Re: A burning issue

[Brad Miller <miller@ACORN.CS.ROCHESTER.EDU>]

    Date: Tue, 13 Sep 88 11:27 EDT
    From: JSP@IBM.COM (J. Scott Penberthy)

    I only intended for Symbolics to determine whether a given message
    contains "company secrets" (read: source that they don't want released
    to the general public).  The patches would be offered on an as-is basis,
    where comments can be made via the 1lispm-patches0 mailing list.

The point I was trying to make is that if the two parties already have
source licences, these source secrets are available to them already, and
thus should not require Symbolics blessing to transfer changes to these 
sources between them.

The simple way this system could work is if I have a fix, I can advert it or
send it to another party if I believe him to have a source licence, and cc
my trasmission (or pointer to an FTP directory) to Symbolics so it is an
above the table transaction. If Symbolics feels sufficient care isn't being
taken to protect their work from other parties w/o source licence, then they
can naysay the methodology used. In other words, this is a move from a
system where specific permission must be gained on a per item basis from
Symbolics to one in which blanket permission is assumed, but Symbolics may
make specific objections (e.g. as in the case of putting the Zmail sources
on some net fileserver with a minor change in one file and saying "come and
get it", vs. "I have a patch, send me mail if you have a source licence and
I will send it/tell you where you can get it." 

As for verification of source licencehood, the mailing list idea is a good
one, if maintained by Symbolics, based on release level. You get on the list
iff you have a source licence for the release level discussed. Presumably
earlier releases can be gated into later releases, but not vice-versa.

----
Brad Miller		U. Rochester Comp Sci Dept.
miller@cs.rochester.edu {...allegra!rochester!miller}