I think switching to RC5 is a bad idea ...

[Michael]

Edit - deflamed as per the request of DanC


If there really was a TEAM thought in the RC5 people, they would not raid the SETI team. We finally have been able to recruit a fair number of people and get our WU count up where it is making a difference and then this?

I thought that there was a gentleman's agreement about not recruiting each other's members. I guess I was wrong. This poor idea will end up wrecking both teams. SETI will lose it's recruiting power and never will recover. RC5 could take YEARS to complete, or at least many months.

At no point have ANY of the main RC5 people discussed coming over to SETI when the contest ended. OGR or the next DNET contest has always been the next one in line.

I am doing 5K RC5 blocks a day (with 800K for the team) and the second thing I did in my new job today was to install RC5 (after I loaded SETI). I was going to convert my new department over to RC5 and get at elast one new member to honor DanC's request.

Michael

 

[DanC]

Toned down out of respect for Michael - who has been an avid RC5 proponent.

 

[Michael]

He's offering them AFTER RC5 is done. What does that mean?

Has no one in the RC5 team been reading the SETI threads recently. We're on a major roll in getting new members. We do not need a raid (and I consider this a raid because of the time it will take to complete RC5) right now.

I was working to recruit new RC5 members, not just SETI. The difference was I was trying to recruit totally new members.

Michael

 

[Russ]

I knew this was coming. Frankly, I'm surprised this rant didn't come up sooner.

Michael,

All Train did was offer a trade-off. Some SETI people volunteered to help out. This is EXACTLY what has been suggested by SETI team members in the past; only in reverse.

Have you stopped to consider the enormous amount of power that could go to the SETI effort if this works out? Any idea how many SETI WU's can be done by the massive number of systems in the RC5 effort?

It's sad that we can't seem to go even a day around here without arguing, threats, yelling and screaming.

Russ, NCNE

 

[DanC]

Let's all have a beer ok?

Michael and I are talking... This was not intended to be a raid - I think we've established that. I also believe that Michael's feelings are valid.

We must pursue endeavors that benefit ALL PROJECTS within TA. That should be our goal and our purpose.

We aplogize sincerely for any offense that was caused. It was not Train's, nor my intent. Let's get past this ok?

 

[networkman]

I can see your point, Michael. The "one" machine that I'm building for the RC5 effort is a PII-266, which unfortunately doesn't do too well in SETI anyways, so it should do something else. But I, like you, do not plan to switch the rest of my herd from the SETI effort to the RC5 project. Likewise, I will continue to recruit new people for the TA SETI effort and not RC5.

I also see your point regarding the lack of willingness of people to leave the RC5 project for SETI once the current RC5 project is completed; OGR and the next dnet project have always been discussed.

And before anyone accuses me of being negative, I wasn't involved in the threads about the "it won't work" subject. What I'm typing now is exactly what I see going on. IMHO, Michael's assesment is essentially accurate: Some RC5 folks are looking to "rob Peter to pay Paul." Well, I'm willing to donate some to Paul, but that's it.

Edit: It should be noted that I while I was typing my reply to Michael, that both Russ and DanC replied BEFORE my message was posted; having seen what they've written, I don't see the need for apologies, but I also do not see any need to re-cant what I've originally written either.

 

[Wolfie]

Michael

First, I would like to say...... you have the right to say whatever you feel is right.

BUT, We are not "raiding" the SETI team. I guess when you say

<< At no point have ANY of the main RC5 people discussed coming over to SETI when the contest ended. >>

I would like to know who these &quot;MAIN RC5 PEOPLE&quot; are. I guess I am a nobody. I stated that I would join SETI. Even tho I don't have the LARGEST herd in RC5 but I thought I would help the SETI team when this thing was all over. I even thought about staying with SETI if RC5 started up again. Don't make me second guess myself please.

Wolf

[EDIT]Forget what I said. I was typing this out when there was no other replys to this post...... no reason to read what I said.... just ignore it... period.....[/EDIT]

 

[Michael]

Russ,

The only suggestion that recurrs is that the SETI team would love to have RC5 power directed their way AFTER the contest is over.

If you expected this type of reaction, then why were you not arguing against it to cut down on the internal team division?

If all 150 odd active SETI people switched over to RC5, we would boost the effort by 20% or so. That's 200K blocks (say 300K blocks) per day. How much faster would that make the contest run? The offer to help once RC5 is done if the SETI folks help now will hurt SETI a lot and not really help RC5.

What RC5 needs is new members that are not doing any DC project at all. That is what DanC asked us to do, and that's what I spent 1/2 hour today trying to do (I have 2-3 very likely candidates). Recruiting from other projects is a zero sum game and should be avaoided.

Michael

 

[DanC]

I must say I agree with Michael. If there are any voluntary additions - that is entirely up to members, just as it is for RC5 folks to do SETI. That seems fair.

