![]() |
|
|
#199 |
|
6809 > 6502
"""""""""""""""""""
Aug 2003
101×103 Posts
23×1,223 Posts |
Our church has been dealing with that for years. Our name dates to the 1860's. But since there is not central org, we can't protect the name. We have a television program that dates from the 1960's. Never thought about getting a formal TM on the name until the internet age.
|
|
|
|
|
#200 |
|
If I May
"Chris Halsall"
Sep 2002
Barbados
2×5×7×139 Posts |
|
|
|
|
|
#201 | |
|
Quasi Admin Thing
May 2005
17068 Posts |
Quote:
Due to your threats (or at least that is how they can be read), I've adviced Rebirther to shut down Trial Factoring and make a news post as to why GPU72 disappear. This may have serious negative consequence for GIMPs, because you will litterally loose hundreds and eventually thousands of GPUs all willing and able to compute for GIMPs. What puzzle me about you and outspoken members of GIMPs community, is the fact that you would really like to see your subproject and the main project speedup and move faster than the day before, but now it is the second time, that Trial Factoring of Mersenne numbers is brought to BOINC and it has been just the same amount of times, that GIMPs users has forced the eminent help from BOINC to hibernation and an eventual stop, due to lack of will to cooporate. I simply cannot see that you Chris has the right to act the way you do. BOINC could have gone lone rider and made thousands of your users results not needed, by poaching the reservation that GPU72 has made and return the results even at high bit levels, faster than your users could return and complete them. So maybe another time, start out by showing some respect and tolerance. You have only to a very small extent shown respect and tolerance, in your way of dealing with Rebirther, even after I told you, that there is sometime the need to say the same thing, just using other words, when the counterpart does not have the english or american english language as their maternity language. Finally I will say, that whatever time you may feel you have wasted, both I and Rebirther has wasted a lot more time, so step out of the role of victim and as mentioned earlier, control your inner R.D.Silverman. As mentioned to Rebirther, I'm sorry that I suggested BOINC in the beginning, I should have learned by now, that projects mostly directed and/or run from these forums frown upon BOINC and refuse to see how much benefit they could actually have from BOINC. Take care and maybe you and the real admins of GIMPs, should try and figure out if you are in fact owner and leader of this project or if you just has to play by the books, like BOINC were and would had continued to do. Should you eventually have a logical explanation towards your latest attack on BOINC and Rebirther, feel free to share that explanation. |
|
|
|
|
|
#202 | ||
|
If I May
"Chris Halsall"
Sep 2002
Barbados
230028 Posts |
Quote:
Quote:
Even the suggestion of setting up different classes of work, as already exists for other work types on SRBase, was rejected out of hand. "Nah, we're just going to do the easy stuff. Others can do what's actually needed. I hope Reb continues with his GIMFS work. Just not using the GPU72 handle, since it will cause confusion with the real GPU72 BOINC project I'm working on, which will do some of the "heavy lifting" work needed eminently. Good day to you. P.S. Your threat of having BOINC poach work was unnecessary. But, not unsurprising. You seem to take the position that GIMPS has to change it's years' long position of how work is done because BOINC is different. We don't have to. Last fiddled with by chalsall on 2020-04-07 at 18:42 Reason: s/operator, of GIMPS/leader, of GIMPS/; |
||
|
|
|
|
#203 |
|
6809 > 6502
"""""""""""""""""""
Aug 2003
101×103 Posts
23×1,223 Posts |
KEP and REB,
You 2 must realise that Chris (chalsall) has been doing this work for a while. George has seen that his approach at managing the a huge amount of GPU TF firepower is sane. Chris has been looking at the stats and managing the work (all the while letting people do what they like.) One of the big factors is managing the biggest benefit now, vs later. Doing work that will only benefit GIMPS 5 years from now doesn't help much. As the name of Chris's project GPU to 72 implies it was a goal to take exponents up to 72 bits. While was higher than the standard Prime95 level for the exponents of the time. The real goal is eliminate as many exponents as possible right in front of the "wave front". But, because there are many many exponents assigned in Cat 4 that never get completed, Cat 3 and even 2 become a different wave front. GPU72 want to have exponents at the highest possible bit level right before they are needed. Ideally the bit level is a wedge shape in front of the first time checks (the fat end toward the first time checks). The steeper the wedge that we can manage, the better. The further we move along, the more computing power is available and the higher that we can and need to push the bit level. Keep in mind that it is not just the Cat 4 wave that Chris is trying to manage, but the Cat 3 and wedging in the P-1 work to their maximum efficiency. George has given his approval (various ways) for Chris to manage this. It benefits GIMPS by allowing the CPU's to move the wavefront forward faster. Chris is now asking you to help him where the help would best benefit GIMPS. If 77-78 is too tough to get BOINC folks to do, just take exponents to 77. Grabbing all the 72-73 for exponents that won't see a first time check for 2 years yet is less helpful. It is making the wedge more shallow. Like manufacturing, we are going for "just in time delivery" of highly factored exponents. Please respect that. |
|
|
|
|
#204 |
|
P90 years forever!
