Suggestions for Future contents/features.

Printable View

06-23-2014, 12:11 AM
Peter J. Shane

Suggestions for Future contents/features.

Whilst I consider that FreeDC is the best in all the available statistical Sites, sometimes I sit back and wonder why is certain features listed as is or why something else is not included and I tell myself Do not Rock the Boat wait until somebody fixes it. Today, I realize that unless I state my perspective, nothing will change for the others more than likely think the way I did, that is do not fix something that is not broken. or wait for somebody to fix it. The trouble is the user "Somebody" is very allusive.

Having said that this is what I would do if I could program (which I can not) just to satisfy my way of thinking but I also concede that others might not see it the same way.

1. Divide the Active Projects into BOINC and Non Boinc Projects. I believe that the majority of users follow BOINC projects although a minority are likely to follow both. From the majority point of view when searching for BOINC projects, it would be a waste of time looking up a Non Boinc Project. A Division in this field would make the task easier and less confusing..

2. Update projects location to either Active or Retired. Again, it is taking too much time to research weather a program is active or not. We should have a clear policy when a project is no longer active. For example:Mersene@Home is still listed as active after 503 days of not posting stats, Similarly Sierpinski Base5 PRP has been inactive for 719 days.

3. Food for future features. The above suggestion I think could be easily accommodated. My thoughts for the future is not so simple I am afraid. Often what appears to be simple is not.

My future thoughts for change is not so simple and I am sure will rock the boat. You might recall the days of CPU only crunching, some users were complained that it is not fair for they are not in the same playing field as those of IT Techs in charge of large corporations that the individual did not pay for yet can become number one in the world by using machines that do not belong to them. Later when GPU's came in force a similar argument pursued for those who wanted to remain on CPU only had no chance against those who owned GPU's. The competition became harder because the goal post was moved out of reach to some. In its infancy Dr David Anderson and his team tried to rectify some of the anomalies in the system with creating the so called "New Credit" system but unfortunately he created more problems than it was worth. There is another problem emerging as we speak. There are users now whose daily score accedes 100Million by using rigs created to mine bitcoins. Whilst Bitcoin Utopia can utilize those machines granting such high credits, the rest of the world Boinc users have no chance at all of catching up let alone being in the top. Such equipment is not only costly but is not available in a lot of countries making it harder still to compete. It is time to think of creating Divisions as future equipment currently being worked on is likely to be even more powerful. I am thinking something on the line of Division A would consists of those users whose out put is more than 100Million a day, starting the Division at a certain lowest point. Division B would be those who have not yet reached the lowest point in Division A but similarly output more than 10 millions or so a day. Further Division as needed including a division for corporate users who do not own the machines but are in charge of them. I am not thinking of friends who might help you with one or two machines but those busyness types who do not pay for the rigs. It is human nature to compete so let us compete on more even fields like a ladder. The higher you are attempting to claim on the ladder, the harder it is and those on the top would reflect users who have the means and can afford to be in a Division of their own.

This is just food for thought but unless we start thinking what to do in the future I doubt BOINC will remain in its present form as the days of Desktop computing is nearing the end and the cost of running equipment is increasing almost daily leaving no option but to give up on BOINC..
06-23-2014, 12:55 PM
ChertseyAl

Simplest solution is for all projects to stop granting any credit, or if they do don't export it. Then there is no need for stats comparison sites, no competion, no league tables, no need for divisions. Everyone can then just crunch what they like without fear of being judged by their performance and just enjoy the feeling that they are doing something useful for humanity. Problem solved.

BOINC is a pretty good tool for distributing work and collecting results. It's pretty bad as a basis for competitive sports like credit-whoring and badge collecting.

When I'm in charge I'm going to scrap all credit, and scrap all stats sites.

After I've got my next couple of 100K milestones and reached 25 million total credit of course ;)

Cheers,

Al.
06-26-2014, 10:19 AM
Bok

Quote:

Originally Posted by Cruncher Pete

Whilst I consider that FreeDC is the best in all the available statistical Sites, sometimes I sit back and wonder why is certain features listed as is or why something else is not included and I tell myself Do not Rock the Boat wait until somebody fixes it. Today, I realize that unless I state my perspective, nothing will change for the others more than likely think the way I did, that is do not fix something that is not broken. or wait for somebody to fix it. The trouble is the user "Somebody" is very allusive.

Having said that this is what I would do if I could program (which I can not) just to satisfy my way of thinking but I also concede that others might not see it the same way.

1. Divide the Active Projects into BOINC and Non Boinc Projects. I believe that the majority of users follow BOINC projects although a minority are likely to follow both. From the majority point of view when searching for BOINC projects, it would be a waste of time looking up a Non Boinc Project. A Division in this field would make the task easier and less confusing..

I suspect you are perhaps using the index.php page and perhaps usermain.php? Those are mixed between BOINC and nonBOINC, but honestly I never use them personally.

If you are only interested in BOINC projects then just use your userbycpid page as the jumping off point. For a list of only BOINC projects use the Projmain page which is linked on the side as 'overall' (not the best naming I would say though). The stats themselves within the database are totally separated between BOINC and nonBOINC

Quote:

2. Update projects location to either Active or Retired. Again, it is taking too much time to research weather a program is active or not. We should have a clear policy when a project is no longer active. For example:Mersene@Home is still listed as active after 503 days of not posting stats, Similarly Sierpinski Base5 PRP has been inactive for 719 days.

