Science United ???

Message boards : Questions and problems : Science United ???
Message board moderation

To post messages, you must log in.

AuthorMessage
MossyRock

Send message
Joined: 19 Oct 13
Posts: 29
United States
Message 84184 - Posted: 4 Jan 2018, 6:53:51 UTC

The URL http://boinc.berkeley.edu goes to a site called "Science United." It says "Science United alpha test."

The URL https://boinc.berkeley.edu goes to the familiar BOINC home page.

As you can see, the only difference is the "s" in the http.

What is up with this? What is "Science United?"

Thanks.
ID: 84184 · Report as offensive
Profile Jord
Volunteer tester
Help desk expert
Avatar

Send message
Joined: 29 Aug 05
Posts: 15477
Netherlands
Message 84185 - Posted: 4 Jan 2018, 7:29:06 UTC
Last modified: 7 Jan 2018, 18:16:32 UTC

Please read {redacted}.

It's David's new thing. See https://boinc.berkeley.edu/dev/forum_thread.php?id=11647#78714 for more on that.
ID: 84185 · Report as offensive
Richard Haselgrove
Volunteer tester
Help desk expert

Send message
Joined: 5 Oct 06
Posts: 5077
United Kingdom
Message 84188 - Posted: 4 Jan 2018, 10:08:53 UTC

Existing projects, like for example SETI@Home, have a 'powered by BOINC' button on their front page. That links to the non-https versions of the BOINC url - hence, to an alpha test page for Science United.

Existing projects like NumberFields@home which still use the traditional BOINC web interface have a 'Join our project' panel on the front page. Step 2 is 'download BOINC': again, this uses the non-https form of the link. This takes you to a login page for SU.

I suppose this is one way to secure success for a new, private, non-community-managed, project: kill the one you're leaving behind.

This, like so much recent BOINC development, hasn't been thought through at a full systems analysis level.
ID: 84188 · Report as offensive
Profile Jord
Volunteer tester
Help desk expert
Avatar

Send message
Joined: 29 Aug 05
Posts: 15477
Netherlands
Message 84189 - Posted: 4 Jan 2018, 10:22:31 UTC
Last modified: 7 Jan 2018, 18:16:51 UTC

The 'normal' page for it seems to be {redacted}, because all of the join/login pages link to this domain on a secure page as well. I'll ask David if he intents to use the not-secure page as well, but it seems to be a fluke.
ID: 84189 · Report as offensive
Richard Haselgrove
Volunteer tester
Help desk expert

Send message
Joined: 5 Oct 06
Posts: 5077
United Kingdom
Message 84190 - Posted: 4 Jan 2018, 10:30:22 UTC - in response to Message 84189.  

Well, I've sent out an APB to project administrators to cover the situation until David wakes up in the Pacific morning.
ID: 84190 · Report as offensive
Richard Haselgrove
Volunteer tester
Help desk expert

Send message
Joined: 5 Oct 06
Posts: 5077
United Kingdom
Message 84191 - Posted: 4 Jan 2018, 10:45:11 UTC

Most of the 'home page' links here and in the Wikis - like the BOINC logo you see top-left as you read this - link to SU as well. This SUcks.
ID: 84191 · Report as offensive
MossyRock

Send message
Joined: 19 Oct 13
Posts: 29
United States
Message 84204 - Posted: 4 Jan 2018, 17:11:02 UTC

Well, I certainly hope that this is an unplanned move/mistake, because to foist something like this on us without warning, with no explanation or transition, will result in a lot of p*ssed off crunchers.

I read through the links explaining the history and reasoning behind a move to the "Science United" model and can understand the desire to simplify volunteer computing (VC) for people who are not that tech savvy in order to grow VC participation.

However, the removal of granular control of what I'm crunching left me stone cold. I would think that the majority of the very dedicated crunchers, who are passionate about certain projects, would feel the same way.

