How much have your points increased?

[Wrah]

But how would you do it with the console version?

And, with the GUI, can you switch back to FAH after you've checked the box in GAH and still have it work?

Just start the console up with -configonly parameter, when it asks if you want the change the advanced options answer *yes*. It will then ask "Client Type (no-pref/fah/gah)?", answer "gah". Then the next question will be if you want to upload work in batches: "Minimize networking -- get and submit work in batches (no/yes)?". Go yes again and that's it.

In the winclient it's right there:




Note: I noticed that older versions of the winclient do *not* have this option. You have to have the latest version.
(Is this the latest version? The version I got here is 4.0.1.0 dated from 8-dec 2003)


And you can allways switch back to fah. I switch back and forth a lot to avoid stuff like the latest 638-storm. Not that the tinkers you get with gah are that great either.

[dasman]

And you can allways switch back to fah. I switch back and forth a lot to avoid stuff like the latest 638-storm. Not that the tinkers you get with gah are that great either.

That's not really what I meant, but I think seeing the screenshot answered the question. I was wondering if after clicking the box, could you switch back to FAH and still have the batch DLing work?

The screenshot appears it appear that you can't, it is "Get and return G@h work in batches." So I'm guessing that the selection doesn't carry over to FAH

I want to batch DL F@H (give me 8 Dgro's please ). Looks like I'll try UD Monitor over the weekend...

Dave


PS Wrah, do you know of a switch that does the same thing as the checkbox for the console version? I do have an old (pre-SSE) system that could be pressed into service at home, but I'd want to run the console as a service (it's my server and I don't want to keep it logged in to run the client). Right now, that machine isn't folding...

[Wrah]

Nope, that doesn't work. I tried putting it manually in the client.cfg too but it has no effect. Stanford is doing this on purpose. The tinkers you'll recieve when you switch to gah do not have a deadline, meaning stanford is no hurry to get result back of these projects. So they allow you to store a batch. But I wouldn't be surprised if the setting disappeared completly in a next version.
If I found a way to get 10 Dgromacs in the queue I would have PM'ed everybody by now.

Wrah, do you know of a switch that does the same thing as the checkbox for the console version?

um, didn't I describe that at the top of my last post or are we talking about different things?

[dasman]

um, didn't I describe that at the top of my last post or are we talking about different things?

I was so mesmerized by the pretty screen clip that I blew right by that

Dave

[haysdb]

bump

[geordie]

I think you have to be patient, David.

We really need to give it a couple more weeks before any comparisons between old and new will give anything like sensible results. Without getting all statistical, the sample sizes are just too small at the minute so any changes could just as easily be down to natural variance caused by different wu assignment.

Hope you've been keeping a record of weekly points from before the change.
Even more difficult will be finding folders that haven't changed hardware / folding time etc over the last 8 weeks or so.

[haysdb]

Grimm (who once tried to rip me a new orifice over the points issue) and I now AGREE on a new scale, and Jason also agrees:

I think Grimm's scale seems good. I'll test it out a little later tonight to see how it looks.

Here is the scale we agreed on:

Code: Select all

Color  Multiplier  Points
Red        16       1600
Maroon     12       1200    
Brown       8        800
Green       4        400
Blue        2        200
Teal        1        100

This new scale is based on the Stanford formula for assigning WU values, which uses a factor of 110 PPD for their benchmark 2.8GHz P4. This is a pretty stout cpu, more powerful than can be considered "typical", so 100 PPD was proposed as the baseline. 100 is also a nice round number. Rather than just pulling the numbers out of our ass, they are based on the Stanford formula.

David


There is one major wildcard: how plentiful will DGromacs be going forward. This is the one thing that could cause the 24hr averages to change dramatically from day to day, week to week. However, since Vijay has said they will come in spurts, and since the benchmark machine has SSE2 disabled, I think it's reasonable and fair to ignore them in terms of computing the new points cutoffs.

[haysdb]

Here is how the new points cutoffs will affect SPCR:

Code: Select all

        Old  New
Red      2    3   +mas92264

Maroon   2    2   +CharlieChan

Brown    5    8   +genetomcat, BMWBMW, tangled, shens

Green   13   15   +MikeChin, ColdFlame, bcassell, Hortalonas, knarf242, shad0

Blue     ?   +2   +WarpedPlatter, Wrah, PaleMelanesian, captain_android, Kemokim,
                   mobquet, olos, burcakb

Teal     ?   +4   +powergyoza, jvanlaven, Fastman, Mormakil, Graham_Astor,
                   KnightFallen, Pete_Amer, ojg, mortinal, Michael_qrt, Putz,
                   Drewdog2323
                   Close: kloppe, RedSnertz, Paul, chylld, Vergozan, Choy
               

Obviously the biggest effect is raising 12 worthy folders from black to teal, with 6 others within a point or two.

I am pleased with these changes. They feel "right".

David

[haysdb]

Jason has implemented the new points cutoffs.

Beneficiaries of the change (those who have changed color because of the change):

• Red: mas92264
• Maroon: myself and CharlieChan
• Brown: Zyzzyx, genetomcat, shens
• Green: Hortalonus, Lenny, Mike_Chin, Sbabb, Semm, ColdFlame, sthayashi, Kemokim
• Blue: mjrusso45, WarpedPlatter, Wrah, CoolGav, wussboy, captain_android, wumpus, bucakb, Copper, Hao_Dong
• Teal: Rusty075, Choy, NeilBlanchard, DanOnKeys, jvanlaven, Buddabing, Mormakil, Unregistered, Einrik, DryFire, sonofdbn, blahblahbloo, Michael_qrt, samwc912, quksilver, Putz, kloppe, Cameron_Ashizawa

David

[Zyzzyx]

I'd like to point that I would be 'Brown' now, even with the old point ranges. Currently at ~930ppd.