Change in BA requirements

[orion]

Well, again just for perspective:
Bruce attempts to explain and "defend" the current system. When we are discussing failures, or flaws in the current system, this can become counterproductive.
Attaching personal blame is likewise counterproductive. What needs to be addressed is the structure of the system.

Why shouldn't bruce, or any other user, be allowed to have and voice a positive opinion?

Surely any debate about 'flaws' and 'failures' should allow comment from people who feel like there 'aren't any?'.

Isn't a "positive" view point in the eye of the beholder?

What you may consider positive I may not, like wise what I may consider positive you may not.

Is a lack of clear, precise and timely information from PG a positive thing? Is starting a thread in the beta section that had nothing to do with beta but all about ones views of the points being too high with BA's a positive thing?

Now I'm starting to sound like a broken record

[mdk777]

Why shouldn't bruce, or any other user, be allowed to have and voice a positive opinion?

Surely any debate about 'flaws' and 'failures' should allow comment from people who feel like there 'aren't any?'.

yup, no problem with that!
But after that, just explaining what I see as a legitimate cause of frustration:
The systematic deletion of threads.

When one has the power to enforce censorship,the shaping of what is the acceptable record of debate; one really has to expect some backlash.

[mdk777]

OK you might say, but what does your vendetta on communication, transparency, and accountability have to do with the specific topic?
We are tired of hearing about how you think open communication would avoid these situations.

Well, from memory, I can explain in concrete terms how it applies.

When launched, PG wrote a description of the requirements for BA.
They stated that they wanted 8 physical cores. They went further to specifically write that they did not want donors to attempt to OC a high-end 4 core to meet the time requirements.
Subsequently, intel launched a 4 core 8 thread processor that many donors promptly applied to BA.

This went on for a long time.

I pointed out in several discussions that this practice was in direct opposition to the stated requirements.

PG reviewed the situation and determined they would allow 8 threads.

However, rather than to add an addendum to the original description, making a notice that the requirements for participation were now revised...

THE ORIGINAL POST WAS SIMPLY REDACTED.

THE LINE EXPRESSLY STATING THAT OC 4-core was highly discouraged...was simply deleted.
Core count was simply replaced by thread in subsequent discussions.

So, now we are debating again if cores, or threads, or simply ability to make deadlines is important.

Well, given the history on this subject and the way it was communicated in the past, is there any wonder that donors are confused?

[PantherX]

...THE LINE EXPRESSLY STATING THAT OC 4-core was highly discouraged...was simply deleted...

By chance, do you mean this or not:

...Q: Can I try running these units on my super-overclocked, liquid-cooled quad-core system?
A: No. In our experience, fast quad-core systems tend to come in over the 4-day timeout and would thus 1) not contribute to the scientific goals of finishing these projects quickly and 2) not receive bonuses. Quad-core systems can make important contributions to standard SMP projects, and we'd encourage you to apply them there...

Source -> viewtopic.php?p=105038#p105038

[mdk777]

Q: What about Core i7 chips with 8 virtual cores?
A: In our testing, these have been found somewhat marginal for completing work within the target deadlines (ideally 3 days per WU). If you have a fast Core i7 that is a dedicated folder, feel free to give it a try. If you're doing a lot of other things with that system, standard SMP may be a better bet.

As I mentioned, BA was launched before I7 ...hence, by definition this is revision of the original post.

The original post did not say "feel free to give it a try."
It said it was forbidden.

As I recall, the original post had a line that went something like:

Can I use my highly OC, water-cooled super powerful gaming rig on these?
answer: No, we want only 8 physical cores. These are not intended for 1P gaming rigs.

You prove my point.

Again, you might claim my memory is flawed, but since we don't have a record of the revisions, since the modifications were merely edited into the description instead of being added as an addendum/change....we can't know who is correct.

Again, my point.

Will I win a debate on past mistakes and miscommunications in this forum?
Not likely. My point is to show how communication can either lead to an improvement, a clarification of concepts...or conversely into a institutionalized wall, a defensive structure that actually precludes the exchange of helpful information.

[orion]

This one too.

What systems can run these work units?
Right now, only Linux and OS/X systems can run these work units, and they require 8 or more cores. We prefer 8+ *physical* cores, although fast Core i7 machines that are dedicated folders have proven sufficient during the testing process. The points incentives are designed to match appropriate resources to points value; if your machine is marginal for the extra-large work units, you're probably better off running standard SMP.

[k1wi]

I started folding on my I7 920 (it came with my HD4770's drivers) and I still remember the launch of BigADV, so I7 must have already been out? Maybe my history is getting vague.

I see an argument for more concise, available and visible information.