If the existing BOINC model can exist simultaneously with this "new" model and I can still use the existing BOINC model, I'll be ok.

However, if the existing BOINC model is replaced with this "new" model, I think it will be a disaster.
ID: 84204 · Report as offensive
MossyRock

Send message
Joined: 19 Oct 13
Posts: 29
United States
Message 84208 - Posted: 4 Jan 2018, 21:04:06 UTC

In reading that post more carefully, I see that David Anderson said, "TBD is not a BOINC replacement. It's a new project manager for volunteers who
want to use it, and it's technology to help HPC centers add BOINC back ends."

I hope that this approach is maintained.
ID: 84208 · Report as offensive
Profile Jord
Volunteer tester
Help desk expert
Avatar

Send message
Joined: 29 Aug 05
Posts: 15477
Netherlands
Message 84213 - Posted: 5 Jan 2018, 3:08:16 UTC

The URL to the insecure front page of the BOINC website is now again pointing at the BOINC website.
ID: 84213 · Report as offensive
Richard Haselgrove
Volunteer tester
Help desk expert

Send message
Joined: 5 Oct 06
Posts: 5077
United Kingdom
Message 84218 - Posted: 5 Jan 2018, 11:08:26 UTC - in response to Message 84213.  

The URL to the insecure front page of the BOINC website is now again pointing at the BOINC website.
Yes. It's been confirmed to me (private email) that it was a configuration error, now fixed.
ID: 84218 · Report as offensive
Richard Haselgrove
Volunteer tester
Help desk expert

Send message
Joined: 5 Oct 06
Posts: 5077
United Kingdom
Message 84255 - Posted: 7 Jan 2018, 12:41:27 UTC

Gary Charpentier has reposted a blog from the Planetary Society about the genesis of SETI@Home, and hence of BOINC.

One paragraph caught my eye:

In 1999, Berkeley released the result, SETI@home, and since then more than 8 million people have downloaded the program and donated spare computing power to help search for intelligent life. The open-source software, BOINC, on which SETI@home is based, is now used for other projects. This led to what Werthimer calls "the democratization of supercomputing," where users can choose individual research programs to assist.
Is Science United in danger of de-democratising it again?
ID: 84255 · Report as offensive
Profile Gary Charpentier
Avatar

Send message
Joined: 23 Feb 08
Posts: 2462
United States
Message 84259 - Posted: 7 Jan 2018, 15:21:53 UTC - in response to Message 84256.  

Circa 2004 one project ran antrax briefly in collaboration with the dod (so happens the N Korean soldier who fled to SK had antrax antibodies). Later all references were removed this project ever ran. SU of course opens the door to รจ.g. run bio-warfare type sciences again without anyone agreeing, hidden behind a general 'medical research' tickbox, i.e. Clear framing of what's in/out is needed or trust runs risking to be out the window at the first discussion.

How does any of us know what the actual crunching of a project is for? Any duplicitous person can easily fake documentation saying it is for A when it is for B.
ID: 84259 · Report as offensive
Richard Haselgrove
Volunteer tester
Help desk expert

Send message
Joined: 5 Oct 06
Posts: 5077
United Kingdom
Message 84260 - Posted: 7 Jan 2018, 17:07:55 UTC - in response to Message 84256.  

Circa 2004 one project ran antrax briefly in collaboration with the dod
Was that a separate project just grabbing the open-source software (as anybody can) and putting it out there, or was it one of the established projects doing a bit of contract work on the side?

If it was an open-source grab and run, I don't think we have to worry so much. It would still be possible now, but people would have to find the project on the web and attach to it manually, just the same as they do now. The Alpha site for SU says:

Eventually the selection of science projects, and the allocation of computing power among them, will be determined by an international committee of scientists. For now, all projects vetted by BOINC are included, and they have equal allocations.
The projects list here is the same as the list of projects shown when you 'Add project' using BOINC Manager. It doesn't prevent a project going rogue after being added to that list, but it is one level of safeguard.

