The idea of spendable cobblestones

Message boards : Number crunching : The idea of spendable cobblestones
Message board moderation

To post messages, you must log in.

1 · 2 · 3 · Next

AuthorMessage
Profile Janus
Volunteer moderator
Project administrator
Avatar

Send message
Joined: 16 Jun 04
Posts: 4563
Credit: 2,097,282
RAC: 0
Message 84 - Posted: 18 Feb 2005, 17:07:57 UTC
Last modified: 18 Feb 2005, 17:09:42 UTC

BURP, unlike other projects, will need a currency - something that you get when rendering other people's sessions and spend when having your own session rendered.

For all BOINC projects you get Cobblestones for the work you do. Now, in BURP it would be a good idea to introduce the Spendable Cobblestone.

When you do work, both your cobblestone amount and spendable coblestone amount increases (with the same value). You can then spend some of your spendable cobblestones on rendering an animation.

An example:
- A newcommer has 0 CS and 0 SCS
- The newcommer renders 10 frames for some animation and now has 28.2 CS and 28.2 SCS
- The newcommer gets all excited about this and starts his own render of 3 frames at the cost of 1 SCS per frame. He now has 28.2 CS and 25.2 SCS

-------
You cannot render anything if you do not have any SCS - this is to avoid people creating accounts and rendering without contributing something in return.
-------
When submitting an animation for rendering you can decide what priority it should be given.
Low priority renders will cost almost no SCS, high priority renders will be more expensive (SCS wise) but will be rendered faster.
-------
SCS, just as CS, are non-tradeable. (To avoid abuse)
-------
From time to time official competitions can be held where the winner will get XXXX SCS
-------
Anyone have any further ideas or comments to all of the above stuff?
ID: 84 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
benstabler

Send message
Joined: 8 Nov 04
Posts: 121
Credit: 601
RAC: 0
Message 87 - Posted: 18 Feb 2005, 19:20:37 UTC

"The newcommer gets all exited about this"

you spelt excited wrong.



Other than that, the only thought I have is that there seems no point in the non-spendible cobblestones. Why have them? I think it is just a point for confusion.
ID: 87 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Janus
Volunteer moderator
Project administrator
Avatar

Send message
Joined: 16 Jun 04
Posts: 4563
Credit: 2,097,282
RAC: 0
Message 91 - Posted: 20 Feb 2005, 12:29:15 UTC - in response to Message 87.  

> Other than that, the only thought I have is that there seems no point in the
> non-spendible cobblestones. Why have them? I think it is just a point for
> confusion.

Simply to have some total number of what you have contributed, so that people can compare their contributions - compete and stuff.
You can't do that with SCS alone because one could have spent all of them even though he/she has crunched thousands of frames.
Furthermore the CS is a BOINC-wide standard, so users from other BOINC projects (including the global statistics websites) will expect it to be here.
ID: 91 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
benstabler

Send message
Joined: 8 Nov 04
Posts: 121
Credit: 601
RAC: 0
Message 92 - Posted: 21 Feb 2005, 18:34:54 UTC

A benifit with having only SCS is that you can see your stats in the boinc client, unless you're planning on writing code for another column. At the moment it would only show CS, and this isn't terribly useful if you want to know if you have enough SCS for the render you want to do.

In my opinion, we dont need CS, but if you keep them then I suggest that they are less important and should be scaled back.

later
lightning
ID: 92 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Janus
Volunteer moderator
Project administrator
Avatar

Send message
Joined: 16 Jun 04
Posts: 4563
Credit: 2,097,282
RAC: 0
Message 94 - Posted: 22 Feb 2005, 18:23:21 UTC - in response to Message 92.  

> A benifit with having only SCS is that you can see your stats in the boinc
> client, unless you're planning on writing code for another column. At the
> moment it would only show CS, and this isn't terribly useful if you want to
> know if you have enough SCS for the render you want to do.
>
> In my opinion, we dont need CS, but if you keep them then I suggest that they
> are less important and should be scaled back.

Yes, the webpage will show you the needed amount of SCS when you submit a session. So I don't think this will be a problem. You will also be able to right-click the project in the client to get the information from there.