This is something that's asked about every now and then. Personally I'm not a real fan of separating them, though the mechanism within stats is there, I'd be more in favor of just indicating somehow that they were inactive a bit better. On the userbycpid page for instance I wouldn't want them separated ever as the retired ones still count towards the overall total and the MM's. I'm open to specific suggestions of course.

Quote:

3. Food for future features. The above suggestion I think could be easily accommodated. My thoughts for the future is not so simple I am afraid. Often what appears to be simple is not.

My future thoughts for change is not so simple and I am sure will rock the boat. You might recall the days of CPU only crunching, some users were complained that it is not fair for they are not in the same playing field as those of IT Techs in charge of large corporations that the individual did not pay for yet can become number one in the world by using machines that do not belong to them. Later when GPU's came in force a similar argument pursued for those who wanted to remain on CPU only had no chance against those who owned GPU's. The competition became harder because the goal post was moved out of reach to some. In its infancy Dr David Anderson and his team tried to rectify some of the anomalies in the system with creating the so called "New Credit" system but unfortunately he created more problems than it was worth. There is another problem emerging as we speak. There are users now whose daily score accedes 100Million by using rigs created to mine bitcoins. Whilst Bitcoin Utopia can utilize those machines granting such high credits, the rest of the world Boinc users have no chance at all of catching up let alone being in the top. Such equipment is not only costly but is not available in a lot of countries making it harder still to compete. It is time to think of creating Divisions as future equipment currently being worked on is likely to be even more powerful. I am thinking something on the line of Division A would consists of those users whose out put is more than 100Million a day, starting the Division at a certain lowest point. Division B would be those who have not yet reached the lowest point in Division A but similarly output more than 10 millions or so a day. Further Division as needed including a division for corporate users who do not own the machines but are in charge of them. I am not thinking of friends who might help you with one or two machines but those busyness types who do not pay for the rigs. It is human nature to compete so let us compete on more even fields like a ladder. The higher you are attempting to claim on the ladder, the harder it is and those on the top would reflect users who have the means and can afford to be in a Division of their own.

This is just food for thought but unless we start thinking what to do in the future I doubt BOINC will remain in its present form as the days of Desktop computing is nearing the end and the cost of running equipment is increasing almost daily leaving no option but to give up on BOINC..

Whilst I kind of agree and the new BitcoinUtopia ASIC credits are causing this now, much like GPU's did when they first started being used, it's the same situation. There is no simple, accurate way to determine whether credits came from CPU/GPU/ASIC/Other in the current framework of BOINC. I've pointed it out to the developers before that I would like to see it separated out so it can be distinguished. In my opinion if 'subproject' data was exported as a standard function of BOINC like PrimeGrid,Yoyo@Home and NFS do (which I've attempted to get put in a number of times to no avail) this could then easily be extended to 'mark' each subproject as CPU/GPU/ASIC etc and would allow this.

Unfortunately right now they don't do that and even if they did it retroactively most projects wouldn't have the data available to recreate it, it would only be going forward (though good guesses could probably be made).

Does that answer anything , some or nothing ? :) As always I'm open to specific suggestions.
07-14-2014, 12:58 PM
ChertseyAl

Because of this thread, I got curious. So I bought a dirt cheap bottom of the range GPU (the only ones that will work in my dinosaur machines) and an ASIC bitminer to play with.

The junky GPU is capable of generating about half of the RAC that my entire 'farm' can produce. And that's a bottom of the range model. Am I junking all of my dinosaurs and buying machines with powerful GPUs? Nope. Not that many projects actually have applications for GPUs, so that would drastically reduce the number of projects that I participate in.

The bitminer generates insane credit compared to my CPUs and my little GPU. Except of course it's not insane at all. It's probably about right given the performance of the device compared to anything else. Am I plugging USB bitminers into all of my hosts? Nope. I'm quite happy scraping into the top 100 participants at that project. I'll leave the 99 big boys ahead of me to do most of the work ;)

Back on topic (sort of), yes, it's made my stats pie charts look a bit odd. But I can still easily see my credit in other projects and my milestones.

So, I'm still crunching away on a project that grants dismal credit just to get my 50th 100K milestone. Meanwhile I'll probably end up leaving the GPU on Einstein, and the bitminer can raise a few cents for worthy causes. And it may indeed end up funding other BOINC projects, which would be quite cool. I'll probably end up with a billion credits overall, but still be grinding away on other projects for a few thousand credits per day :)

Final thought with reference to: "unless we start thinking what to do in the future I doubt BOINC will remain in its present form as the days of Desktop computing is nearing the end and the cost of running equipment is increasing almost daily leaving no option but to give up on BOINC" ... My little ASIC tears through over 2 Ghashes per second and consumes 2.5 watts. Imagine what an ASIC dedicated to protein folding could do for projects like Rosetta, Folding etc ... Maybe a someone should develop one as a kickstarter venture funded by bitcoins mined by ASICS :)

Cheers,

Al.