Gridcoin project Sourcefinder out of work
Since a few days, BOINC project - Sourcefinder has been whitelisted on the Gridcoin network.
While checking the latest server status, I noticed that the project has run out of work. Further analysing the daily awarded credits over the past 40 days, output has been really limited as well (see chart below).
For example the Number 1 on Gridcoinstats has a RAC output which is equivalent to 3 Intel i5-4590 processor systems, earning 400 GRC per day.
It is not my intention to criticise the fact that the Number 1 is receiving this amount GRC on a daily basis, it is the system we need to look at to avoid this from happening in the future.
To me there are 2 elements that we should evaluate before whitelisting a project:
- When a project is whitelisted, it is likely that a few hundred or more systems of Team Gridcoin will start crunching. This represents a certain credit or RAC amount. We could calculate this minimum amount based on a standard set of machines. Once the project reaches this minimum output, the first condition is met.
- The project should have a minimum output over a longer period of time, for example 30 days. Once the project reaches the 30 days with a minimum output on daily basis, the second condition is met.
When both condition are in met, we could whitelist the project.
What is your view? Should we set standards for minimum output levels and consistent availability of workunits over a longer period of time? Any other parameters?
Thanks for reading. Follow me for more BOINC & Gridcoin related articles.
The Sourcefinder debacle has raised some very good questions regarding how we add projects to the whitelist. There was a lengthy conversation on slack regarding this. We discussed it briefly at the hangout on Saturday.
I think we need to take a look at the whole process and i think @GUK is putting some ideas into a proposal on the subject.
Personally, I think we need a period of time where we test the project on the testnet, or through some other means to ensure that there will be no superblock problems when we bring the project onto production. I also think we need a stable and open channel of communication between Gridcoin and projects so we can give them lots of notice before we bombard their work with the power of the Gridcoin network. This communication should include some stats on what to expect when a project is whitelisted. This way admins will be better prepared with WU. I also think there should be a large media push for each project we add. By this I mean there should be at least a week of announcements including write-ups on the project, its purpose, its admins, and its history. This would ensure that no one with a huge amount of processing power can monopolize the WU from the new project.
These are just some random ideas though. I'm not sure what is practical and what is wanted.
Thank you for posting this. I wonder what Dutch has to say about this? He has already received thousands of GRC and will receive thousands more, despite doing no work for it. This seems very unfair to all of the people who are spending money on electricity and who would only receive 400 GRC in maybe one or two months.
That being said, I just looked on their website and saw this message:
"There's certainly been a rather large uptake all of a sudden.
I'll be taking measures to allow me to dispatch more work during the week, rather than only on Wednesdays."
So maybe that's the problem. Nevertheless, we should keep an eye on this.
Forgot to attribute that quote - it's from Sam, the project administrator (who also holds a bunch of other titles).
Here's the link:
https://sourcefinder.theskynet.org/duchamp/forum_thread.php?id=252
The biggest problem with small projects is the small amount of hardware behind them. I was reading posts on the official message boards and it seems as though the admin of this project is treating it like a hobby and only had access to the server on Wednesdays. Even then the server seemed small and was only capable of 30GB of WU's.
Before white listing a project the admins of the project should be consulted with and an action plan come up with in order to handle the additional influx of users.
My recommendation would be for the users who received a large amount of Gridcoin from this project for little work done is to donate it to the project admin of Sourcefinder so that they may be able to put the money to good use in upgrading the back end hardware of the project. Of course it probably wont happen.