The "pricing" scheme I had in mind:

A render with standard priority:
100 SCS for the entire render - no matter length or complexity

A render with high priority (will skip the standard priority render que):
5 SCS / part *

* One part is usually a frame, but if subframe rendering is used a frame can consist of several parts.
ID: 94 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Kent Mein

Send message
Joined: 23 Feb 05
Posts: 1
Credit: 0
RAC: 0
Message 95 - Posted: 23 Feb 2005, 14:44:55 UTC

While I agree that something will probably be needed at some time in the future.
I think that making it so new people can not submit jobs will just turn people
away from the project.

If you look at bittorrent new users can still do stuff they just have sucky
privlages until they have done some work.

I also think it will be hard to find jobs to run at least starting out. I would
say you need to run it without restrictions for awhile to see if there is actually a problem with too much work.

Also, where are the patches to blenders cvs that need to be applied for burp?
ID: 95 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Janus
Volunteer moderator
Project administrator
Avatar

Send message
Joined: 16 Jun 04
Posts: 4563
Credit: 2,097,282
RAC: 0
Message 97 - Posted: 23 Feb 2005, 17:12:13 UTC - in response to Message 95.  

> While I agree that something will probably be needed at some time in the
> future.
> I think that making it so new people can not submit jobs will just turn people
> away from the project.
>
> If you look at bittorrent new users can still do stuff they just have sucky
> privlages until they have done some work.
>
> I also think it will be hard to find jobs to run at least starting out. I
> would
> say you need to run it without restrictions for awhile to see if there is
> actually a problem with too much work.

Yes, I guess you are right. So the standard priority renders will be free to start out with.

> Also, where are the patches to blenders cvs that need to be applied for burp?

I'll answer this question here soon...
ID: 97 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Razorirr

Send message
Joined: 15 Jan 05
Posts: 21
Credit: 1
RAC: 0
Message 129 - Posted: 3 Mar 2005, 4:49:59 UTC

ideas on credit you can go with whatever or none of them makes no difference to me
1) if you had scs and cs put one of them in as average credit. it would be easyer then getting the developers of Boinc and the other projects to change their stuff.
2) have scs not be in the credit slots since it would mess with the stats page designers. but have it somehow be built as a user unchangeable thing into our prefs.
3) dont have cs since the majority of us will never submit anything and coud just use our scs as cs since it will only go up.
4) make your own version of boinc thats compatiable with the other projects. ill test it for you on windows. predictor at home actually did this. they came out with windows 3.22 while the boinc people at berkeley were cooking up 4.xx
ID: 129 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Razorirr

Send message
Joined: 15 Jan 05
Posts: 21
Credit: 1
RAC: 0
Message 130 - Posted: 3 Mar 2005, 4:56:06 UTC

i just realized i could be totally wrong about three
ID: 130 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Janus
Volunteer moderator
Project administrator
Avatar

Send message
Joined: 16 Jun 04
Posts: 4563
Credit: 2,097,282
RAC: 0
Message 146 - Posted: 3 Mar 2005, 20:24:17 UTC - in response to Message 129.  

I think I will make the SCS visible only on the website - because the only time you need to know about it is when you need to render something. So it will be shown on the pages related to that and on the user account page.

SCS will not be exported in XML stats since that would mess up the format.
ID: 146 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Peter M. Nielsen
Project donor

Send message
Joined: 14 Mar 05
Posts: 130
Credit: 307
RAC: 0
Message 208 - Posted: 15 Mar 2005, 11:50:44 UTC - in response to Message 146.  

Well, the idea is pretty good - one hitch though. If you have a lot of frames you want to render, you might need to wait half a year before you have enough credit. Also you need to make it possible to buy SCS otherwise you will never get the project running - everybody starts with = credit so no-one can actually submit work

- Peter
> I think I will make the SCS visible only on the website - because the only
> time you need to know about it is when you need to render something. So it
> will be shown on the pages related to that and on the user account page.
>
> SCS will not be exported in XML stats since that would mess up the format.
>
ID: 208 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Keck_Komputers
Avatar

Send message
Joined: 6 Mar 05
Posts: 94
Credit: 1,378,284
RAC: 0
Message 216 - Posted: 16 Mar 2005, 10:07:10 UTC

I think just using the total credit would work, since most people wouldn't spend it and the ones that do most likely would not care if they had theirs lowered. It may be an idea to start everyone out with 1000 CS or something depending on how much credit the average rendering job costs. That should get the stats freaks on board real quick too.

Making the CS buyable could be a good idea. It would provide some real income for the project and may also get some stats junkies to support the project without actually doing any work too.

Using the RAC spot or modifing the client is probably not a good idea. Because changing the standard functions may cause problems later when updates happen.
BOINC WIKI

BOINCing since 2002/12/8
ID: 216 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Peter M. Nielsen
Project donor

Send message
Joined: 14 Mar 05
Posts: 130
Credit: 307
RAC: 0
Message 218 - Posted: 16 Mar 2005, 10:22:58 UTC - in response to Message 216.  

> I think just using the total credit would work, since most people wouldn't
> spend it and the ones that do most likely would not care if they had theirs
> lowered. It may be an idea to start everyone out with 1000 CS or something
> depending on how much credit the average rendering job costs. That should get
> the stats freaks on board real quick too.
>
> Making the CS buyable could be a good idea. It would provide some real income
> for the project and may also get some stats junkies to support the project
> without actually doing any work too.
>
> Using the RAC spot or modifing the client is probably not a good idea. Because
> changing the standard functions may cause problems later when updates happen.
>

Well, I second to most of it, but I still think companies etc. should have the option to "buy" some computerpower by buying copplestones...

- Peter

ID: 218 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Janus
Volunteer moderator
Project administrator
Avatar

Send message
Joined: 16 Jun 04
Posts: 4563
Credit: 2,097,282
RAC: 0
Message 222 - Posted: 16 Mar 2005, 12:37:38 UTC - in response to Message 218.  
Last modified: 16 Mar 2005, 12:38:27 UTC

> Well, I second to most of it, but I still think companies etc. should have the
> option to "buy" some computerpower by buying copplestones...

Due to licensing issues we do not (yet) do commercial renders. That includes paying for cobblestones.

Recently one of the projects we rely on changed their license to a truly opensource license, so the licensing problems are becomming less of a concern now. However, I don't want to have anything to do with administrating money from users - at least untill the project has been live in a couple of months. Any such system would have to be completely automated.

Do anyone know a way to do such a thing without spending hours and hours administrating the money transfers?
ID: 222 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Peter M. Nielsen
Project donor

Send message
Joined: 14 Mar 05
Posts: 130
Credit: 307
RAC: 0
Message 223 - Posted: 16 Mar 2005, 13:20:15 UTC - in response to Message 222.  

> Due to licensing issues we do not (yet) do commercial renders. That includes
> paying for cobblestones.
>
> Recently one of the projects we rely on changed their license to a truly
> opensource license, so the licensing problems are becomming less of a concern
> now. However, I don't want to have anything to do with administrating money
> from users - at least untill the project has been live in a couple of months.
> Any such system would have to be completely automated.
>
> Do anyone know a way to do such a thing without spending hours and hours
> administrating the money transfers?

I agree with you on the aspect of not having to implement a money transfer before alpha and maybe beta testing is complete. Anyhow - the way I suppose the system should be set up is to have the SCS-credit in a seperate table in the database and on a daily/hourly basis you can update the table from boinc (by doing a diff) and payment. The payment could be a paypal-solution where companies and users pay and have their userid included. You then use paypals gateway solution which invokes a script on your site which credits the scs-table. You don't have to do anything manual this way. If you are interested I could set up a framework for it....

- Peter
ID: 223 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Janus
Volunteer moderator
Project administrator
Avatar

Send message
Joined: 16 Jun 04
Posts: 4563
Credit: 2,097,282
RAC: 0
Message 225 - Posted: 16 Mar 2005, 13:56:00 UTC - in response to Message 223.  

> I agree with you on the aspect of not having to implement a money transfer
> before alpha and maybe beta testing is complete. Anyhow - the way I suppose
> the system should be set up is to have the SCS-credit in a seperate table in
> the database and on a daily/hourly basis you can update the table from boinc
> (by doing a diff) and payment.

Actually it wouldn't be difficult to do that "live".

> The payment could be a paypal-solution where
> companies and users pay and have their userid included. You then use paypals
> gateway solution which invokes a script on your site which credits the
> scs-table. You don't have to do anything manual this way. If you are
> interested I could set up a framework for it....

It seems that such a framework already exists. I downloaded it and tested it out. It looks very interesting. It even looks like our website layout can be used troughout the payment process - which was exactly what I was looking for: an integrated sollution.
ID: 225 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Peter M. Nielsen
Project donor

Send message
Joined: 14 Mar 05
Posts: 130
Credit: 307
RAC: 0
Message 226 - Posted: 16 Mar 2005, 14:05:38 UTC - in response to Message 225.  

