Just posted this over at BOINC....:
Thanks - I was being alittle provocative with the set top box thing, but do you want to take a bet on when a Cable box has the power to do BOINC and CPDN? 2 years? 5 years? and how many people will have desktop computers by then - a smaller % than have a set top box, I'd guess. So when does the computing power in a household shift from desktop plus laptop to multiple laptops plus....other devices?There are 100 million US households alone, and it's getting those people added to the 50,000 BOINC CPDN users which seems to be the challenge, rather than getting the 50,000 slightly happier.I applaud the removal of unsupported hardware - but if you go further and consider the absolute newbie who is learning about this stuff via facebook on the brand-new laptop they got for their college course in Comparative History, just seeing a website that talks about Linux is enough to have them say "not for me". The dedicated can have all that stuff - but I'd hide it well away from the millions of people you are trying to attract.A 1% greater uptake amongst 10,000,000 people is far more important than a 25% performance boost for every one of the existing 50,000......In my mind this is no longer *mainly* a technical issue - it is a communications issue. There are plenty of technical issues and always will be, but you can simply ignore 50% of them and spend that time and money on communications / user interface and have a greater positive effect on the total number of cycles dedicated to life-changing research, IM(very)HO.For example:- You got a Mac - sorry.- You got a non-Intel processor - sorry (and no I don't work for Intel, and I don't like anti-competitive practices, but is the point to spend time on being even-handed or on getting computer cycles?)- You are on anything other than Windows version X - sorry (I know this is heresy, but consider what you're actually trying to achieve here - zetaflops, right?)- Meanwhile, since you have the type of system we support it is all very simple ...click this button, say "Yes" to that security question, pick your projects, choose "low overhead" or "high overhead", link with your Facebook account - and you're saving the world!!!If you want help - here is the ONE place you go to, and it is all very simple and hierarchical - no more than 8 or 9 choices on each page (human brain starts starts processing with 60% of the efficiency if the list goes over 9.....)Consider the above for a minute - could this increase the total number of cycles dedicated to these projects?Could you also do that other stuff (linux, GPU processing, Athlon, etc etc) - later, and under the hood?If so why not do it?Thanks for listening - again, I thing this is a truly incredible technical achievement, and I really like the new website. Maybe GridRepublic and Progress Thru Processors actually IS the answer I'm talking about and where I am right now IS the "under the hood" I'm talking about. If so - send everyone there with a footnote that if they are technically ambitous you'd love to see them on this site?To come full circle and also to partially disagree with myself - working on the technology will get the increasing % of installed-based computing power in set-top boxes and cellphones etc doing this stuff, and it's the relentless pursuit of the user-experience that will get the *millions* of vaguely-interested computer users signing up for this stuff. In summary I believe everything should work backwards from the question. "Why isn't 50% of the idle computer power in the world running BOINC projects? "Very interested to hear your commentsQuentin
Hi Quentin,
That is a very interesting and thought provoking post indeed. I too think that getting people involved is partially about communication and presentation, but it is also about perception and trust. We can introduce volunteer computing to everyone but if they do not trust the source of the information or the program then they will never install it on their computer. Also if it as you said too difficult or too technical people will never want to participate or try to understand.
Also something we should try to achieve is the idea of the default. Default installation with the program already available on the computers and other processing devices as the studies show that most users do not change their defaults. This would allow us to surpass the 1% easily I think. This would also probably be the easiest way to get onto set top boxes and other indirect interfaced boxes. The phones and mobile devices would be accessible via an interface, but default installation is the easiest way to be on these.
My vision is a process that is as simple as just pick your projects and hit start and everything would setup automatically. Where you can monitor your computer, change projects or have us manage or recommend projects for you. Creating a socially active website where you can interact with other members in many ways to get involved actively as well as passively. Where those who want to get deep in the science and technology can, while enabling people who just think it is a good idea to be able to come and stay here let me help, but I do not want to mess with the settings. Volunteer computing really enables the citizen the chance to say and vote what research is important to them and help advance it directly.
I was in a similar position questioning what is the right path to get the more computer user involved not too much more than two years ago and I came to the conclusion that we need a combined effort and image for volunteer computing for BOINC projects. Between actually talking with people about the issue and my own discussion in the BOINC community I realized that GridRepublic held the key ingredients to establish the trust, ease of use, and simplicity desired by the mainstream user. It is a nonprofit, it is an account manager who's interface was easy to use (yes, it can be easier and we are working on that but we are working with limited resources for development), and the people wanted to get more people involved because they really believe in the power of volunteer computing.
Since I joined we have launched the community section on our website and established our first public sponsor. I hope we can accomplish more this year. The more people willing to help by donating, finding funding, writing documentation, and many other areas the faster we can focus resources on the areas of development needed. We have many ideas and limited resources to work with.
We would love to have your support and involvement in the advancement of the volunteer computing cause with GridRepublic. If you would like to get more involved let us know.
Thanks for the post - and more later, but I can write OK - so my be able to help with manuals. Let me know what I can do.
My current (disposable) email is quentinsub-01 at yahoo.com then I can send you a real email address, and I am on Facebook too, of course.
All the best
Quentin
