We will soon release a new version of SETI@home, based on BOINC, a general-purpose platform for distributed computing projects like SETI@home. Eventually other distributed computing projects (like Folding@home and ClimatePrediction.net) will also use BOINC, and you’ll be able to share your computer time among projects of your choosing.
BOINC makes it possible for us to release new versions of SETI@home without requiring you to download and install new software. It will also let us do new analyses of the SETI@home data, looking for other types of radio signals (such as short pulses from evaporating black holes).
Switching to SETI@home/BOINC will be easy:
Create an account using the SETI@home web site. This will give you an account ID, sent by email. Note: accounts will created automatically for all current SETI@home users (see below).
Download and install the BOINC client program when it becomes available.
Uninstall SETI@home.
When you first run the BOINC client, enter your SETI@home account ID.
Stages of the transition
We don’t have an exact timetable yet, but the transition will be staged as follows:
We will make a snapshot of SETI@home user information (accounts, teams, profiles) and will use it to initialize the SETI@home/BOINC database. We will then launch the SETI@home/BOINC project.
Over the next month or so we will send email to all SETI@home users, giving them their new account IDs and recommending that they switch to BOINC. During this period you can continue to run the current SETI@home, and your results will be recorded and used. New SETI@home accounts and team changes can be made, but will not be carried over to BOINC.
We will turn off the current SETI@home server. At this point you will need to switch to SETI@home/BOINC.
What will happen to my workunit totals?
BOINC keeps track of your computer’s work in terms of actual computation, not workunits. This is necessary because BOINC projects may have workunits of many different “sizes”. Because of this change, all SETI@home/BOINC accounts will start with zero credit.
However, the workunit totals from the current SETI@home will be recorded in our database, and a section of our web site will show the final totals.
What will happen with SETI@home teams?
All current SETI@home teams, and their membership, will be copied over to SETI@home/BOINC. The work totals for teams will also start from zero.
What will happen with programs like SETIQueue and SETI Monitor?
These programs (which have been very useful with SETI@home Classic) won’t work with SETI@home/BOINC. Their main function (buffering multiple work units) is provided by BOINC itself, which allows you to specify how much work your computer should get each time it contacts the server.
Will the format of input and output files change?
Yes. Programs which display information about the signals found in SETI@home work units will need to be modified to support the new (XML) data formats used by SETI@home/BOINC. Authors of such programs are encouraged to contact the SETI@home team for sample input and output files.
An excellent post Sir Ulli… I just want to re-post a few snippets from that to ensure they are not missed:
Stages of the transition
We don’t have an exact timetable yet, but the transition will be staged as follows:
[list=“1”][li]We will make a snapshot of SETI@home user information (accounts, teams, profiles) and will use it to initialize the SETI@home/BOINC database. We will then launch the SETI@home/BOINC project.
[/li]> [li]Over the next month or so we will send email to all SETI@home users, giving them their new account IDs and recommending that they switch to BOINC. During this period you can continue to run the current SETI@home, and your results will be recorded and used. New SETI@home accounts and team changes can be made, but will not be carried over to BOINC.
[/li]> [li]We will turn off the current SETI@home server. At this point you will need to switch to SETI@home/BOINC.[/list][/li] What will happen to my workunit totals?
BOINC keeps track of your computer’s work in terms of actual computation, not workunits. This is necessary because BOINC projects may have workunits of many different “sizes”. Because of this change, all SETI@home/BOINC accounts will start with zero credit.
However, the workunit totals from the current SETI@home will be recorded in our database, and a section of our web site will show the final totals.
What will happen with SETI@home teams?
All current SETI@home teams, and their membership, will be copied over to SETI@home/BOINC. The work totals for teams will also start from zero.
So make sure you still have access to the email address for your account so you can get your new ID…
I’m presuming that all the current beta testing results for BOINC will be killed off, along with all the Teams/users?
Originally posted by Kwazy_Wabbit
[B]thanks for the heads up
Guess I wont make it into the top 1000 by the end of the current project [/B]
Don’t fret m8, SETI 1 isn’t over yet. When I first read that piece I thought “shoot, only a month till SETI2!”, but it seems to mean that a month after they eventually take a snapshot of the database will be the end of SETI 1… and the snapshot date is probably at least a good couple of months away, 'cause there’s not a public release of the SETI/BOINC client yet (it’s being tested in-house by the berkeley guys I think).
