Re: package dependency as applied math (was: Re: tape deity)
q26g.fsfjgh@dhh.gt.org> <[🔎] 87zpik3b30.fsf@tiamat.datasync.com> <87wwdonnif.fsfjgh@
dhh.gt.org>
Mime-Version: 1.0 (generated by tm-edit 7.106)
Content-Type: text/plain; charset=US-ASCII
From: Manoj Srivastava <srivasta@datasync.com>
Date: 21 Mar 1998 18:49:04 -0600
In-Reply-To: john@dhh.gt.org's message of "21 Mar 1998 08:10:00 -0600"
Message-ID: <87soob8s8v.fsf@tiamat.datasync.com>
Lines: 47
X-Mailer: Gnus v5.6.3/Emacs 19.34
Hi,
>>"John" == John Hasler <john@dhh.gt.org> writes:
>> The issue is more complex than simple graph theory algorithms;...
John> But perhaps not more complex than sophisticated ones.
John> This is exactly what I mean by the general dependency problem.
John> A rigorous analysis might be valuable without producing a
John> complete closed form solution.
John> A related approach is to build models.
All I shall say is that a number of people (Myself, the author
of dftp, and the Deity team) have been looking at this issue for
several months now. I hope that my analysis of the problem was
rigourous enough, but hey, anyone can make mistakes. Do you have
specific weaknesses in our efforts in mind?
Jason has a simulator built for this effort. Several
variations of ordering algorithms have been tested in sim runs; and
we have pored over the results and compared them to manually fine
tuned actions. But this is a fair to middling complexity problem; and
we could use extra pairs of eyes.
In my opinion, we have not ignored the mathematical
ramifications of dpendencies as graphs, but it has been half a decade
since I looked at the state of the art in graph algorithms. Doubtless
other members of the team have been in closer touch, though ;-) (I
also think that topological sorting may be a relative dead area
recently; this is not routing algorithms, after all, but I could be
wrong, and there may be cutting edge link ordering algorithms that I
am unaware of).
However, if these efforts are found lacking, we shall be
greatly appreciative of keener minds analizing the problem, or
presenting us with models that lend us insight we have been missing.
*Any* help is gratefully accepted.
manoj
--
People who are excessively concerned about the environment invariably
turn out to own a great deal of land. There are damn few unemployed
and renters in the ecology movement. -- Frank Mankiewicz
Manoj Srivastava <srivasta@acm.org> <http://www.datasync.com/%7Esrivasta/>
Key C7261095 fingerprint = CB D9 F4 12 68 07 E4 05 CC 2D 27 12 1D F5 E8 6E
--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Reply to: