LLRnet server rally March 8th-9th

[mdettweiler]

Originally Posted by kar_bon

ohhh, here're some guys Star Trek fans like me too!?

35 years ago i saw TOS the first time and since then i'm a fan. my favorite is TNG.
(time goes by... so faaaast )

and for the Star Date look here: http://www.hillschmidt.de/gbr/sternenzeit.htm

PS: and NPLB goes where no primesearcher has gone before!!!

Ah I see, we're in a negative stardate like I expected.
-314818.26173851435

Of course, as per the article Mini-Geek suggested, they're pretty much arbitrary in the shows (at least in TOS they were), so it doesn't really mean much. For example, when I punch in today's date in the year 2267 (right smack in the middle of TOS), I still get a negative stardate (-55820.49863647899). In the shows that would probably be around stardate 3000-4000 (somewhere in there).

After looking at the calculator page a little bit more closely, though, I see that the base date is January 1, 2323 (which the website calls Stardate 0). However, that is obviously way off, because the 23rd century would actually be the 2200's. (You'd think the people who made the website would check for a common mistake like that...lol!)

[Mini-Geek]

Originally Posted by Anonymous

Ah I see, we're in a negative stardate like I expected.
-314818.26173851435

Of course, as per the article Mini-Geek suggested, they're pretty much arbitrary in the shows (at least in TOS they were), so it doesn't really mean much. For example, when I punch in today's date in the year 2267 (right smack in the middle of TOS), I still get a negative stardate (-55820.49863647899). In the shows that would probably be around stardate 3000-4000 (somewhere in there).

After looking at the calculator page a little bit more closely, though, I see that the base date is January 1, 2323 (which the website calls Stardate 0). However, that is obviously way off, because the 23rd century would actually be the 2200's. (You'd think the people who made the website would check for a common mistake like that...lol!)

Maybe they're basing it on a TNG definition, while 23rd century/2200s being positive stardate was from TOS only.

Also, from Memory Alpha (like Wikipedia but Star Trek-centered):

Stardates were in use as far back as the 2150s, although Earth had not yet adopted the system. Among the species who did utilize stardates at the time were the Xindi. In 2154, Degra, a Xindi-Primate, sent a coded message to Enterprise containing a stardate for when Enterprise should rendezvous with Degra's ship at specified coordinates, which were also contained within the message. Vulcans also seemed to have an understanding for stardates at that time, as T'Pol was able to calculate that the given stardate was three days away. (ENT: "Damage")

Would they really use negative dates based on some arbitrary point in the future, or has the definition of stardates changed since then?
And:

In the 24th century, approximately one thousand (1,000) stardate units pass for every Earth year, although there are variations in the system at some points in the timeline. (Star Trek: The Next Generation; Star Trek: Deep Space Nine; Star Trek: Voyager)

So 1 stardate unit = 1 milliyear. Also see http://memory-alpha.org/en/wiki/Star...nconsistencies reminding you it's sci-fi, emphasis on the "fi".

I think it's changed within Star Trek things, and not just TOS where it was a tool and afterthought simply to make it "the future".

[mdettweiler]

Originally Posted by Mini-Geek

Maybe they're basing it on a TNG definition, while 23rd century/2200s being positive stardate was from TOS only.

Also, from Memory Alpha (like Wikipedia but Star Trek-centered):Would they really use negative dates based on some arbitrary point in the future, or has the definition of stardates changed since then?
And:So 1 stardate unit = 1 milliyear. Also see http://memory-alpha.org/en/wiki/Star...nconsistencies reminding you it's sci-fi, emphasis on the "fi".

I think it's changed within Star Trek things, and not just TOS where it was a tool and afterthought simply to make it "the future".

Hmm...according to the Wikipedia page you gave a link for, though, even though in TNG and later, stardates were more consistent, they were still later than the ones in TOS (i.e. they didn't start over from a new base date). So, even if it's using the TOS definition of a stardate, a date from the 2260's shouldn't be negative.

