response 87 of 92:

Oct 15 00:41 UTC 2006

Joe's right, I do want to change that.  I'm pretty disappointed that
discussion about it has stopped in garage (and I would assume the staff
conference and mailing list, too).  But I also want to update grexsrc and
work on getting a local grexdoc that's mirrorable to remote systems via some
standard protocol (cvsup, cvsync, etc).  Ultimately, a merge of grexdoc and
grexsoft would be nice.  Getting an updated newuser running would be nice; I
found - and fixed - lots of bugs in my private version derived from the
sources running on grex.  Removing unused code and support for absolutely
ancient operating systems (4.2BSD [ca. 1983], Ultrix, AIX version 3,
NeXTstep) and just focusing on POSIX+a simple system dependent library as a
target environment reduced line counts by a third; I figure I can strip out
another 1,000 or so lines by simplifying and correcting some of the logic.
There's at least one faily major security hole lurking in newuser which,
only due to a misconfiguration, isn't exploitable on grex at the moment.  I
think working towards a viable replacement for Picospan is a worthy goal.

Like Cyklone, I think process is important and it would be well to impose
some on grex.

Yes, root access would be required for a lot of this.