[ntp:questions] Re: NT binaries for ntp-4.2.0

David L. Mills mills at udel.edu
Thu Oct 23 04:19:37 UTC 2003


David,

The issue is recovering old documentation about old software versions,
right? However bad that documentation may be, we don't get to change it
now. If I take your point seriously, modern shrinkware groupies will
find no reason to extract that old stuff anyway, so why are we debating
it?

You misunderstand the purpose of the documentation from here. It is
intended for reference, record and archive purposes and is probably too
abstract for shrinkware groupies. I describe how the dang things works
in engineering terms, reveal the options and features and the principles
behind the design. If I didn't do that, who else would? There are other
projects, such as the faq and twitchy designed to bring comfort from
this babble and I cheer that.

Meanwhile, don't lecture me on how things have changed in the last ten
years. I have been after all a university professor for thirty years.
While your description of shrinkware groupies may apply to some, it does
not in any way apply to serious researchers here or elsewhere.

Dave

David Woolley wrote:
> 
> In article <3F95BF47.2F1D551A at udel.edu>, David L. Mills <mills at udel.edu> wrote:
> 
> > Geeze I get tired of being the boogeyman. Once upon a time before you
> > came around we used to keep old versions intact with code and
> > documentation in the archive. If folks needed documentation from an old
> 
> Once upon a time may have been ten years ago.  The people for whom
> documentation is an issue nowadays would run a mile before going near
> source code and many can't even install a binary from a zipped archive
> any longer; they need a self executing installer, for Windows, or an
> RPM, for Linux.  Even on SCO Unix, they probably would be put off by
> a source archive. They are home users and glorified office managers,
> not traditional system programmers.
> 
> They also very rarely read reference documentation; if there are any
> forks in the document (a hypertext decision tree) don't expect them to
> backtrack.  The documentation needs to flow in one direction only and
> be prescriptive, not descriptive.



More information about the questions mailing list