Message boards : News : NSF funds new model for BOINC
Message board moderation
Previous · 1 · 2 · 3 · 4 · Next
Author | Message |
---|---|
Send message Joined: 8 Nov 10 Posts: 310 |
By the way, I am all in favor of supporting quasi-commercial research, if that happens to be included. That is much closer to real-world applications. This notion that if no one can profit from it then it must be good research comes from people who are naive in the field. Because you see real problems. The pitfall of academic research is that you may be working on a protein (for example) that has interesting properties to scientists, but is of no great relevance to any disease. A lot of computer time is spent, a paper is published, and success is declared. But nothing happens with the research after that. It happens all the time. Commercial work is the opposite. They start with a real problem, and try to find a solution. At least they are working in the right ball park. |
Send message Joined: 27 Jan 16 Posts: 16 |
Commercial work is the opposite. They start with a real problem, and try to find a solution. At least they are working in the right ball park. "I have become death, the destroyer of worlds" Robert J Oppenheimer (quoting the Bhagavad-Gita) On his work in creating the thermo-nuclear device. The problem with commercially motivated science is that the main focus becomes what we CAN do, not SHOULD we do it..... Advancement without understanding is the hallmark of science for profit... For example Celebrex, the revolutionary pain killing drug, touted as the savior of everyone in chronic pain.... Sold to millions for millions, and prescribed like candy. Until it started killing people..... Science with a profit motive is not the best approach in the practical world..... |
Send message Joined: 1 Jul 16 Posts: 146 |
As to the matter of the back-office required by the platform on the project - David has identified that as being an issue for many smaller projects in his paper, and unless there is a substantial change in the underlying philosophy I can't see that changing significantly. There are certain functions that don't change, e.g. user management, application development, data management. This is the real benefactor with this as I see it. Small projects or projects with inconsistent work can plugin to a 'grid' of sorts. There are already several grid projects and this seems to be more open to any work. |
Send message Joined: 27 Jan 16 Posts: 16 |
This is the real benefactor with this as I see it. Small projects or projects with inconsistent work can plugin to a 'grid' of sorts. There are already several grid projects and this seems to be more open to any work. That is only part of the equation, a very valid part yes, but the smaller part nonetheless. The bigger part is plugging the crunchers into the national shared supercomputer grid. Making them available as a general resource to computer science. But to do that you have to assure two things, always on and no choice... He's aiming at the cellphone network of connected android devices..... Basically the young that really don't pay attention to what is happening on their hardware. The other issue is revealed in his position that Boinc is a dead end as far as academia is concerned (TACC specifically) as the developer and main touter, this has the effect of leaving him behind, by getting Boinc patched into TACC his creation is now relevant again... He's back on the cutting edge of DC in the scientific community.... We cannot ignore those personal motivations either since this has never been about altruism.... |
Send message Joined: 8 Nov 10 Posts: 310 |
Sorry, that is propaganda, not experience. I have been there (though not in the pharmaceutical world.) |
Send message Joined: 8 Jun 17 Posts: 1 |
First of all: Smart phones may have free cycles, but they already have, at best, enough battery to last through the workday. Using their free cycles is not going to make anyone happy. Second: Yes, people are motivated by science goals, not credit, that seems like a very fair statement, BUT... The key point here is that they're motivated by concrete goals. Here's an example: I'm very much interested in space and our universe, as such im in LHC@Home and Cosmology@Home. Now this is all good, if i selected my goals then your algorithm would likely donate my computing power to those two along with other things. Problem here though is that I don't actually want to donate ANY of my cycles to other projects. Lets take Milkyway@Home as an example (this will be a long example, sorry for that, but i hope it clears up the thought process here). I loved Milkyway@Home description, all keywords match, even goal is great, etc. After searching through their site though - it appears that their "Science" page which is meant to explain what calculations they're performing is incredibly outdated. To the point that there's a quote in there about a paper 'to be published soon in 2013'. The only other link they have there is to their YouTube page, which unsurprisingly has about 6-8 very short very seemingly low quality/scale simulations that I'm certain my computer alone could simulate in real-time. All those videos are 3+ years old. After that the only posts from the project are random nonsense and self-praise for the " only black girl that got a degree in [somthing bla bla] at our university". This is NOT what i care about. I don't want to give a single cycle to people who try to sell themselves through some cheap advertisements and give no