[arkayn]

Looks like Mac's that are not running the optimized app are unable to get work right now due to a bug in the code. They are getting a "Seti@hom Enhanced is not available for your type" error message when contacting the server.

[BTBlomberg]

arkayn,

I have 2 G4 10.5.4 Laptops and 1 Intel 10.5.4 Hackitosh Running SETI optimized and have just checked the logs and I am getting work just fine today on all 3. Could it have been a connection problem at SETI?

[arkayn]

Those of us running the optimized app have no problem, it is the ones running the standard app that cannot get work.

[BTBlomberg]

Sorry, I did not read the NOT in your post.

I guess there are more reasons to run optimized than just the better performance then.

[arkayn]

I have already gotten 2 people over at SETI up and running with the optimized apps now.

[BTBlomberg]

If you are running the real client you get automatic app updates, so I wonder if they pushed a buggy version recently. Optimized, obviously, overrides this.

[The Ancient One]

I installed the optimized apps and ran the installer, but I'm still not running. Is there a step-by step guide for idiots?

[The Ancient One]

Maybe if I installed the optimized apps in the correct folder it would produce better results.

[Todd Madson]

Sorry about the long silence. I've had sick cats and work has been nuts.

RAC is still down (presently 621.10).

I upgraded the ram from 3.5 gigs to 6, we'll see if that has
any effect at all (maybe I should run it out of a ramdisk).

Weirdly, it looks like Seti still has yet to calculate credit for
work units on this computer dating back to the 12th of July.

No idea why that is happening.

[Todd Madson]

I'm running Boinc Manager 5.5.0 as well as the Alex optimized client but am getting this nonsense:

"No Work From Project, deferring scheduler requests for 1 hr, 45 minutes and so and so seconds".

Looks like my recent RAC peaked at 838 something on July 17th and it's dropped ever since,
it got as low as 569 on August 8th and raised slightly since then.

It is giving me "found app_info.xml: using anonymous platform" as was indicated was one
sign Alex's app was going, as well as "using setiathome_enhanced_version 527".

[BTBlomberg]

They had server problems over the weekend and are trying to clear all those WUs we have returned after it was back online, then after the disk space is cleared they can make new WUs to crunch. The uploaders were not working right for 2 or more days, so it is all backed up. New WUs are going out, but it's the luck of when you connect and if they have some when you do. I finally got some on a couple of machines, but not all.

As for not calculating credit, you may be waiting on someone else to return the same WUs back. That would be the only reason for that, well other than a technical issue.

I have been seeing a lot of small WUs. Statistically they should boost ones RAC, but then there is the lag time in switching WUs that may factor in there too. Not sure.

[BTBlomberg]

Originally Posted by Todd Madson 

I'm running Boinc Manager 5.5.0 as well as the Alex optimized client but am getting this nonsense:

I just upgraded to 6.2.15 Friday and it is good. At first I thought it was the problem with the uploads until I read the SETI technical forum. It's fine now.

Originally Posted by Todd Madson 

It is giving me "found app_info.xml: using anonymous platform" as was indicated was one
sign Alex's app was going, as well as "using setiathome_enhanced_version 527".

This is right. You won't get Autopulse, but they have no optimized app for that yet anyway and it's long one to run.

[arkayn]

They just added a preference on the site to select whether you want to crunch Astropulse or Multibeam or both.

[Todd Madson]

Some interesting things I've noticed:

I've watched my G5 and it usually has 2 work units in the tasks window
with 10-20 minutes to complete on each.

They finish and it then uploads and then it seems to sit there for ages
waiting for more work units to work on with no work to work on.

I run temperature monitor so I can tell just by when the temperatures
on the processors are low when the machine is working on work units
and when it is not, it seems to sit idle for long periods.

My Vista laptop on the other hand probably at least 100 work units queued
up and waiting at any given time while two are being crunched.

I realize the C2D chips are more efficient but I get the idea the problem is
I've got a machine that could be working on more work but isn't doing so.

I realize they had a server outage last weekend but I think I'll update the
boinc app at home to see if this helps.

It's almost like there's a "built in waiting module" for this app now on the
PPC side.

[arkayn]

Check the messages tab, it will probably say that it is requesting work but not receiving any. The work just started flowing last night after the weekend crash.

[Todd Madson]

Problem fixed. Apparently in the last version of the software it would only work on two work
units at a time, upload them then sit there for hours until it decided it wanted to work on
some more work.

I downloaded 6.2.15 just a few minutes ago and now I've got a PLETHORA of work units to
crunch on so I think it fixed the problem.

What a weird deal - the computer was just sitting there waiting most of the time.

Checking temperature monitors' graphs this thing was idle most of yesterday.

[BTBlomberg]

SETI@ Home has not been right all week, since the problem last weekend. New WU have been trickling out and getting them have a luck of the draw (if you connect and they have some from the splitter you get them, otherwise not). Also, you have a lot more control with the new Boinc apps. I had set up preferences individually on each machine and just changed them all with the web site settings and that was nice that it overwrote them on all clients.

Hopefully they get their stuff together after Matt (at Berkley) gets back. He has been on vacation since this started.

[Billy]

Looks like there has been a big drop in credits granted for Seti work starting about 2 months ago.

http://www.allprojectstats.com/prgr....jekt=15&type=8

http://www.allprojectstats.com/prgr....jekt=15&type=5

[arkayn]

I am down about 1100 RAC since they implemented the change. 4200-----3100

[Gareth Johnston]

This adjustment in credit is slightly more than 15% or is it just me.

I was tinkering with number last night and it works out to be between 24% and 28% depending on the units.