[rc5] Cyberian Discussion
nugget at slacker.com
Thu Jul 31 18:57:35 EDT 1997
This thread has now gone well beyond the scope of this mailing list and
should be allowed to die of natural causes. Further postings will be
frowned upon by the list management.
Let me take this opportunity to address some of the issues raised, however
and hopefully there will be no further confusion.
1. The Bovine effort has no plans to cooperate with the cyberian effort.
It was their decision to start a competing effort. We honor and respect
that decision, but also recognize that the prospect of duplicating work was
a known liability in choosing not to cooperate with us. In other words,
they chose their road and now they have to travel it.
Our resolve in this decision is bolstered by the fact that at NO TIME, both
before and after the inception of the cyberian effort have their
coordinators made any effort to contact any of the bovine coordinators.
(Despite their claims to the contrary)
Therefore, we can only assume that they have no real desire to cooperate.
We are willing to accept their decision in this matter.
2. Their stats pages aren't very healthy either, although between ours and
theirs, ours is probably "more broken". I do find it somewhat discouraging
that their stats system is breaking under what I would consider to be a
very light load, but that's just the nature of the beast. I still stand by
our decision to place the V2.003 code release above stats repair higher on
our list of priorities. Unfortunatly, both these tasks are reliant on the
same person (who has had limited time to devote to the effort recently.
3. Although my own experiences do not reflect this, it's quite possible
that some people are getting faster checking times with the cyberian
client. I would, however, advise those people who are throwing figures
around to sit with the client and a stopwatch to determine keyrates and not
to rely on the client to report a true rate. I do know for a fact that the
older (and possibly current) cyberian win32 client had a bug which caused
it to report innacurate keyrates. Remember: the cyberian block size is
exactly 1/16th the size of the bovine block size. Time how long the latest
bovine takes to complete a block and then time the cyberian client. It's
the only way to be sure.
Regardless, once you factor in the 10% lossage (this figure climbs 0.25%
each day) that is inherent in the duplicated work with the cyberian client,
I honestly don't think you'll see any substantive differences. Also, our
code is getting faster too. THe new GUI Win32 client shows a marked
improvement on Pentium chips and these optimizations will be present in the
general 2.003 release as well.
4. If you are frustrated by the rivalry and feel that the goal of proving
that 56-bit encryption is weak should transcend any competition between
efforts I would remind you that we at the Bovine effort had NOTHING TO DO
with the split effort that we have right now. We've made every effort to
accomodate people's needs and requests in an attempt to port the client to
every imaginable platform and to improve its features and capabilities. We
would have welcomed the cyberian coders to our own development team
enthusiastically and would love to have gotten their (obviously skilled)
coders involved in our V2 source development. Had they bothered to contact
us before starting a competing effort, perhaps this is what would have
5. The bovine effort did not start at key #1, nor do we distribute the keys
6. I'm sorry if you feel that Project Gutenburg is a poor choice of
charity. I guess it's impossible to expect us all to agree on a worthy
cause with so many to choose from. As I recall, the decision to donate to
Gutenberg was made very early on in the project by an informal group of
people congregated in the IRC channel. I don't recall hearing any
arguments against at the time, but most of the alternates that have been
suggested on this list were discussed at the time.
If you have any further comments or questions regarding the cyberian
effort, I'd simply ask that you move it to a private channel and let the
list drift back on topic. Perhaps it's time to consider creating an
rc5-chat list where we can be a little more relaxed regarding topics.
|David McNett |To ensure privacy and data integrity this message has|
|nugget at slacker.com|been encrypted using dual rounds of ROT-13 encryption|
|Birmingham, AL USA|Please encrypt all important correspondence with PGP!|
To unsubscribe, send email to majordomo at llamas.net with 'unsubscribe rc5' in the body.
More information about the rc5