RSA Secret-Key Challenge Termination

On Mon, May 21, 2007 at 08:53:35AM -0400, Greg Lobring
wrote:
> Personally I don't foresee myself continuing to crunch
RC5-72 keys with the
> discontinuation of the challenge by RSA. In fact, I've
already disabled the
> RC5-72 project on all my clients, they are doing only
OGR work now. As it
> stands with yesterday's stats, it could be well over
1,000 years before
> RC5-72 is done. Even with progressing technology, I
don't see that
> diminishing too much in my lifetime. I've put over
1,630 days (over 4 years)
> into the project, and seeing it at only 0.415% complete
is a little
> depressing!

Something worth mentioning... we have been working on cores
for both PS3
and Nvidia (yes, as in video cards) that obtain keyrates
pushing
200Mkeys/sec. To put that in perspective, 200Mk/s would put
you at the
#66 spot in yesterdays stats. It would take less than 700 of
these
clients to *double* what the rest of the entire network did
yesterday.

To answer the next logical question... no, we don't have
optimized OGR
cores for those devices yet, and it's not clear (to me at
least) what it
would take to make that happen. An advantage to RC5 is that
it's fairly
easy to translate into assembly or similar low-level
languages to gain
huge performance benefits. OGR is substantially more complex
which makes
this much harder to do.

Personally, I think it would be a shame to officially close
RC5-72
without at least taking these new cores for a spin.
-- 
Jim C. Nasby, Database Architect            decibeldistributed.net
Give your computer some brain candy! www.distributed.net
Team #1828

Windows: "Where do you want to go today?"
Linux: "Where do you want to go tomorrow?"
FreeBSD: "Are you guys coming, or what?"
_______________________________________________
rc5 mailing list
rc5lists.distributed.net
htt
p://lists.distributed.net/mailman/listinfo/rc5

On Mon, May 21, 2007 at 08:40:30AM -0700, Kevin McCoy
wrote:
> 
> I'd like to echo Greg's comments. I stopped doing RC5
long ago and
> took up OGR instead since the RC5 project time frame is
a bit
> absurd.
> I do think we will lose a lot of participants that were
interested
> in the prize money or "sticking it to the
man", so the next new
> project should be chosen well.
> So for now, lets focus on OGR and whatever new project
comes next.

There's something I should mention about "what comes
next"... One of the
biggest drivers for that will be people stepping up to bring
new
projects to distributed.net. All of the client code that's
needed is
available for download, and we'll happily work with folks on
the
required back-end changes. There are some folks that are
discussing
various new projects, but nothing will move that process
along like
someone willing to actually get some code running.
-- 
Jim C. Nasby, Database Architect            decibeldistributed.net
Give your computer some brain candy! www.distributed.net
Team #1828

Windows: "Where do you want to go today?"
Linux: "Where do you want to go tomorrow?"
FreeBSD: "Are you guys coming, or what?"
_______________________________________________
rc5 mailing list
rc5lists.distributed.net
htt
p://lists.distributed.net/mailman/listinfo/rc5

On Mon, May 21, 2007 at 08:40:30AM -0700, Kevin McCoy
wrote:
>    I'd like to echo Greg's comments. I stopped doing
RC5 long ago and took up
>    OGR instead since the RC5 project time frame is a
bit absurd.

For a while now I've viewed RC5-72 as a great way to show
off how much CPU you can waste. This has its admirable and
despicable aspects, but its worth realizing that some of us
will give you our CPU cycles as long as there are stats to
prove that we did it.

>    I do think we will lose a lot of participants that
were interested in the
>    prize money or "sticking it to the man",
so the next new project should be
>    chosen well.

I think we can come up with some way to chew up CPU cycles
that is useful and sexy.  But if you have to choose one or
the other, pick sexy and we can always convince a few of the
accolytes that OGR and other worthwhile causes is a better
way to waste their CPU cycles.

>    So for now, lets focus on OGR and whatever new
project comes next.

I appreciate that d.net is a volunteer-only organization and
it is awesome what has been accomplished in the spare time
of so few, but I hope that this transition encourages
everyone to focus on creating alternative projects and
creating the stats infrastructure to keep track of it all.

Thanks again and keep on crunching.

