|
New Linkage
There is a relatively new (about 2 weeks old) kid on the Ars Technica
Distributed front. This kid goes by the name E. Roll.....as in Team
Egg Roll. So what do these bad boys do? Team Egg Roll
is the Ars entrant in the Folding@Home
Project. The Folding@Home project, is a distributed processing
project involved in the simulation of protein folding. Being a
biochemist in my former life, I know how complicated the folding of
proteins can be. Nature does a pretty good job of it, but with a
heck a lot of amino acids in a protein, the possibilities of how a protein
folds into tertiary and quaternary structures can be endless. This
project uses the CPU cycles of home computers to help in the simulation of
these folding processes. Currently Team Egg Roll is holding down the
#1 team position after rocketing up to #1 in only a couple of weeks.
They have been added to the Links on the main menu....go check them out!
There is another TLC stats page out there...(SAY IT AIN'T SO!).
Krell, of the TLC powerhouse Krell Worshippers has put up a stats
page which keeps daily track of the top 1000 member club.
This page started out because Krell didn't want to wait for the stats
pages here to be up....hey I don't blame him! He is over on the
other side of the pond there :). Even though the link doesn't have a
permanent spot in the menu to the left, go check them out anyway.
Shooting for the Stars...but Ending
Up in the Drink
I just caught this
article over at wired. Apparently there has been a contest
to see who could be the first sort of "public" group to send a 2
kg payload to an altitude of 200km. The winner gets $250K. The
contest ends on Nov 8th. There has been several interested groups in
trying to claim the prize, but escaping Earth's atmosphere is a bit harder
than it sounds. Problems with both getting the rocket to do the job,
and trying to jump through hoops getting the go ahead from the FAA and
other Departments of Red-Tape has derailed most of the groups. But
one group tried today. They lifted their rocket up to around 70K
feet, and then tried to launch the rocket and payload....problem is that
the rocket decided to fire down instead of up....and ended up somewhere in
the Gulf Of Mexico. There is only one other active group....but they
do not seem to be shooting for the deadline time. They want to do it
right, and will launch "when they are ready".
Notice Any Difference?
In page downloads that is... A couple of days ago I started
implementing a little space saving program to whittle down the size of the
chart html pages. I'm not sure if you have taken a look at the size
of the chart pages, but they tend to be on the hefty side. The
problem is due to the charts being created and saved as .html from within
Excel. When I save as an .html file, Excel decides to throw in alot
of cr*p which bloats the html code even more than a Microsoft product
usually does. I guess some people bitched and moaned about how bad
this code was, and Microsoft has a utility that can strip out much of this
unnecessary code and lighten things up with. Now wouldn't it be
great if M$ could incorporate into Office so I can use this
automatically? Heck no. But it does help. The original
size of the chart html file that Excel creates is somewhere around 150K in
size. This utility strips it down to around 110K in size. It
sure helps out in my upload to the site, and it should make it a little
faster to download on your end.
Two More Utilities
There are a couple of new versions of SETI utilities out recently and they
may be of interest to ya'll. The first is Seti
Monitor, this utility offers many different features, and also
gives some caching ability for the GUI client, if that is your bag. Check
it out here. The other utility is called SETI
Stats. This utility is for "checking on your SETI@home
ranking, graphically displaying results and creating signature files based
on your personal data". You can check SETI Stats out here.
Team Lamb Chop News!
Things keep rolling along with TLC. Alot of times when I do the
updates, the numbers just seem to astound. How amazing are these
numbers? Let me count the ways....
- Averaging over 10,000 WU/Day.
Right now this is over twice as many WU per day than any other team.
- Support AIDS Research. Today
marks the 11th day in a row that they have posted 1,000+ work
units. I have never seen this kind of production from
anyone. Two of the last three days they almost broke the 3,000
WU barrier.
- 10 members averaging over 100 WU/Day. These
are consistent high producers for the team. This isn't a big
post one day and cache up a bunch of results....these are consistent
crunchers.
- The other 3000+ Members on the team.
Too often the top producers and the top of the charts/graphs get all
of the attention, but what makes this team is everyone including those
who may only product 1 or two work units per day. Even without
SAR or the other members averaging 100 WU/Day, this team would STILL
crank out about 8000 WU/Day. This would still be about twice
any other team. This team is just strong top to bottom. I
really wish I could figure out some way of highlighting some of the
consistent crunchers down in the ranks, but I haven't yet :/.
Speaking about the underrated ppl....lets talk about svdsinner.
Here is someone who has been a constant cruncher, and doesn't get thrust
into the spotlight that often. But a couple of days ago svdsinner
passed C. Eric Smith for the team #2 spot. He has also moved up to
the #138 position overall quite quietly. Give 'em his due, hard work
has paid off pretty well. Unfortunately the stay at #2 too
long. The SAR juggernaut is poised to pass him within a couple of
weeks. Congrats svdsinner!