[AMDave]

Originally Posted by gd_barnes

... I see that you have the Grommit avatar at the AMD users site, which is listed as your home page. Being the diligent mod that he is, I'm sure he found that and used it.

Actually I thought it so good that I copied it over to there from here (hope he doesn't mind)

I am up to 4 cores for this weekend
2 x 0.4xx per cycle and
2 x 0.9xx per cycle
They are all on port 100 at the moment.
Should I keep them there or pounce on port 5000?

[mdettweiler]

Originally Posted by AMDave

Actually I thought it so good that I copied it over to there from here (hope he doesn't mind)

I am up to 4 cores for this weekend
2 x 0.4xx per cycle and
2 x 0.9xx per cycle
They are all on port 100 at the moment.
Should I keep them there or pounce on port 5000?

I'll have two cores on the rally this time, both from my Core 2 Duo 2.2Ghz. Yeah, I know, I probably should try to come up with a little bit more, though...I'll see what I can come up with.

As to when you should move your cores to port 5000, Dave, it's your call.

[gd_barnes]

Originally Posted by AMDave

Actually I thought it so good that I copied it over to there from here (hope he doesn't mind)

I am up to 4 cores for this weekend
2 x 0.4xx per cycle and
2 x 0.9xx per cycle
They are all on port 100 at the moment.
Should I keep them there or pounce on port 5000?

Pounce on 5000 anytime. My work machines were moved over this afternoon. I'm switching over my home machines shortly. I've had problems trying to switch over shortly before the rally.

Let's see if we can get to n=390-395K by the end of the rally!

[IronBits]

Bring it on! I'll be here during the rally to keep an eye on the Server.
I hope I have enough k/n pairs
Port 500=67394 lines in knpairs.txt
Port 5000=55554 lines knpairs.txt

[AMDave]

K.
I just finished the switch.
now 4 on 5000 for 48 hours (about 3000 pairs I think)
then back to 100.

[mdettweiler]

Originally Posted by IronBits

Bring it on! I'll be here during the rally to keep an eye on the Server.
I hope I have enough k/n pairs
Port 500=67394 lines in knpairs.txt
Port 5000=55554 lines knpairs.txt

Hmm. Why don't we load up the port 5000 server with another 50K or so? It definitely won't do any harm, since even if we didn't run out of k/n pairs during the rally, we'd still need to load it up shortly after (assuming the rally produces as much as we're expecting it to).

[gd_barnes]

Originally Posted by Anonymous

Hmm. Why don't we load up the port 5000 server with another 50K or so? It definitely won't do any harm, since even if we didn't run out of k/n pairs during the rally, we'd still need to load it up shortly after (assuming the rally produces as much as we're expecting it to).

50000+ should be well more than sufficient for the rally without Carlos and Beyond. We tested 28000 the first rally with Carlos. I expect more than that but not substantially more since we've received no report of any other extremely heavy hitters, i.e. 20+ cores or 4+ quads.

If we load another 50000, Carlos will be manually testing n=~430K before we've even gotten to n=400K on the server. I'm attempting to do a little better job of keeping things at least somewhat in n-value order. That's why I've delayed posting manual files until we got close here. I'll do that here shortly since it's clear that n=400K-410K should be OK.


Gary

[mdettweiler]

Originally Posted by gd_barnes

50000+ should be well more than sufficient for the rally without Carlos and Beyond. We tested 28000 the first rally with Carlos. I expect more than that but not substantially more since we've received no report of any other extremely heavy hitters, i.e. 20+ cores or 4+ quads.

If we load another 50000, Carlos will be manually testing n=~430K before we've even gotten to n=400K on the server. I'm attempting to do a little better job of keeping things at least somewhat in n-value order. That's why I've delayed posting manual files until we got close here. I'll do that here shortly since it's clear that n=400K-410K should be OK.


Gary

Ah, I see.