> It seems that such a framework already exists. I downloaded it and tested it
> out. It looks very interesting. It even looks like our website layout can be
> used troughout the payment process - which was exactly what I was looking for:
> an integrated sollution.

Great, just to be curious and it is a bit off topic - who are the developers/contributors behind this project?

- Peter
ID: 226 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile fred

Send message
Joined: 30 Oct 04
Posts: 13
Credit: 426
RAC: 0
Message 227 - Posted: 16 Mar 2005, 16:20:03 UTC

the spendable cobblestones seems a very good idea ! but i think that there should be a free service for free. Thoses projects would be rendered only if there is no other projects of higher priority running.

What could be nice too, is being able to spend cobblestones "live" as soon as you get them you can spend them to increase the priority of the work that hasn't been rendered yet :)
ID: 227 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Kasmok
Project donor

Send message
Joined: 9 Mar 05
Posts: 18
Credit: 19,038
RAC: 0
Message 229 - Posted: 16 Mar 2005, 23:59:42 UTC

I am wondering if there will be a need for "spendable cobblestones." In general there are tens of thousands of computers running Boinc. Some of which will participate in this project to yield a huge amount of processing power.

First, I am curious if there is any projection for the rendering demands that will be generated by the public? Will they be able to keep the system busy? IMP.org's renderer is almost always idle (and they have only a few dozen renderers.)

Second, I suppose most Boinc participants, like myself, are strictly interested in data crunching and would never render anything.

Lastly, it would seem to me that the greatest way to get data crunchers is to ensure that the processor time is spent on the best possible projects.

P.S. Great Boinc poject! :)

ID: 229 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Janus
Volunteer moderator
Project administrator
Avatar

Send message
Joined: 16 Jun 04
Posts: 4563
Credit: 2,097,282
RAC: 0
Message 230 - Posted: 17 Mar 2005, 0:58:51 UTC - in response to Message 229.  
Last modified: 17 Mar 2005, 1:06:38 UTC

> Great, just to be curious and it is a bit off topic - who are the
> developers/contributors behind this project?

Have a look at the contributor list

> the spendable cobblestones seems a very good idea ! but i think that there
> should be a free service for free. Thoses projects would be rendered only if
> there is no other projects of higher priority running.

I thought about that but came to the conclusion that it is better to use "Free +1" in order to avoid people that don't even install the client to have access to the resources.

BURP will always be "for free" in the way that as long as you help others you will be able to get your own stuff rendered.

I'm still considering which "prices" to use and I guess this will have to be adjusted several times during both alpha and beta since it really has to be based on how much work is to be done compared to how much work can be done by BURP. In other words: if the system is completely overloaded the price in SCS will rise and vice versa.

> What could be nice too, is being able to spend cobblestones "live" as soon as
> you get them you can spend them to increase the priority of the work that
> hasn't been rendered yet :)

Noted, a good idea. But don't expect it to be there until the final release.

> I am wondering if there will be a need for "spendable cobblestones." In
> general there are tens of thousands of computers running Boinc. Some of which
> will participate in this project to yield a huge amount of processing power.

Noone knows. Perhaps we discover that the amount of CPU power far outdoes what is needed. Perhaps it is the other way around... Time will tell. I just want to be prepared and the best way to do that is to make the SCS system. It can then always be disabled if it isn't needed.

> First, I am curious if there is any projection for the rendering demands that
> will be generated by the public? Will they be able to keep the system busy?
> IMP.org's renderer is almost always idle (and they have only a few dozen
> renderers.)

This is some of the material that will be collected during the alpha test (when people are allowed to actually submit and render animations).
As far as I know IMP.org is used in a much different way than BURP will. Basicly animations are submitted to the network by a central team (the project leader actually I guess). With BURP anyone will be able to render their animation at any time. No need to wait for a project leader to pick it up and submit it.

My guess is that BURP is going to be a bit like IMP.org but with a lot more animations and with the user in control.

Keeping the system busy is not a goal we want to pursue. If there is no work there is no work.

> Second, I suppose most Boinc participants, like myself, are strictly
> interested in data crunching and would never render [ie. submit] anything.
> Lastly, it would seem to me that the greatest way to get data crunchers is to
> ensure that the processor time is spent on the best possible projects.

Yes or simply render everything and have a good system for categorizing and rating it afterwards. This is something that I'm working on at the moment.
ID: 230 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
1 · 2 · 3 · Next

Message boards : Number crunching : The idea of spendable cobblestones