I am new to this project and I have noticed the results table plus a database where the project-generated results are stored. What I could not find is which part of these results was produced by which user? Could you please ensure that each contributor finds his/her results somewhere? One way to accomplish this is to simply extend the account webpage with a section tabulating the individual results as it is done in many other distributed computing projects. Another one would be to extend the existing database.

Michael.

---

President of Rechenkraft.net

Could you point me to an example of an extended account web page showing individual contributions? I use the standard BOINC code and I don't think that is a built in feature. If I see how other projects are doing it, I might be able to incorporate such a feature, although I hesitate because that makes it really hard to update the web code with the latest BOINC version.

Also note that the search algorithms by nature find multiple representatives for the same field. The search usually finds each field 10 times over. So we can't associate a single user with any given field, but rather a list of users, and that could get ugly. To make things worse, we wouldn't know the list of users until all the results in a batch were collated, so the list cant be updated in real time. I suggest a better metric would be a count of the number of fields found per user - the assimilator could update this value in the database each time a user returned a result and then this could be displayed in various places such as the account page (going forward we could even create new badges based on this number). Does that sound reasonable?

In fact, the individual user findings of many mathematics projects are show e.g. even on external Stats pages such as those of Free-DC.
Ad hoc examples are Primegrid, Amicable Numbers and Yoyo@home (my team / organization colleague: You may contact him directly and even ask for code, if you like).

Also, Rakesearch has implemented a results section on the user's account page upon my suggestion.
Since Yoyo@home (for certain reasons) uses "older" BOINC server software while the other projects use the recent one, it is clear that the feature I suggested is implementable.

I don't think a real-time update of the user findings is mandatory, so your suggestion sounds reasonable. The ultimate goal is just to ensure that the project becomes a bit more interesting to potential contributers. And also, I think it is standard in mathematics to name those who helped to achieve scientific findings, right?

Michael.

---

President of Rechenkraft.net

Ok, I will look into this. It may take a week or two, as I don't know exactly what the change will entail yet and I'm pretty busy with my "day job".

..very nice.

:)
Michael.

---

President of Rechenkraft.net

I've completed the first stage of this - the necessary database changes to store the field counts for each user. I even updated the assimilator to update the database when it assimilates the results.

The septic search naturally finds many more fields than the decic search. This turns out to put a heavy burden on the assimilator. For this reason, I will tabulate the septic results after that search completes in about 30 to 60 days.

The next step is to update the web code to put the user field counts on various pages. I will start that this weekend.

Every user can now find their individual field counts on the top participants page and also on their accounts page. Let me know if there are any other pages where you would like to see this extra information and I will try to add it.

For now the septic count can be ignored since that count is currently disabled in the assimilator. The septic counts currently listed are from a 4 hour window when the assimilator was automatically updating them. After I noticed the assimilator had a big backlog of results to process, I disabled the database updates. I will try to find time to write an update script that I can run periodically to update the septic counts; that way we wont have to wait until the end of the septic search to see the counts.

Good progress! ;-)

Michael.

---

President of Rechenkraft.net

I now have a script for updating the septic counts. I ran it last night and will run it every 3 or 4 days as blocks of the septics get finished.

Ver good. I just figured that I now have 118 septic counts.

Michael.

---

President of Rechenkraft.net

I now have a script for updating the septic counts. I ran it last night and will run it every 3 or 4 days as blocks of the septics get finished.

What about the decic counts? Is there a script, too?

Michael.

---

President of Rechenkraft.net

As I tried to explain above, the decic counts are done in real time by the assimilator every time a result is returned.

The two searches are fundamentally different in that the decic search targets those fields ramified over a small set of primes. The septic search finds all degree 7 fields with discriminant below a fixed bound. For this reason we find much fewer decics, and therefore it is less of a burden on the server to update them in real time.

OK, thank you for the quick explanation.

Michael.

---

President of Rechenkraft.net

What about the bounded app?
Most of my points come from this application.


[OT]
As you may have notice I´m back to my personal prime search and factoring stuff. But, I will come back at some time. I need to be under Top 10 from my country; atm I´m at position 11 and needing a bit more than 2 millions points to reach TOP10.
[/OT]

What about the bounded app?
Most of my points come from this application.


[OT]
As you may have notice I´m back to my personal prime search and factoring stuff. But, I will come back at some time. I need to be under Top 10 from my country; atm I´m at position 11 and needing a bit more than 2 millions points to reach TOP10.
[/OT]

That app completed its search over 2 years ago, so I had no plans to go back that far. And I dont even have the db archive files any more which are needed for associating users with fields. Sorry!

I have one more suggestion regarding the results details:
At present the user account just shows the number of identified results per category. Since there appears to be even a database of the identified results, it would actually be nice to not only know the number of contributions but to also link to these in the database such that the user can actually take a look at all the results found.
Of course it would also be nice to even indicate the finders name in that database.

Michael.

---

President of Rechenkraft.net

So there are several problems with doing this. The database is only saving the counts of the fields found. There is a good reason for this - the septic search has already found over 10 million fields of which about 1 million are unique. So each of these 1 million unique fields would be associated with many users. To store all that information in the database would be burdensome and not worth the trouble. With searches of this nature, I would argue what's important is what portion of the final results can be attributed to a particular user, and this is what the counts do. I think what you are suggesting makes more sense if the objects being found are rare, such as Mersenne primes or amicable pairs.