

Ported U*ix programs to Plan 9?

[forsyth]

>> I think the only place where the "plan 9 model" is dramatically
>> different is in window system software.

i disagree.  9P, precise control of the contents of multiple namespaces,
user-level file servers, and the simplicity of the process model
make a big difference in the way one designs
the implementation of both old and new functions
in the Plan 9 environment.  many of the features i listed appear
in some form in other systems, but usually they are sufficiently
hard to use that people don't bother.  by contrast, in Plan 9
they are used almost casually, and to good effect.

>> I should clarify this:  In Unix, most programs take standard input
>> and produce standard output.  The same is true in Plan 9.  So when

a Plan 9 program can also make things visible in a name space,
to allow them to be manipulated by both new and existing commands.
that possibility can dramatically change the design of a program or
system of programs.

Ported U*ix programs to Plan 9?

[mike]

>Er, rather than porting existing stuff (GNU, etc) shouldn't
>you be writing NEW programs with the plan9 model?

I think the only place where the "plan 9 model" is dramatically
different is in window system software.

There's a large amount of unix software that fits plan 9 just
fine, and it would be awfully stupid to reinvent wheels when
there's genuinely new things to work on instead...

Ported U*ix programs to Plan 9?

[Roger]

> Er, rather than porting existing stuff (GNU, etc) shouldn't
> you be writing NEW programs with the plan9 model?

sometimes it's non-trivial to re-invent _every_ wheel
before using it, however nice it might be... :-)

IMHO, one of the nice things about plan 9 is how easily
most `naive' unix programs port to it.

  rog.

Ported U*ix programs to Plan 9?

[Mike]

The ape library and the associated "pcc" compiler have worked very well
for me for porting a variety of unix programs to plan 9.

The most annoying aspect of the whole process is not porting C code,
but porting makefiles and the associated shell scripts.

To this end, I have ported the almquist /bin/sh from netbsd, and gnu
make.  Using these tools I have been able to configure and build a
variety* of gnu software right out of the box.  I've also ported the
pbmplus suite of bitmap tools, although that required some small
source changes.

In the process I've found a few bugs in the ape library; most notably
the wait() function does not return exit status correctly.  What I
think happened is the kernel recently started putting the string
"<program>:" into the exit status, and the ape library has not been
updated to know this.

I eventually (september or october) will make all this stuff
available to others, but right now my time is limited.

	Mike

* sorry, no emacs. :-)  It can be done--in 1991, out of sheer masochism,
and to annoy rob, I ported gnu emacs to plan 9.  Fortunately, it has
been lost to the mists of time...

Ported U*ix programs to Plan 9?

[Magnus]

Is there someone, somewhere who has ported a U*ix program to Plan 9.
I'm sure there is, so I wonder where I can get more information about
these ports.

We haven't got our system running yet - we'll try on wednesday - so I
don't really know if it's even interesting to port any programs! 

(Wouldn't it be nice to have Emacs on your system... :-)

Actually I had in mind something for managing mail and News.
Impossible?

Regards,
Homann
--
   Magnus Homann	    "They don't call it the net of a
   d0asta@dtek.chalmers.se   million lies for nothing" -- Vinge (?)