apparent results. Speaking of the results and goal. Their description is "Milkyway@Home uses the BOINC platform to harness volunteered computing resources, creating a highly accurate three dimensional model of the Milky Way galaxy". There is no three dimensional model that i can find anywhere, maybe it's not accessible, maybe not complete, maybe not even visual, but either way there does not seem to be any relevant communication from the projects owners. Anyways, enough rambling. Most importantly: My point here is that volunteering is like donating. I may support a certain cause, but not all organizations that claim to work towards it. I like technology, but i don't support Apple, i simply don't like their products. Same applies to science, I might have scientific goals in mind, but that does not in any way mean that I will give my cpu cycles to any organization that happens to relate to my interests. I hope this makes everyone's view a bit clearer. |
Send message Joined: 27 Jan 16 Posts: 16 |
....... Propaganda? Ok, at least we know your biases. And are a believer in non-responsible science. (probably in the pursuit of greater profits, also, it is interesting that you didn't comment on Dr Oppenheimer's opinion of his life's work when he finally realized what he had done in the absolute pursuit of his goal) Refusal to recognize that science in the pursuit of knowledge has killed and destroyed more than all the religions of the world combined is an ideopathic response to one who questions the efficacy of the absolute scientific model. (all science is good no matter what it results in) Not trying to slam your opinion, just pointing out the logical fallacy inherent in it. |
Send message Joined: 29 Aug 05 Posts: 15551 |
Can you keep the personal attacks out of the conversation, please? |
Send message Joined: 27 Jan 16 Posts: 16 |
Can you keep the personal attacks out of the conversation, please? I don't see any personal attack, But, that being said, Will do.... (but it is a two way street) And thanks for the reminder of why I don't post here very much.... I will now leave the conversation to the more erudite amongst us... |
Send message Joined: 4 Aug 16 Posts: 13 |
I do not post in these forums normally but as an older member of the community I always worry when I see a possibly one dimensional commercial idea spring to life from within a purely voluntary process. I have 'dabbled' in the BOINC world since the early days when SETI began and only with the passage of time have I become a convert. The ethos of BOINC and its projects has always been the contribution to science by normal everyday people with normal everyday equipment. To me BOINC is an open-source volunteer oriented computing grid that combines the processing power of all individual users for the purposes of scientific research. It's free, and harnesses the unused clock cycles from processors and graphics cards to attempt to cure cancer/aids/ebola/malaria, map the milkyway, crack enigma codes, etc.. Yes there are a few of us who really take this seriously and 'crunch' science projects for the joy of earning the credits and badges of honor. We crunch away even though we are spending our hard earned money to pay for the electricity that keeps those computation cycles going and take joy from the fact that 'we are doing our bit'. I never had the reason, or the inclination, to join any of the array of teams that are available in the BOINC community in fact I've always been a bit of a loner. However in recent years I became aware of one of the teams who, whilst 'crunching' on a long list of BOINC projects, were running a sideline project which has after a few years has turned into the Gridcoin Community. The Gridcoin Community are BOINC users who love their projects just as much as the all the other BOINC users. However here's the difference: Gridcoin was created as a method to help compensate citizen scientists donating their spare resources, to help the common good, by solving complex problems through distributed computing. This compensation may then be used by BOINCers to defray electricity expenses, be donated to their favorite project to assist in its funding, or simply used to buy equipment. Gridcoin rewards BOINC computation using the Distributed Proof of Research (DPOR) reward mechanism, which is a combination of Proof of BOINC (POB) and Proof of Stake (POSv2). Proof of Work (POW) consensus mechanisms are not utilised by the Gridcoin network, making the Gridcoin cryptocurrency network far more energy efficient than any existing POW cryptocurrencies. Gridcoin (Ticker: GRC) is a decentralized, open source math-based digital asset (cryptocurrency). It performs transactions peer-to-peer cryptographically without the need for a central issuing authority. Gridcoin was the first block chain protocol that delivered a working algorithm that equally rewards and cryptographically proves solving BOINC (Berkeley Open Infrastructure for Network Computing) hosted work, which can be virtually any kind of distributed computing process (ASIC/GPU/CPU/Sensor/Etc). So how have I benefited from my being a citizen scientist and donating my spare resources to BOINC projects? As I mentioned earlier, I'm from the older generation, yes even older than Mr. Anderson. Since I really began to get more involved in the Gridcoin BOINC Community I have discovered how many brilliant minds there are out there within this community and after investing some time in the community I have found that although, I may be elderly in my