Random Numbers?

[drezha]

For some reason my internet connection got cut off last night about 15 minutes after I went to bed. :evil:

And then RC5 started work on random numbers. Now how does it know what random numbers to do? I assume there might be some crossover of work there then if it's totally random.

[Brucifer]

There is a blurb on their site about that. I keep 2 or 3 days work on my systems that I run OGR/RC5-72 on, just so that there is soemthing to crunch should I have an ISP issue.

[NeoGen]

Don't worry drezha, the number of possible (and still untested) RC5 combinations is so large that it can afford to throw some randoms at it with a neglegible probability of duplicating work.

After all, in almost 4 years of activity only 0.356% of that interval has been completed yet.

[drezha]

Cheers.

4 years worth of work? :S
Fairly strong code then...
A BOINC port might help :roll:

[Brucifer]

There's more to life than boinc. Look at all the boinc efforts and their problems, and then look at the overall reliability of distributed.net and they are doing much better. At least in my opinion.

And the problem isn't a lack of participants, it's the massive volume of work that has to be performed to crack the code. As it is there's quite a number of participants that are huge fans of the distributed.net clients.

[drezha]

When I was on about BOINC ports I wasn't being serious. It was a V sign to BOINC. I'm a fan of the dnet client, it was just a bit of a biatch to setup nicely. Was also taking the mick that everything seems to BOINC'ing these days and IMHO there splitting up time that people can use towards one project.

Using lifemapper as an example, at TPR we were big fans of SETI classic and lifemapper. Lifemapper finished and most of those people went to Genome. Now Genome's finished as well.
While I like the idea of BOINC, it doesn't work out as well. IMHO it's causing the time that a project may complete longer to complete Espcially the jumping around the projects like a jack in the box. Yes I know a lot of you do it here. Please don't take it as an insult, it's just my opion. Varity is the spice of life et al but you can have a little to much.

As for BOINC problems, thats because most of them only have 1 server. I guess the upshot to both dnet and FAH is that they have multiple servers in case one is down. In fact I'm suprised Berkley dont for SETI...

And as for the random numbers...I like it...at least it gives the machine something to do if the machine gets disconnected. Even if it's not a untested number, at least it's giving the project some checking ability in the data may have been corruptted etc.

Just my 2 cents.

[Brucifer]

When I was on about BOINC ports I wasn't being serious. It was a V sign to BOINC. I'm a fan of the dnet client, it was just a bit of a biatch to setup nicely. Was also taking the mick that everything seems to BOINC'ing these days and IMHO there splitting up time that people can use towards one project.

Ah so... I understand the jab there.

I agree with you on the boinc thing. Too many of the projects just don't have the infrastructure behind them to have a good reliable project. But there are some good solid boinc projects, too bad that they are in the minority. I did a good bit of crunching on LM too, I really liked that project. Too bad that it went down the tubes.