One more for the spotlight. Vomesh
has been crunching for TLC for a while now. He shot up the TLC
sometime last year, but went into a dry spell for a good while. He
had reached up to the top 530 or so members overall in the SETI project,
but slowly he lost spots, and fell in danger of falling out of the top
1000 members. Somewhere around the beginning of August he turned
things around a bit, and then turned it completely around at the beginning
of September. In danger of being out of the top 1000 he has rocketed
back up and now stands at #205 overall...and # 7 on TLC. He should
reach #3 sometime within a month. -zAmboni
Got News?
Hrmm..I guess this is as good a time as any to update the news :)...been a
while eh? Where to start...
One of the news stories that got lost in the past week was the
advancement of The Knights Who Say Ni!.
KWSN passed up Microsoft sometime last week for the #8 position in the
overall scheme of things. The Knights were boosted up by several ex
TLC members, particularly Virus and Moses and they are looking to reach #7
sometime soon by overtaking Intel.
The past couple of weeks has seen a period of quite a bit of action on
TLC concerning additions and subtractions of members. Unfortunately,
the distributed
forum has seen quite a bit of talk about this, with people
questioning the motives and identities of different team members.
Things took on an air of soap operaishness (is that a word?) as people got
riled up. Luckily cooler heads prevailed over the situation, and it
looks like TLC will come out the better and stronger. People may
join TLC for a test ride, but that doesn't mean that they will eventually
make a purchase. Also with TLC being one of the premier teams in the
SETI project there may be some people out there to undermine the team and
cause dissention (as Larry Loen eloquently pointed out in this post).
If you want a bit more of an up close you can take a look at the various
threads in the distributed forum for your perusal. Just a word of
warning...don't get too out of hand, Rat Bastard
and I now have moderator privileges on the forum so don't make us get out
the red pencil on your posts!
A Little More on v3.0 CLI WU times
One thing of note in the update to the v3.0 run time prediction article a
week or so ago was that low angle range work unit times were significantly
longer using the CLI than when running the GUI client. For those low
angle range work units the run times on my Win98SE machine had run times
longer than even the mid angle range work units, and took quite longer
than the "expected" run times from Roleof's equations.
Other people have seen the same phenomena with the CLI, but here is an
interesting thing about the CLI. There were several people including
Lawrence Kirby, who noted that the run
times of the CLI on Win2K were on par with the times for mid angle range
work units. Therefore Win2K times were faster than Win98 times (even
though the Win2K times were still longer than the GUI times)....Strange
eh? I still haven't heard a good explanation why the CLI would
behave differently than the GUI on these work units though....maybe we
will eventually get an explanation.
Even More Stats on the SETI Servers?
A bit ago there was a post on the newsgroup asking if they would expand
the top team pages to something like a top 500 instead of a top
200....Eric K came up with a reply:
>Slightly off-topic, but any plans on
explanding the Top 200 stats for
>businesses, clubs, etc.... to possibly the Top 500?
That wouldn't be too difficult. I'll bring it up at the next
meeting.
Might mean that the lists update less frequently, though.
Eric
Well...right now it already seems pretty hit and miss with even the top
200 teams being updated on a regular basis, and they seem to be at the
whims of the stats servers...and this brings me to my next point. Is
there a good measurement (for us out here that is) of the S@H server
health? I have my own little way of seeing how well the stats
servers are handling things, and this deals with two different
items. The first item is how well the SETI servers are updating the
processed work units and adding them to the team totals. For the
most part of the past 5-6 months, the stats servers cannot keep up with
the daily influx of work units processed for each team. When the
servers were down on the 18th to replace the RAID configuration, things
seemed to pick up on the servers...but now they are back to the way they
were...lagging behind and increasing the backlog of work units.
There is another way I gauge the health of the servers, and this is when I
download the stats each day from the SETI servers. I pull the
entire list of members from the top 10 teams each day, and when this
happens the servers have to run through the cgi scripts on the servers to
get this data. How quickly I can pull the data gives me a judge of
how well the servers are holding up. In the past week or so, things
haven't been looking good. I am getting consistent non connects to
the servers for pulling the data (equivalent to a "page not
found" on a web browser), dropped or incomplete downloads from the
servers, and also extremely long times getting the data back.
The situation doesn't look good, but again it also may be explained by a
lower priority on the servers for stats processing...
Yet Another SETI App.
I received an email yesterday from Kevin Noble pointing out there is
another SETI app out there that I haven't looked into. You may want
to take a look at it if it sounds good for ya...I'll let Kevin tell ya:
The site mentions a number of great apps, some
that I use like SETI Driver and SETISPY. Another one that I use is
SETISERVICE, it is a simplified way of getting the SETI executable to
run as a service and easily deployed to other networked workstations on
NT. Setting it for non-interactive to the users makes it
non-intrusive in appearance.
The page is at http://www.geocities.com/petiepooo/seti.html
Weekly Stuff...
I haven't updated them yet, but I should get them up by the end of the
night. -zAmboni
New SETIWatch / SETILog
Mark Loukko, the author
of SETIWatch and SETILog, has released new versions for the V3 client. The
boast a few new features, such as displaying info on angle ranges. Here's
a snip from the release info:
New features for
SETIWatch 3.0
- A new tab called
"Best Results" has been added. This tab contains the best
Peak, Triplet, Pulse and Gaussian for the current running work unit.
