🚀 go-pugleaf

ward@cfa.harvard.EDU (Steve Ward) writes:

>In response to the recently posted lamentation of silence in this
>newsgroup here is a mild reminder that some good things are underway
>but need the support of everyone.  A repost of the 32K software
>archive site info is enclosed.
>
>In fact the response to a call for contributions of software to these
>archives was excellent, resulting in many pledges.  Understandably,
>all volunteering contributors deferred their actual contributions until
>after the holidays, many of them taking vacations and undertaking
>travels domestic and foreign.  I trust you all enjoyed yourselves and
>have returned with recharged batteries!

Having just returned from the holidays...

I have just sent to the archives four shar files which contain Unix style
code development tools I wrote for the Series 32000.  The tools are
equivalent to Unix's ar, ld, nm, ar, and ranlib.  I use them to generate
code for the 32016 system whose design I posted to the net about a month
ago.  So far I have used them to generate a nearly complete Minix port and
a ROM debugger with disassembler.

I also have a GNU C compiler (gcc) which is compatible with the above tools.
However, I am not sure how to distribute it.  GNU requests that you
distribute ALL of gcc if you distribute any of it.  Since I only wrote
a few kbytes of changes to the backend, I sure do not want to distribute
all umpteen megabytes of gcc.  Since my version of gcc currently generates
code for just one machine, I doubt GNU wants to add my  version to their
distribution.  Furthermore, my backend is presumably slightly dependent
on the gcc version.  Hence, even if I did distribute just the backend,
others might have trouble obtaining the possibly obsolete gcc version
that it needs.  Any ideas?

Bruce Culbertson
culberts@hplabs.hp.com