results not needed

[Madpoo]

Originally Posted by TheJudger

A whole bunch of "result no needed" today. Many (all?) of them reported by Kevin Jaget earlier today? Feels like wasting resources!

That user doesn't seem to do TF work on assigned exponents. Last time they checked in TF results for an *assigned* exponent was back in October 2012.

They do get assignments for other types of work (P-1 and LL tests).

So... not sure what's going on there that they're just doing TF work without checking if it's already assigned to someone for TF (like GPU Factoring, for instance).

[chalsall]

Originally Posted by Madpoo

That user doesn't seem to do TF work on assigned exponents. Last time they checked in TF results for an *assigned* exponent was back in October 2012.

To put on the table, I do my very best to facilitate work.

It pisses me off when people "game" the system.

[Gordon]

Originally Posted by chalsall

To put on the table, I do my very best to facilitate work.

It pisses me off when people "game" the system.

Remembering of course what we always say "your equipment, your money, do what you like". As you yourself have stated on here many times before.

[kladner]

Originally Posted by Gordon

Remembering of course what we always say "your equipment, your money, do what you like". As you yourself have stated on here many times before.

True. But doesn't that suppose that people will work within the system so as not to waste efforts?

[LaurV]

Originally Posted by chalsall

It pisses me off when people "game" the system.

It pisses me off when people "game" the system without making any profit*, just to annoy the others.

-----
* Milking the system is ok, IMO...

[TheJudger]

Hi,

again good to know that this is cause by "human mistakes".
I'm not sure how much the individual waves are ahead of each other but perhaps there should be a delay when someone unassigns some work until they are given to another user?

Oliver

[TheJudger]

Hi,

this time user "JC":

no factor for M73730911 from 2^73 to 2^74 [mfaktc 0.21 barrett76_mul32_gs]
no factor for M73730911 from 2^74 to 2^75 [mfaktc 0.21 barrett76_mul32_gs]
no factor for M73730843 from 2^73 to 2^74 [mfaktc 0.21 barrett76_mul32_gs]
no factor for M73730843 from 2^74 to 2^75 [mfaktc 0.21 barrett76_mul32_gs]
no factor for M73730977 from 2^73 to 2^74 [mfaktc 0.21 barrett76_mul32_gs]
no factor for M73730977 from 2^74 to 2^75 [mfaktc 0.21 barrett76_mul32_gs]
no factor for M73730863 from 2^73 to 2^74 [mfaktc 0.21 barrett76_mul32_gs]
no factor for M73730863 from 2^74 to 2^75 [mfaktc 0.21 barrett76_mul32_gs]
no factor for M73730981 from 2^73 to 2^74 [mfaktc 0.21 barrett76_mul32_gs]
no factor for M73730869 from 2^73 to 2^74 [mfaktc 0.21 barrett76_mul32_gs]
no factor for M73730869 from 2^74 to 2^75 [mfaktc 0.21 barrett76_mul32_gs]

I think there should be either a longer registration time for assignments or a (longer) grace time before reassign.

Oliver

[chalsall]

Originally Posted by TheJudger

I think there should be either a longer registration time for assignments or a (longer) grace time before reassign.

Awww, sh*t...

Sorry about that. So everyone knows the system no longer expires candidates _unless_ the assignment is over 30 days old ***AND*** no new assignments have been given nor reported in that period ***AND*** the assignment has not been extended. This has been the case for almost a year now.

I will increase this expiry period to be 60 days, although the notices and warnings will continue to claim it's 30 days.

Also please note that there is a warning (in red) given on the "Current Assignments" page if any candidates are at risk of being expired, and said candidates are also rendered in red.

I don't really know how much more I can do short of completing the assignments myself so others don't take the risk (which I sometimes do).

[TheJudger]

Hi Chris,

for GIMPS it doesn't really matter if you or me doing duplicate work. Just lets try to avoid (read: reduce to a feasible minimum) duplicate work.

Oliver

[TheJudger]

Originally Posted by chalsall

Also please note that there is a warning (in red) given on the "Current Assignments" page if any candidates are at risk of being expired, and said candidates are also rendered in red.

When I report my results once a week I check my current assignments a hour later and there are often little to no assignments older than 10 days.

Oliver

[chalsall]

Originally Posted by TheJudger

When I report my results once a week I check my current assignments a hour later and there are often little to no assignments older than 10 days.

Understood and appreciated. Not all workers are as vigilant.

To put on the table, I don't think anyone is being intentionally disruptive. With the introduction of fetching and submission automation many simple "fire and forget".

Even with Primenet's LL / DC / P-1 efforts this happens. Someone makes a promise of completion but then doesn't deliver; the candidate is then reassigned and then the original assignee completes.

It is very difficult to manage independent actors....