- Angle Range has been
added to the "Completed Work Units" tab.
New features for SETILog 3.1
- Now records the best
Triplet, Pulse and Gaussian in the SETILog.csv file.
- You will notice a
message the first time you run SETILog 3.1 and a previous SETILog.csv
file exists.
- This message box will
prompt you to convert your SETILog.csv file into the new 3.1 format
(backup copy of your csv file is created).
Both programs are located
right here. And
for those of you not familiar with them, SETILog is a CLI launcher which
logs the results of each workunit into a database. It's also handy for
hiding the client- I run it on all my machines at home to clean up the
startbar. SETIWatch is a complimentary program that lets you monitor the
progress of your clients- including those located over a network. It's
VERY helpful for those of us with small (or large) Seti farms to contend
with.
Thanks to Netsnipe
for the information.
May his Seti machines never be subject to random network outages -Rat
Small Addendum
Netsnipe
wrote back with a piece of advice for those using SETIDriver in
conjunction with SETIWatch and SETILog. I'm not familiar with SETIDriver,
so I'll let him explain:
I've also noticed that in order to use SETIDriver
and SETIWatch together, that
SETILog must be set as the CL client under the SETIDriver settings and
the 'Create SETILog.csv' checkbox in SETIDriver must be
turned off because the latest SETIDriver 1.5.0.3
only generates 2.x csv records. But another side
affect of changing clients in SETIDriver is that it automatically deletes
'setiathome-3.0.i386-winnt-cmdline.exe' when changing clients, so users
must beware of this when upgrading and restore of a backup of setiathome-3.0.i386-winnt-cmdline.exe
after changing SETIDriver's client setting
to SETILog.
We now leave you to (The message has been
moderated by Rat Bastard) -Rat
Predicting v3.0 Run Times Updated!
I just updated the article on predicting version 3.0 run times. You
can check the update here.
There are some things of interest, including some tweaking on the
prediction of run times, and also a major finding of a difference between
the run times of the GUI and CLI versions of the v3.0 client! If you
want to check out the original article
it may be good for some background. -zAmboni
S@H Berkeley Connection Problems
Well over the past couple of days, there have been some connection
problems for people trying to access the Berkeley servers. Good
thing is that there doesn't seem to be any problems with the Berkeley
servers themselves, but instead it is problems with the U.C. Berkeley
Internet infrastructure. Here are the Tech
News releases on these problems.
October 16, 2000
A power outage on campus resulted in
most of the Berkeley campus being disconnected from the rest of the
Internet between the hours 0200 and 0500 UT. Power has been restored,
but some campus systems are still out.
October 14, 2000
A problem on campus resulted in our connectivity to the outside
world being reduced to 20 Mbps overnight last night. This resulted in
numerous dropped connections and generally slow service.
Mutations Abound
Waffling more than a Presidential candidate on tough debate questions,
Virus has changed allegiances again. Mutating from Virus
- TLC to KWSN - Virus, Team Lamb Chop gains some stability in
exchange for a loss in work units. The Virus Caravan stop lasted for
just about a week at TLC, before moving back to the Knights Who Say
Ni. But, as many of us know here in the US, the cold and flu season
will soon be upon us. Even if you can shake an infection, there is
no guarantee that you wont be immune from another Virus later down the
line.
New Version of SetiSpy Available
Now available for download is version 3.0.1 of SetiSpy. This version
fixes a few small bugs and will update your version to the latest and
greatest iteration. You can download
it here. I also wanted to point out that SetiSpy has
been featured on the ZDNet
Download center, and has been rated as 5 Stars! Way to go
Roelof! Just a word of warning, the newest version of SetiSpy is not
yet available for download on ZDNet (only 3.0.0) so you may want to go to
the first link if you want the latest.
Last 10 Results Page?
You know that you have tried to access it, but it just isn't there
anymore. When will it be back? Well don't look for it to be
available anytime soon. Erik Korpela
gives us the lowdown from the newsgroups:
Probably not for a while. At this point
any database effort given to this detracts from the effort to sift
through the signals in the database for potential E.T.'s. Our hope
was the we'd be getting another system to use for postprocessing, but
that has fallen through for the time being.
Weekly Stuff
Even with the team's losses today...this week has been pretty good team
wise. You can check out some of the weekly
team numbers here. The two big newcomers have taken over the
tops in the charts this week. Support AIDS
research and Antimoses were
on the top of both the movers and work units for the week this past
week. Each one of them taking a top spot. I know Vomesh
must be thinking "what do I need to do?" He has cranked up
the production to pass up even Admins spend CPUs, but then a two newcomers
hope in to take the spotlight away from him! There was one TLC
member who left the top 1000 this week, but we do have another to take his
place. Krell Worshippers debuted
in the top 1000 this week....next week there may be as many as 3 more new
members in the top 1000 club!
WU Time Update...
I may have an update to the Predicting WU Times article up sometime
tonight...depends on if I get caught up in other things....If I don't get
it up today, then hopefully tomorrow. -zAmboni
-Front Page |