Aug 2002
Yeehaw, FL
2×53×71 Posts |
Kep, Rebirther, Chris,
Can we please tone down the acrimony? Everyone is investing their personal time in good faith to further a common goal. I'd like to believe that what we have here is a communication problem. Can we give this another go at reaching a mutually satisfactory compromise? I suspect Rebirther is trying to keep his customers happy and his customers, especially with slow GPUs, are telling him they prefer shorter work assignments. If this is not correct, what exactly is the reluctance to work on 2^77 and 2^78 assignments? Are there other issues that we should know about that make your managing a BOINC server difficult? Is the offer to create new mfaktc / mfakto versions that work on a subset of the classes or support fractional bit levels at all helpful? Forgive me/we for not understanding the BOINC world, it is NOT that a BOINC contribution is frowned upon. The more I/we understand the BOINC issues, the more possible solutions we can propose. As possible compromise solutions that might keep both parties happy: 1) Can the 2^77 and 2^78 assignments be sent only to faster BOINC GPUs? 2) Chris: If BOINC would take assignments from GPU72 to 2^76 would that be useful to GPU72 and perhaps palatable to BOINC users? 3) Chris, you've a long, long history of supporting the GIMPS philosophy of letting the customer do what they find most suitable. If push comes to shove would you consider letting rebirther continue using the GPU72 trademark -- perhaps as a GPU72 sub-project that "softens up the long-term GPU72 needs"? Essentially becoming a GPU72 for slower GPUs whereas your BOINC efforts morph into GPU72 for "manly" GPUs. Peace, George |
|
|
|
|
#205 | |
|
If I May
"Chris Halsall"
Sep 2002
Barbados
2×5×7×139 Posts |
Quote:
Reb has demonstrated absolutely no willingness to be cooperative. Thus, I feel no desire to be either. Last fiddled with by chalsall on 2020-04-08 at 13:16 Reason: s/May 7th/May 7th, 2020/; |
|
|
|
|
|
#206 | |
|
Quasi Admin Thing
May 2005
2·3·7·23 Posts |
Quote:
I really regret getting Reb involved in this and at least in interaction with you. Seriously, you are and have been out of line for quite some time. Read the statement above and then you will know what I referded to in terms of threatening behaviour - and now you are doing it again. If you had initially any kind of objections against Reb using that name, maybe you should have taken steps to prevent it. It was after all your own demand and suggestion that BOINC reserved work through GPU72. And at George, do you really want to have your project associated to such a horrible person? Now Chris is just doing it in the open, what he did to Reb in private messages and to be honest, that kind of behavior (wether Chris is angry or not) is ultimately as a whole damaging towards your project. Please take in to consideration that eventually some of your superusers may not think positive about your great project when they see what kind of attitude you have to tolerate from selfproclaimed "directors/subleaders" of your project. I expect your kind personality to tell Chris a thing or 2. Personally I would really like for Chris to stop attacking and threatening Rebirther and maybe ultimately take a break from the forum and take a lesson in cooporating with other people. Last fiddled with by KEP on 2020-04-08 at 15:39 |
|
|
|
|
|
#207 | |||
|
If I May
"Chris Halsall"
Sep 2002
Barbados
2×5×7×139 Posts |
Sign... I really don't need this crap.
Quote:
Quote:
To be clear, I bent over backward to help in the PMs. Answering the same questions again and again. And this is the thanks I get. Quote:
And, for the record... Not once have I attacked or insulted Reb nor yourself. You, on the other hand, are on record for apologizing for inappropriately "hash" language towards me. |
|||
|
|
|
|
#208 | |
|
Quasi Admin Thing
May 2005
96610 Posts |
Quote:
Now could some of the moderators delete this thread, it is really not that helpfull for anyone anymore, especially not for GPU72 and GIMPs as a whole, also it is currently way off topic and since BOINC and GPU72 is no longer connected, there really is no sense in keeping this thread. Thank you! |
|
|
|
|
|
#209 |
|
P90 years forever!
Aug 2002
Yeehaw, FL
2×53×71 Posts |
Closing the thread.
|
|
|
![]() |
| Thread Tools | |
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| Chess World Championship Match -- 2013, 2014, 2016 | Raman | Chess | 34 | 2016-12-01 01:59 |
| mprime ETA and primenet "days to go" do not match | blip | Software | 1 | 2015-11-20 16:43 |
| less v4 reservations being made | tha | PrimeNet | 8 | 2008-08-14 08:26 |
| LL test doesn't match benchmark | drew | Hardware | 12 | 2008-07-26 03:50 |
| WE MADE IT!!!!!!!!!!!!!!!!!!!!!! | eric_v | Twin Prime Search | 89 | 2007-01-23 15:33 |