-- 
</chris>

"The problem with troubleshooting is that trouble
shoots back!"
_______________________________________________
rc5 mailing list
rc5lists.distributed.net
htt
p://lists.distributed.net/mailman/listinfo/rc5

I don't see a lot of point running RC5-72 any more either as
I don't
believe it will advance knowledge etc.  I suppose you could
keep it
running so those who are running platforms like HP-UX that
don't have a client
v2.9008 or higher and thus can't participate in OGRp2 can
still have
stats to show off and maybe you can test and learn from the
nvidia and
PS3 cores.  It costs little to keep it going I guess. I
expect a
mass exodus from RC5-72 so if you can find the time and
manpower I would
try to get another project started before OGRp2 ends. 

_______________________________________________
rc5 mailing list
rc5lists.distributed.net
htt
p://lists.distributed.net/mailman/listinfo/rc5

Jim C. Nasby wrote:
> On Mon, May 21, 2007 at 08:53:35AM -0400, Greg Lobring
wrote:
>   
>> Personally I don't foresee myself continuing to
crunch RC5-72 keys with the
>> discontinuation of the challenge by RSA. In fact,
I've already disabled the
>> RC5-72 project on all my clients, they are doing
only OGR work now. As it
>> stands with yesterday's stats, it could be well
over 1,000 years before
>> RC5-72 is done. Even with progressing technology, I
don't see that
>> diminishing too much in my lifetime. I've put over
1,630 days (over 4 years)
>> into the project, and seeing it at only 0.415%
complete is a little
>> depressing!
>>     
>
>
> Personally, I think it would be a shame to officially
close RC5-72
> without at least taking these new cores for a spin.
>   
Absolutely! Even if only so that the developers of the core
could see 
the fruits of their labor, and see "what could have
been".   Another 
question - would we be able to confirm if the key was found,
or can only 
the RSA confirm this?

-- 
[   Joseph Kaye   ]


_______________________________________________
rc5 mailing list
rc5lists.distributed.net
htt
p://lists.distributed.net/mailman/listinfo/rc5

On 5/21/07, Joseph Kaye <jkayeisd.net> wrote:
> Jim C. Nasby wrote:
> > On Mon, May 21, 2007 at 08:53:35AM -0400, Greg
Lobring wrote:
> >
> >> Personally I don't foresee myself continuing
to crunch RC5-72 keys with the
> >> discontinuation of the challenge by RSA. In
fact, I've already disabled the
> >> RC5-72 project on all my clients, they are
doing only OGR work now. As it
> >> stands with yesterday's stats, it could be
well over 1,000 years before
> >> RC5-72 is done. Even with progressing
technology, I don't see that
> >> diminishing too much in my lifetime. I've put
over 1,630 days (over 4 years)
> >> into the project, and seeing it at only 0.415%
complete is a little
> >> depressing!
> >>
> >
> >
> > Personally, I think it would be a shame to
officially close RC5-72
> > without at least taking these new cores for a
spin.
> >
> Absolutely! Even if only so that the developers of the
core could see
> the fruits of their labor, and see "what could
have been".   Another
> question - would we be able to confirm if the key was
found, or can only
> the RSA confirm this?

Once the key is found, anyone can decrypt the ciphertext
with the key
and if the plaintext starts with ``The unknown message
is:'', we have found
the key. RSA's role is only to verify that we have found the
key and give us
the prize money.
_______________________________________________
rc5 mailing list
rc5lists.distributed.net
htt
p://lists.distributed.net/mailman/listinfo/rc5

At 18:21 -0600 on 05/21/2007, Open Phugu wrote about Re:
[RC5] RSA 
Secret-Key Challenge Termination:

>Once the key is found, anyone can decrypt the ciphertext
with the key
>and if the plaintext starts with ``The unknown message
is:'', we have found
>the key. RSA's role is only to verify that we have found
the key and give us
>the prize money.

At 20:23 -0400 on 05/21/2007, Scott Dodson wrote about Re:
[RC5] RSA 
Secret-Key Challenge Termination:

>Only they'd be able to confirm it for sure, however,
assuming that
>there's only one key that produces coherent verbage it
could be assumed
>we've found it.


Given the way the RC5 cypher hashes, I think the string
"The unknown 
message is:" is longer than the chunk size so that the
first part of 
the string is hashed/encoded separately from the the next
section 
(which would include the start of the actual payload). Thus
if you 
recover a message that begins "The unknown message
is:", I think that 
this would be a unique result that could not be the result
of using 
any other key. I MAY be wrong but that is my impression. A
backup 
check would be the actually read with the Mark-1 Eyeball (as
opposed 
to a computer program) the complete text to see that it is
coherent.

_______________________________________________
rc5 mailing list
rc5lists.distributed.net
htt
p://lists.distributed.net/mailman/listinfo/rc5