eyes, I too have a niche here. My interest in science has not abated but has been heightened and my new found interest in blockchain technology has taken me places I would never have gone before. I am of the belief that BOINC should not fundamentally change. The concept is sound. We do however need to bring it in out into the light and make it more mainstream worldwide. Development within BOINC to stream line the product should include equipping it with more sensible security which will make it more attractive to the mainstream institutions (commercial or other wise). The main reason cited by organisations for not running or allowing BOINC to be run on machines with in the IT industry and beyond is security. Citizen scientists should be offered challenging, interesting and useful projects to donate their spare resources to in order to help the common good, by solving complex problems through distributed computing. It is simply not enough with to keep BOINC, already deemed to be the largest computing grid in the world, as it is. We need internal innovation not a move to channel present and potential users to a new format. The ideas I see in forums, on IRC and hear in our Gridcoin Hangouts are not 'pipe dreams', they are sensible cogent and considered. Perhaps Mr. Anderson would like to join one of our hangouts in the near future to put forward his ideas for the future of my favorite hobby and obsession. |
Send message Joined: 9 Jun 17 Posts: 5 |
I second the idea of using blockchain currencies to promote VC. If you are looking for computing power to solve problems, there is only one place computing power is being wasted in those quantities: Blockchain Proof of Work Hashing. It is the most logical way to form a competitive ecosystem which benefits scientific projects in need of number crunching power, and the only way to attract serious investment into the BOINC system. |
Send message Joined: 1 Jul 16 Posts: 146 |
I do not post in these forums normally but as an older member of the community I always worry when I see a possibly one dimensional commercial idea spring to life from within a purely voluntary process. That was a straight up ad for Grincoin, much of which taking straight from the gridcoin home page word for word. This is not a team campaign thread. |
Send message Joined: 4 Aug 16 Posts: 13 |
That's unfair and unwarranted. This is not an advert for Gridcoin in any way. I am speaking from the heart when I say that changing BOINC would be a something I wouldn't want to be part of. The mentioning of gridcoin in my post is to tell anyone reading my post that the reason I have stayed BOINCing is a result of my discovery. I would not be able, as a retired person, to continue indulging my passion for science and contributing to the projects I support if I wasn't able to defray my expenses in some way. I thought that there may be other people out there of my age (or younger) who might be interested in knowing about the solution I had found. Very small minded of you to comment in such a way about an innocent post by a concerned user. If I have used any info from a Gridcoin site it was purely to give a rational, intelligent and logical explanation of how the project works. Also I am totally aware that this is not a team campaign thread, I'm not stupid. By the way it's Gridcoin and not Grincoin!! (That was a straight up ad for Grincoin, much of which taking straight from the gridcoin home page word for word. This is not a team campaign thread). |
Send message Joined: 29 Aug 05 Posts: 15551 |
When quoting previous long posts in this thread, and adding a little ditty to it, can people please take the time to cut the quoted material down so it doesn't comprehend the complete quoted post? Keep in mind that there are people reading these forums on small devices, like mobile phones and that that's quite a bit of scrolling to do. |
Send message Joined: 1 Jul 16 Posts: 146 |
That was just a typo. I actually went back to edit my post to make the Gridcoin recruitment rant smaller but there was no BBCode spoiler tag to hide it. I didn't feel like going to another site to see what actually was. The little ditty might be easier seen if the site was setup with better separate of quote and post and between posts. |
Send message Joined: 5 Oct 06 Posts: 5124 |
The quoted text simply becomes part of your reply post. You can delete some or all of it, as you wish - either at the time, or up to an hour later with the 'edit' button. |
Send message Joined: 29 Aug 05 Posts: 15551 |
You have two buttons via which you can answer to people: - the Reply button which gives you an empty reply box, where you can add your little ditty. - the Quote button, which quotes the whole text, for you to the whittle down (with use of Shift + Delete, of Backspace) to the part you want to answer to, or if the post isn't biblical in proportions, the whole post. Both these buttons give the in response to Message N line at the top, always showing which post you answered to. The 'normal' Post to thread button just allows you to post to the thread, not specifically answering to anyone. To look up what part of BBcoide works here, see https://boinc.berkeley.edu/dev/bbcode.php. There are a couple more tags that aren't documented that work as well. By the way, I didn't read the long post as an advert for Gridcoin, more like an opinion on how things can be made better/worse/different than they are now. |
Send message Joined: 27 Jan 16 Posts: 16 |