Cease and desist all actions which could be construed to be inter-team recruiting and focus on TEAM ANANDTECH - and NEW MEMBERS. That is the goal.

 

[networkman]

That's fine, DanC, I can abide by that.

 

[TheyCallMeSAK]

Well, this is certainly something new to me. Again, I have not been able to check out the forums as much as I would like to. I would just like to say that I enjoy Team AnandTech being involved in VARIOUS projects, unlike many other teams. This reply is for everyone, not just those that have replied to this thread already. Like I said, I don't know the whole story, but I don't care!

We obviously have enough TOTAL power to rule any competition we would want to, but I don't think we should look at it that way. Each contest is just as important as the next, in a sense. No matter what TA contest you do it for, JUST ASSIMILATE. Please don't recommend where others should be focussing their power either. I am an RC5er, but I have considered doing other contests also. It has just been easier to keep track of one for the time being.

Our SETI team does need some help! That is very evident! They have all done a great job assimilating and recruiting so far, so this makes me happy . Our OGR-25 effort isn't going quite as well as I would have hoped, but this could also change. And to top it off, we are losing the dailies to the DPCs in the RC5 contest. As you see, we have our power split up, but that makes it all the better!

So come on, guys, just do your best to increase the power of any of these TEAM EFFORTS!! I'm sure with serious work, we could do well, errr BEST , in many projects!

 

[Russ]

<< If you expected this type of reaction, then why were you not arguing against it to cut down on the internal team division?] >>



Michael,

Because we spend entirely too much time around here finding reasons why an idea won't work, instead of seeking solutions, and I'm not going to add to that. More importantly, a lot of SETI people jumped all over it and offered to join in, and I didn't feel like saying to them: &quot;Thanks, but no thanks.&quot;

It essentially boils down to this: I am going to recruit, offer help, and keep trying to build Team AnandTech. I am NOT[/i] going to participate in these arguments anymore.

But, in the final analysis, you are right. New blood is what is needed.

Russ, NCNE

 

[DanC]

Thanks guys -

Obstacles are what you see when you take your eyes off the goal. Since our team focuses on multiple projects, we must collectively (no pun intended) focus on things that benefit all of the projects. To this end, I still firmly believe that Train is on the right track with the database - the newsletter - etc.

If we do not manage our membership and share information, we lose. The DPC do this very, very well. This is their edge. We haven't been fighting the DPC - we've been fighting our own lack of organization. IMHO.

 

[Michael]

I am 100% behind getting brand new members for RC5 and will continue to help to do that. I will also continue to add computers to the RC5 effort at my new work place. When we have as many members as the DPC, then we will be #1 until the contest ends.

Michael

ps -any SETI person who wants to toss some power behind RC5 or OGR is welcome to. Some CPUs are not anywhere near as efficient in SETI as they are in the other projects.

 

[ViRGE]

Ok, who tried to recruit the SETI memebers!?:| We've had a long standing agreement from the Alien Cow Wars(TM) that niether team would try to recruit the other's members, of any condition short of an apocolyspe for either contest. Now, you go about and do this!? Shame, shame on you.

 

[CurtOien]

I would not want all of the SETI guys to switch to RC5, nor should they.
Whatever I can get a new person to run I will.

It's just that, at this time we should take the number one spot in RC5 soon and if trends continue DPC will take it from us shortly afterward.

The sooner we can take the number one spot the more time we will have to get all the people who like to jump on the bandwagon to do so. If we act quickly and forcefully we may be able to get enough people to hold off DPC forever.

Nobody recruited us.
We are making a trade.

 

[DanC]

Virge -
Don't get your panties in a bunch. We've already dealt with it.

 

[CurtOien]

People have made trades in the past with machines that are better for one thing or another.

 

[Tetsuo316]

Michael,

I replied to your post in a previous thread that I also had feelings like yours. Initially I was pessimistic about our (seti members) contribution to the rc5 team, and questioned whether it would make a substantial difference or not. I concluded that what is most benficial out of our involvement with rc5 is not the computing power, but a team spirit. Sure, rc5 could take months, or even another couple of years, but seti isn't going to be over soon - by any means. What's made me so sure about this is the intangible 'feeling' I got while reading the posts all today. This is a team to be reckoned with! This is a team that doesn't stop at its particular program. This is a team that sacrifices. I was excited! This doesn't happen while reading forums on the internet, but it happened today. Our production will suffer some now while individual user branch off, but it is the gesture that we made today that will bring many members to the seti team once rc5 is through. At that time, we will reclaim our current position and sail into the top ten. Our output might not suffer that much at all in the meantime.

 

[Robor]

I agree with Tetsuo316 100%...