I'm more worried about SU being developed 'under the direction' of just one man, with the intention of being moved "to a community-based model" once the money runs out. Having just sat (with Jord) for six months on a working group coping with the fallout from the last time that was attempted, wouldn't it be better to apply community direction right from the start, rather than leaving the community to pick up the pieces after the major decisions have already been made?
ID: 84260 · Report as offensive
Profile Gary Charpentier
Avatar

Send message
Joined: 23 Feb 08
Posts: 2462
United States
Message 84261 - Posted: 7 Jan 2018, 17:55:52 UTC - in response to Message 84260.  

Circa 2004 one project ran antrax briefly in collaboration with the dod
Was that a separate project just grabbing the open-source software (as anybody can) and putting it out there, or was it one of the established projects doing a bit of contract work on the side?

If it was an open-source grab and run, I don't think we have to worry so much. It would still be possible now, but people would have to find the project on the web and attach to it manually, just the same as they do now. The Alpha site for SU says:

Eventually the selection of science projects, and the allocation of computing power among them, will be determined by an international committee of scientists. For now, all projects vetted by BOINC are included, and they have equal allocations.
The projects list here is the same as the list of projects shown when you 'Add project' using BOINC Manager. It doesn't prevent a project going rogue after being added to that list, but it is one level of safeguard.

Well, just for the heck of it I went to CAS@home and looked for the list of applications. There was none. They do list some projects.

That list of BOINC "vetted" projects starts with "projects known to us that we think might be telling the truth about what they do."
ID: 84261 · Report as offensive
Profile Jord
Volunteer tester
Help desk expert
Avatar

Send message
Joined: 29 Aug 05
Posts: 15477
Netherlands
Message 84262 - Posted: 7 Jan 2018, 18:15:19 UTC - in response to Message 84261.  
Last modified: 7 Jan 2018, 18:17:23 UTC

What, you mean http://casathome.ihep.ac.cn/apps.php?
If a BOINC project uses the BOINC front end, they may not show the links but they'll be there, because they're in the source code. You can't make your front page without them.
ID: 84262 · Report as offensive
Richard Haselgrove
Volunteer tester
Help desk expert

Send message
Joined: 5 Oct 06
Posts: 5077
United Kingdom
Message 84263 - Posted: 7 Jan 2018, 18:16:22 UTC - in response to Message 84261.  

Well, just for the heck of it I went to CAS@home and looked for the list of applications. There was none.
http://casathome.ihep.ac.cn/apps.php works just fine - they just don't advertise it on the home page. (That's the benefit of common standards and common software)
ID: 84263 · Report as offensive
Profile Gary Charpentier
Avatar

Send message
Joined: 23 Feb 08
Posts: 2462
United States
Message 84268 - Posted: 7 Jan 2018, 22:45:24 UTC - in response to Message 84263.  

Well, just for the heck of it I went to CAS@home and looked for the list of applications. There was none.
http://casathome.ihep.ac.cn/apps.php works just fine - they just don't advertise it on the home page. (That's the benefit of common standards and common software)

And how quickly could they fork it to allow apps that are not on the now hidden list?
ID: 84268 · Report as offensive
Richard Haselgrove
Volunteer tester
Help desk expert

Send message
Joined: 5 Oct 06
Posts: 5077
United Kingdom
Message 84269 - Posted: 7 Jan 2018, 23:25:35 UTC - in response to Message 84268.  

Well, just for the heck of it I went to CAS@home and looked for the list of applications. There was none.
http://casathome.ihep.ac.cn/apps.php works just fine - they just don't advertise it on the home page. (That's the benefit of common standards and common software)
And how quickly could they fork it to allow apps that are not on the now hidden list?
Hopefully quicker than Synecdoche
ID: 84269 · Report as offensive

Message boards : Questions and problems : Science United ???

Copyright © 2024 University of California.
Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.