True but looking at their lovely little project timetable and it showing the seti1 server being turned off at the end of the month plus this announcement with the “We will soon release a new version of SETI@home” opener I’m leaning towards the remote possibility that they might actually do it at the end of the month…
Then again I could be wrong as I have been in the past. :rolleyes: :flip:
/goes off in search of bits to build a 1G Tbird cruncher…
2 months seems too short to me bearing in mind that there has not even been a limited public beta test so far.
BOINC released a new beta version 2.04 yesterday which is currently running the Astropulse project. Seti units are sat in the database ready for download and they tell us that the seti2 software will automatically download when seti units are encountered but it hasn’t happened yet.
I’m still not sure I’ll particpate in Boinc when its the only option on offer from Berkeley.
One of the best parts of Seti-1 was working out a way to cope with the shortcomings of the client particularly things like add on caching, progress monitoring, stealthing, crunching without a net connection. Taking away all of these challenges and attempting to reproduce the caching stuff inside the client makes the prospect seem a little dull by comparison.
The ability of the client that to download new code and run different tasks on the whim of somebody on the other side of the world is a point for concern. In days gone by a process doing that would have been call a Trojan. This will just about scupper any chance of continuing to use work equipment for Seti.
Even I wouldn’t run boinc on an office machine.
I understand your reservations m8. We’ll just have to see how it works out on the security front. With respect to a challenge, I’m sure there will soon be lots of semi functioning third party add ons and stats utilities to give us a bit of excitement.
Originally posted by The Balrog I understand your reservations m8. We’ll just have to see how it works out on the security front. With respect to a challenge, I’m sure there will soon be lots of semi functioning third party add ons and stats utilities to give us a bit of excitement.
Date: Mon, 06 Oct 2003 22:00:24 -0700
From: David Anderson <davea@ssl.berkeley.edu>
To: boinc-beta@lists.sourceforge.net
Subject: [Boinc-beta] missing download file bug
We may have fixed the problem that resulted in lots of
results being issued after their input files had already been deleted
(causing download failures).
Other project news:
we’re very close (and I mean it!) to releasing the Windows client
for SETI@home/BOINC.
Originally posted by MrTFWitt The ability of the client that to download new code and run different tasks on the whim of somebody on the other side of the world is a point for concern. In days gone by a process doing that would have been call a Trojan. This will just about scupper any chance of continuing to use work equipment for Seti.
Even I wouldn’t run boinc on an office machine.
Don’t you have the option to only run specific processes under BOINC? IIRC you can just specify that you want to run SETI and nothing else…
Originally posted by riddlermarc Don’t you have the option to only run specific processes under BOINC? IIRC you can just specify that you want to run SETI and nothing else…
Yes, but this param is stored together with the other params ‘on the other side of the world’. Borkeley seems not very responsable to this security concerns. Basically a “you can trust us” was all I saw so far.
I personally will not run it on any network (work or home) until I can hold the params used by BOINC locally on my network under my exclusive control.
/wonders if hackers around the world are already waiting to activate the biggest fleet of ghost computers ever seen for the ultimate DDOS attack
Subject: Re: Boinc and Auto Update of what ??? Message 4 of 13
Posted by Eric Heien on Oct 5, 2002 at 02:34
Reply to this message.
The auto update feature is in BOINC, but not in the capacity that you seem to be
thinking of. Projects can update their executables, but participants have to
explicitly sign up for a project to have anything from them be downloaded. To
keep viruses from being distributed in the auto update by malicious hackers,
BOINC uses a 1024 bit RSA code signing scheme. You can take a look at the
current BOINC source code at boinc-status download | SourceForge.net. We’d love to have
any comments on it. Good luck,
We already linked over SETI@Home classic statistics to SETI@home/bOINC. When S@H classic ends, the statistics will be frozen and show your final # WUs, total CPU time, etc.
In the context of the rest of the thread I take this to mean that final SETI 1 scores will be displayed with/alongside/generally adjacent to the SETI 2 stats
[/B]