I've been doing SETI since the client was released and I'm at my peak output right now (60+ WU/day). So is TA SETI. That said, we are still building very slowly and much of our output comes from the top 20 or so members. If (for whatever reason) we lose a few of them it would severely impact our output. However, if some of us switch some of our power to help the rc5 team now, they've offered to help us in the future. Remember that the rc5 team is 4-5 times our size and power. I'm willing to &quot;sacrifice&quot; some output today for the promise of a big payoff in the future. Can you imagine if after the rc5 project completes we get a combined effort behind SETI? IMO, we could easily make up the &quot;lost&quot; ground.

Remember, we're not leaving the SETI project... Not by a long shot.

Rob

 

[Michael]

Repeat from another thread ...

Tetsuo316 - In my opinion 1) The SETI team has managed to grow and expand production enough that we are now gaining on the teams ahead of us. This has given us momentum and is a major help in us recruiting new members. The RC5 people who have been around longer will admitt that they really started gaining members once they started jumping up the rankings. People like to join winners. If the SETI production number drops too much, then the effort will lose that momentum. That will hurt RC5 (and the other DC projects as well) as that will mean less people drawn to the DC forum.

2) I don't know where the idea came that the SETI team switching over would help end RC5 very quickly. Even if we managed to double the Anandtech RC5 output (and we wouldn't, there are only 150 odd active SETI people which is only 1/4 the number of RC5 people), it still could easily take years to finish the project. No one knows where in the keyspace the key is. If it is at the 80% mark, and it has taken 2 years to get to 30%, then it could take another 2 years to end RC5.

3) There has been an agreement for quite some time not to recruit from other projects. I know that Train meant well, but his &quot;offer&quot; is recruiting from SETI. The offer to throw RC5 power behind SETI once the project is done doesn't help if R5 takes another 2 years to run.

Any individual member can make their own decisions. If someone wants to throw computer power behind RC5, then more power to them. However, this should be an individual decision and not as a result of direct recruiting from one project to another.

Of course, that is just my opinion, but I do feel strongly about this issue. This is coming from someone with 800K RC5 blocks who added 10 new computers to RC5 over the last month and who is planning on adding more.

Recruiting completely new members who are not doing a DC project at all (ala Dennilfloss recently and DanC's challenge) is the best path to victory.

Michael

My RC5 stats. Yesterday was my best day ever!

ps - Robor - Since you are a big producer, throwing your SETI machines at RC5 will reduce SETI output by about 3%. In many ways, that is the difference between us gaining on the teams right ahead of us and losing ground. It will be much harder to recruit for SETI if the team is losing ground (human nature, everyone loves the winning team).

 

[Assimilator1]

The trouble is ,when is RC 5 over?

If its 2 years ,thats too long for me.Also the current SETI project is due to be finished in May (unless some one can show 100% otherwise).In which case the damage would be permanent.
On the other hand IF SETI continues with the same stats (ie not reset for another SETI project),then damage would be more short term ,however long that is............
For now I'm staying 100% SETI ,unless I find any 32Mb(or less) PC's to assimilate.

 

[Hellburner]

I believe the Alien Cow Wars severely crippled the S@H effort in the beginning and even turned some RC5'ers away from TA. A year and a half ago the same promises and deals were being offered/rejected, the history of the wars and the (in my mind) broken integrity of the RC5 project (cracked clients) makes this kind of activity questionable.

I whole heartly endorse trading off CPU's for optimal deployment of our forces, with the introduction of the 3.0 S@H clients I believe most of the non-on die L2 cache machines have taken a serious hit in completion times. This would have little effect on the RC5 client so trading off P5 and Katmai core machines for medicino/cu-mine machines could easily benefit all involved.

Since there is so much interinvolvement in the teams anyways I think a S@H miniteam in RC5 is a great idea.

 

[Viztech]

If I could ask a quick question HB...I think I know the answer, but if you could clarify mendicino/katmai core for us that don't know, and also where do the various AMD K6s and Celerons come out in the performance with the new 3.0 clients?

I also agree with the trade off of power for the sake of efficiency. Besides, variety is the spice of life, and SETI is an interesting project.

I am into OGR quite heavily, and would be interested with an appropriate trade to get K6s on OGR. I have several OCed Celerons, but I don't know how they would stack up now in SETI efficiency.

viz

 

[Hellburner]

The Mendicino core is the old style Celeron A's, the Cu-mine core's are the Celeron II's and P3 E's(256k L2 cache). The Katmai are the P2's and P3's with 512k L2 cache.

The Celerons and P3 E's have full speed L2 cache, the P2's and older P3's have half speed L2 cache.

The performance difference with the new S@H client appears to be associated with the full speed L2 cache. While I'm not nearly as familiar with the AMD processor line I know that Thunderbird owners are much happier with the new client, I assume this would be based on the full speed L2 cache. The new client hasn't been out long enough for me to get a good detailed picture of how all the AMD processors stack up.