*Pout*^H^H^H^H^H^H*Grin*

[Engine]

Man, I was just getting all psyched up over OGR, and it's looking like I'll have to go back to RC5 for a while. I'm moving out of my parents place and into my new apartment this weekend. So, I'll be without an internet connection at home for about three weeks (DSL gets installed in the apartment Aug 7!).

Since OGR-24'll probably be over in less than three weeks, I can't really build up a big in-buffer and wait Oh, well, back to RC5 temporarily for me.

 

[ViRGE]

Don't look at it that way, look at it as the fact that you'll have a megaflush for Aug. 7!

 

[imported_Thunder]

Engine,

or... if the prospect of a nice flush on Aug. 7 doesn't sufficiently brighten your day, remember that there's always OGR-25!

Good luck with your move.

-Brian

 

[Joe O]

You could sneakernet.

 

[Engine]

>>remember that there's always OGR-25!

I know If OGR-24 finishes early enough I may go over to my parents place and grab a bunch of OGR-25 stubs from their computer and put them on a floppy to take to my apartment.

>>You could sneakernet.

I _could_, yeah. I could bring a floppy over to my parents place and do a fetch and flush from their computer, but that would be a huge pain. Especially since I'd really have to do it every day or every other day because the ogr-24 stubs are recycled so quickly. I'd be able to do it really easily if the email fetch and flush worked for OGR. Then I could bring a floppy in to work and fetch and flush through email at work everyday.

 

[Joe O]

Engine,
E-mail fetch/flush does work. Eug talked about it in another thread. I'll try to find it and post a link here.

Here's the link: http://forums.anandtech.com/messageview.cfm?catid=39&threadid=199389 Where nexus9 wrote

<<

TO: fetch@distributed.net
Subject: whatever
contest=OGR
numblocks=xxx

where xxx=number of blocks you want to get... >>

 

[Engine]

woohoo!

I remember a .plan update when CSC support was put into the fetch/flush via email, so I assumed that there would be one when OGR support was added. Guess I shouldn't have assumed that

 

[JonB]

The last couple of days of OGR-24 will include lots of recycled stubs. I believe you still get stats credit for completed stubs even if already turned in, though. At least, they did that during the recycle problems in the CSC contest. I only buffer 30 stubs maximum right now. It won't be so bad during OGR-25 since it will be a more controlled start and most participants will be doing OGR by choice, not by accident.

 

[Eug]

Just note that if you flush through email it doesn't get registered on the pproxy stats, if you're using one of those.

(So I'm fetching via email for two computers and flushing through my third one (main one) using the standard method to Mika's.)

 

[nexus9]

Anyone know (approximately) how much longer OGR-25 (26, 27, etc) will take compared to OGR-24?

 

[BoberFett]

Well, I don't know a whole lot about OGR, but I'd guess it will scale like a binary number. 25 will take twice as long as 24. 26 will take twice as long as 25. Repeat as necessary.

 

[Engine]

Actually, I don't think that would be the case. Going from OGR-24 to OGR-25 isn't like increasing by a bit in encryption cracking. This a testing all combinations sort of thing.

It's been over a year and a half since I took any stats in college, but isn't the formula something like (25(25-1)!)/25 for OGR-25, and (24(24-1)!)/24 for OGR-24? So, OGR-25 would be approximately 25 times as big as OGR-24. Likewise, OGR-26 will be 26 times as big as OGR-25. (I'm pretty sure that the formulas there are wrong, but I think I'm right about OGR-25 being 25 times as big as OGR-24)

Of course, if you multiply those formulas out you get like 2E22nodes for OGR-24, which would take a couple thousand years or so. That would be the maximum number of nodes possible, though. Most of those can be ruled out with pruning algorithms and stuff.