[mdk777]

Well, I might also be confabulating beta with official launch timing.
Wish I could confirm the timing by looking back at the original post...oh snap.

[PantherX]

If we go by what is currently publicly available, then:
Intel i7-920 was launched in 2008 Q4 (http://ark.intel.com/Products/Spec/SLBCH)
Bigadv was launched in 2009 July 16 (viewtopic.php?f=24&t=10697)

Of course the above is for public release and not beta testing. While a change-log might have provided additional information/clarification, doing this on a Forum instead of a Wiki (which supports history function) isn't everyone's cup of tea.

Nonetheless, we can all agree that an official donor relations personal will be a valuable asset to the F@H community and will hopefully provide an effective communication channel

[orion]

Nonetheless, we can all agree that an official donor relations personal will be a valuable asset to the F@H community and will hopefully provide an effective communication channel

Indeed

[mdk777]

Well, my memory was pretty accurate about the changes to the BA description, if not to the launch date of I7.

viewtopic.php?f=55&t=15107&start=45#p150215

Quite the opposite

Fact: These were the original directions

Q: Can I try running these units on my super-overclocked, liquid-cooled quad-core system?
A: No. In our experience, fast quad-core systems tend to come in over the 4-day timeout and would thus 1) not contribute to the scientific goals of finishing these projects quickly and 2) not receive bonuses. Quad-core systems can make important contributions to standard SMP projects, and we'd encourage you to apply them there.



Fact: This was added when folders ignored the directions

Q: What about Core i7 chips with 8 virtual cores?
A: In our testing, these have been found somewhat marginal for completing work within the target deadlines (ideally 3 days per WU). If you have a fast Core i7 that is a dedicated folder, feel free to give it a try. If you're doing a lot of other things with that system, standard SMP may be a better bet.



Somehow the original information changed.

[Grandpa_01]

Q: What about Core i7 chips with 8 virtual cores?
A: In our testing, these have been found somewhat marginal for completing work within the target deadlines (ideally 3 days per WU). If you have a fast Core i7 that is a dedicated folder, feel free to give it a try. If you're doing a lot of other things with that system, standard SMP may be a better bet.

As I mentioned, BA was launched before I7 ...hence, by definition this is revision of the original post.

The original post did not say "feel free to give it a try."
It said it was forbidden.

As I recall, the original post had a line that went something like:

Can I use my highly OC, water-cooled super powerful gaming rig on these?
answer: No, we want only 8 physical cores. These are not intended for 1P gaming rigs.

You prove my point.

Again, you might claim my memory is flawed, but since we don't have a record of the revisions, since the modifications were merely edited into the description instead of being added as an addendum/change....we can't know who is correct.

Again, my point.

Will I win a debate on past mistakes and miscommunications in this forum?
Not likely. My point is to show how communication can either lead to an improvement, a clarification of concepts...or conversely into a institutionalized wall, a defensive structure that actually precludes the exchange of helpful information.

I was running Q9650's when bigadv first came out and was unable to run them. Shortly after i7 came out I switched over to 920's to run them, I do not remember i7 being available when they were first released seems to me bigadv had been available for a few months before the i7's were released. The original super OCed message you refer to was about the Q series CPU's I know that because I am one of the few that actually ran and successfully completed a bigadv WU on a Q series CPU actually I completed 2 of them (which everyone said could not be done .) But the bigadv release was very close to the releas of i7 cpu's. I had to sneakernet the WU from 1 of the 920's to test the ability of the Q9650's

I am encouraged by the current FF staff position on thread closure / manipulation / editing by mods etc. everyone of us knows that a very short time ago this thread would have been treated differently, I have not seen any of that happening in this thread yet, perhaps we should just wait and see what happens there. I for one am already seeing a change.

[mdk777]

anyway, to bring the discussion full circle.
As asked by many and unanswered in this thread by any:

Since the BA requirements have been thread and deadline driven for some time,
What is the rational for increasing the core count requirement at this time and with the timing described in the announcement?

[Bill1024]

anyway, to bring the discussion full circle.
As asked by many and unanswered in this thread by any:

Since the BA requirements have been thread and deadline driven for some time,
What is the rational for increasing the core count requirement at this time and with the timing described in the announcement?

Is it the up coming release of the new 8 core 16 thread core Intel desktop CPU in the 1st Q of 1014?
And a backlog of SMP WUs that needs to be folded? Just a wild guess.

[Viper97]

Nonetheless, we can all agree that an official donor relations personal will be a valuable asset to the F@H community and will hopefully provide an effective communication channel

Indeed

If in fact it happens and isn't just lip-sync.