Working on the 28,000,000-30,000,000 range

[chalsall]

Originally Posted by tha

People who reserve exponents as "ANONYMOUS" and want to distinguish themselves could enhance the profile of their assignments by at least filling in a computername. That would keep me from considering an exponent.

I don't entirely disagree.

For the record, 28550909 was "captured" by GPU72, and then made available to our authorized workers. A certain someone then reserved it via the GPU72 manual assignment page, but didn't "claim it".

This actually causes some time consumption for us to clean up 60 days later...

I've recently been thinking of removing DC and LL assignments from availability from the Manual Assignment pages (on GPU72), and only allowing such assignments via the Proxy to ensure the candidate will be immediately "claimed" by the user, and said user's machine, so we all know who we're dealing with.

Thoughts?

[chalsall]

Originally Posted by kracker

That does not guarantee 100% that he/she will not pick it up again.

No, it doesn't.

But if the work is done (e.g. DC), it won't be picked up again; the work will be done.

Another example... look at "CaptainEntropy"; 58239631 being held-up by "Mommy's Laptop", even though this TFing work has long since been done....

[tha]

Originally Posted by kracker

That does not guarantee 100% that he/she will not pick it up again.

Correct. Not 100%, but reasonable close in my opinion.

Unfortunately there are still clients around that request assignments and abandon them directly after. The server is mostly able to steer them away from the sensitive ranges, but apparently these assignments slipped through or at least were not able to distinguish themselves from abandonded assignments in any of the many ways.

Another user (JH) made a lot of similar assignments in the 26M range except that they were worked on, and the "days to go" value of the about 40 assignments never became very negative. So I was able to avoid wasting resources on them. A computer (or deamon) name registered with an assignment would immediately identify an assignment as likely to be completed.

[chalsall]

Originally Posted by tha

So I was able to avoid wasting resources on them. A computer (or deamon) name registered with an assignment would immediately identify an assignment as likely to be completed.

If I may then ask you, what exactly is your personal criteria for what some call "poaching"?

[tha]

Originally Posted by chalsall

Thoughts?

I do not have a GPU so I am not familiar with the difference between reserving and claiming an exponent for assignment using that form. Whatever change to the procedures is needed to prevent this ambiguity will be appreciated.

It now looks as if someone who reserved them might be working on them off line and thinking he has done what is needed to claim them. That would make my resources wasting their time. I still consider that unlikely though since it would be a terrible slow GPU doing this. Or someone is saving up results and plans to upload them as a bunch, unaware or insufficiently aware of the time limit.

[tha]

Originally Posted by chalsall

If I may then ask you, what exactly is your personal criteria for what some call "poaching"?

In the old v4 era abandonded assignments from mostly overclockers were a plague. In that time I tried to filter out the worst to keep the trailing assignments going.

Since the v5 server came online I stopped doing that since the preferred assignment rules recycle them automatically.

Two months ago I spotted some leftovers and started to ring some bells and pull some strings to get things going again.

Poaching in my opinion is taking and finishing an assignment that someone else is still working on in a reasonable way.

The wiggling room is therefore in what is reasonable. What I consider unreasonable is not distinguishing the assignment from abandonded assignments through reports to the server, or claiming so much more time that major milestones are going to depend upon their completion.

[lycorn]

@tha:
The way to "claim" such an assignment is simply to connect to the server to update the "Days to go" (the GPU72 server has already provided us with the key)- Then the exponent starts appearing in the Primenet server pages as being assigned to us
Also note that exponents obtained via GPU72 to be LL/DCed are not necessarily done on a GPU.

[chalsall]

Originally Posted by tha

The wiggling room is therefore in what is reasonable. What I consider unreasonable is not distinguishing the assignment from abandonded assignments through reports to the server, or claiming so much more time that major milestones are going to depend upon their completion.

And you get to decide this, instead of the community, why exactly?

[tha]

Originally Posted by chalsall

And you get to decide this, instead of the community, why exactly?

I feel the community has approved this by faciltating and endorsing this through feedback. The only other way to ensure progress on the mentioned fronts is by enforcing more strict communication between the server and all clients, which would limit some users in their specific usage of the software.

It is a matter of balancing the risk of an exponent being done twice which would waste resources and the risk of an exponent not being done at all or desired results arriving undesirable late.

[c10ck3r]

Originally Posted by chalsall

And you get to decide this, instead of the community, why exactly?

Because the community is overly conservative in its methodology for recycling exponents that are not being worked on. (and this is coming from what some would consider an extreme conservative)...

[chalsall]

Originally Posted by tha

I feel the community has approved this by faciltating and endorsing this through feedback. The only other way to ensure progress on the mentioned fronts is by enforcing more strict communication between the server and all clients, which would limit some users in their specific usage of the software.

It is a matter of balancing the risk of an exponent being done twice which would waste resources and the risk of an exponent not being done at all or desired results arriving undesirable late.

OK. A very fair and reasoned argument. Thank you for that. :sincerely:

I recently spent over two hours in a church where it was said "If anyone knows any reason why this [marriage] cannot lawfully be fulfilled, speak now, or forever hold your silence".