That's unfair and unwarranted. Yet your doing pod casts for Gridcoin from the dev level..... http://www.stitcher.com/podcast/cm/gridcoin-community-hangouts/e/025-04032017-50076991 Right, some non-advertisement for Gridcoin..... Can you say falsehood? How about we try this one..... https://steemd.com/gridcoin/@mercosity/re-stevescoins-suggestions-for-gridcoin-to-earn-market-value-20160815t205813727z How to strategies to improve GC's market share...... So much for the old guy with a hobby hanging around Boinc from the start looking to defray his electric bill.... Need I research more? Seems very fair and in this case warranted to me..... |
Send message Joined: 23 Feb 12 Posts: 198 |
I'm getting the feeling that this is just Charity Engine 2.0 that removes the potential lottery portion and possibly retains the ability to remain on teams. I think David really misses the real reasons virtualbox projects aren't popular and how much further their road to getting it appealing really is. NONE of the Virtualbox projects run smooth for general DC users. I've ran pretty much all of them in depth. Every system has to be tweaked and monitored a great more deal than non Virtualbox projects. Yes, I know there are a few users out there that brag about how their rigs were simple to set up and run solid trying to champion those projects. They are an extreme minority. Though it may reduce the project scientists' overhead, it drastically increases the users. Most DC'ers will not tolerate all that they are being asked to do in order to get it to work. The system overhead, the wasted resources running the VM, the wasted man power of each user trying to get each rig compatible, is a much greater waste than the scientists compiling a few applications and testing. Again, my opinion. I do applaud David for taking a step for "dumbing down" the process of project selection. I have heard a lot of people express why they don't move to BOINC. Simplicity vs. complexity. By giving more options requires a lot more learning to do what should be pretty simple. Someone shouldn't have to spend hours to get it right. It should be as simple as install it and go for most users. WCG has had threads that bring up this topic. Users get bewildered because they have to set up an account with each project, then attach it to BOINC, choose their science selections, and it is all done in multiple places. Why can't it all be done from the application? Well we BOINC'ers understand it is because WCG doesn't code the application. They would get better responses and retention with their own client. That is my opinion to be clear. The BOINC client is wonderful for control freaks. But for new users it is daunting. Everyone wants the most out of their system, but they don't really want that much leg work to get to it. Since BOINC supports and infinite number of different needs projects, it really isn't for the novice. But there needs to be a novice option. Set and forget. I've supported almost every BOINC project that comes up. I don't always care about the science but have other reasons for supporting it. Many of times it is for learning. I will agree that most projects only write a published paper(s) at best. To many, this isn't really an achievement as most of us are smart enough to know that most papers go nowhere in real world practicality and is just a way to get grant/funding money. WCG is one of the few projects to ever tie their results to real world usage. I also applaud David for making it easier (theoretically) for small research groups to get access if this goes anywhere. Yes there are BOINC projects out there that claim to be set up to house multiple researches and to try and appeal to other researchers to user their platform. I've not seen many of them actually onboard any after launching other than WCG. But even WCG requires the research groups to guarantee enough work to make it worth their efforts to onboard it. Hopefully, David's vision makes it also a simple injection method with control measures for the users as I would want to know what I'm crunching. This will also help address the issue some have with requiring users to create multiple accounts using the same user name and password. Now they could have just one to connect with and be done. One concern now is whether a rogue project springs up simply for gathering email address password combos. Data breaches have been going on for years to do this as most users recycle the same login info across almost every site they use. By having an account manager controlling the login process and eliminating the need for each individually privately controlled server to manage them would be a positive thing in my eyes. |
Send message Joined: 27 Jan 16 Posts: 16 |
I do believe that there are some great ideas contained within the proposal, I just hope that some of them are introduced to Boinc itself rather than soley being for the purpose of bridging TACC with the body of Boinc users.... That is the unclear part, if he's proposing that Boinc being used in a way to bridge the gap between super computer clusters then it really doesn't matter to us individual boinc users. (except in the issue of the Gridcoin project) If it is an interface between the boinc user and the academic computing clusters, then it is of immense importance to us.... And many of us would have issues and questions centering around the who, what, when and where of what is being run on our equipment.... We want to know and we have a right to know..... It is our equimpent, isn't it? |
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.