Edwin Digest, SETI edition
A couple of days ago, Edwin said he wanted to set the record straight, and in reply I posted a link to an earlier post where I'd shown him to be willing to lie through his teeth in order to try and win debating points (as it were). August 30, 1999: Edwin posts initial Seti@Home results, including "Total CPU time": 977.5 hours.
Then, in the first interval between results, Edwin talks about how his PC is doing more work than a Mac can.
Why didn't you post better results from your Mac, Jeff?Edwin
How about because Edwin was posting results from more than one PC? And for anyone who doubts that
I think that the similarity between a computer that runs like it's "powered by a gerbil on a wheel" and my PC which has done more work than 80% of those performing this task is far more subtle, yet it's one you imply you can discern.Edwin
My time is high because most of it was done only running the seti client as a screen saver on a machine with 32 MB of RAM. I've gone up to 96 MB, and I set my client to always run in the background, so look for my times to go down. My address is , if you want to verify. I did this work on a AMD 366MHz machine that cost $936.25 with RAM upgrade. How does your Mac compare with that?Edwin
Please note references to "my PC" (singular), and "a machine", and "my client
My Gateway may not be running the seti@home client as fast as some other computers, but the seti client does NOT get slower as I add tasks to those already already running, and all other foreground and background tasks run without any noticeable slowdown. That's great multitasking. Laugh that off.Edwin
Once again, he refers only to "My Gateway" and "the seti client". If you want to read more such references, comp.sys.mac.advocacy+author. Nowhere does he ever suggest that he is actually running Setiathome on more than one machine.
September 19, 1999Edwin posts second set of Seti results. None of us pay attention to the "Total CPU time"; assuming nothing deceptive (more fools us ). Total CPU time: 1654.25
In this interval (calculating from the first and second results): Elapsed time between messages: 20 days Hours available(including an extra 24 to be extra fair): 504 CPU Time used in the interval: 1654.25 - 977.5 = 676.75 As a percentage of elapsed time: 134% Now tell me: how does *one computer* use 676.75 hours of in an interval which at *most* (allowing for the possibility that the first results were posted from 00:00 on August 30, and the second set from 24:00 on September 19) was 504 hours? But let us look on. October 4, 1999: Third set of results from Edwin; "Total CPU time" included. Still, nobody notices the discrepancy. Total CPU time: 2105.25 In this second interval: Elapsed time between messages: 15 days Hours available(without an extra 24, see below): 360 CPU Time used in the interval: 2105.25-1654.25 = 451 As a percentage of elapsed time: 125% Once again Edwin is posting results and the amount of CPU time far exceeds the total time between results. And remember, if we gave him the benefit of the doubt on the last results, then this interval could have at best run from 2400, Sep 19 to 2400 Oct 4. So no extra 24 hours credited this time. But let us move along. The *long* interval: For a long while, Edwin posts very little regarding Seti. He makes a few more remarks about how his PC is doing more work than any Mac until late October, 1999
My Windows box keeps running the Seti client *while* I use it for other work. It's also running the RC5-64 client in addition to the seti client and my other work.
I've seen a few other posts from Edwin on this subject. He bought a cheapo PC and can't figure out why it pucks blood while running Seti_at_home Heck, my 21/2 year old Power Computing clone puts up better numbers..
Why did your "better numbers" result in less work done by you than me? I explained why my speed was bad in several other posts. My total work is up, and my average time is down.
Why did Bill Altenberger's better numbers result in less work? Edwin knows. Because he was being lying. But no results until: Feb 10, 2000: Edwin once again posts update Seti results, including "Total CPU time" as always. Total CPU time: 5884.25
In this third interval: Elapsed time between messages: 129 days Hours available(including an extra 24 to be extra fair): 3096 CPU Time used in the interval: 5884.25-2105.25 = 3779 As a percentage of elapsed time: 122% Are we starting to see a pattern emerging here? Edwin's has a PC that can defy the laws of physics, or he's using more than one! Fourth interval: Edwin is posting about how he's using this as a benchmark between Macs and PCs (but a benchmark of this kind is only meaningful if you compare computers one to one, right?). He posts about how his "PC" can do all this work while he plays games, etc. There's even a post where he asks why I haven't completed more work than he has (perhaps because I wasn't essentially cheating on the comparison ). Perhaps this is what piqued my interest.
I use my computer for doing work and playing games, along with running the seti client. Because my PC has PMT, I can do that and all my normal activities as well. Mac advocates had claimed in the past that they can run the seti client much faster than I do. Yet all they have is sarcastic remarks to post, not any results to match mine.
Comp.sys.mac.advocacy exists to compare the Mac to all other platforms. I gave my seti client results from my PC. Let's see those from your Mac so we can compare them.
I invite you to "show me up" with your G4, if you can. My PC cost $899. How much did you pay for your G4?
Doing the same work on a Mac and a PC and posting the results wouldn't be "stupid," it would be doing the thing this group exists for.
I think this next one may have been what got me so pissed. It was in reply to me (as was the last one above):
That's bullshit. I have 260 blocks done to your 229, and your machine is doing the work faster? The seti client would never do a block in 11 hours as a screen saver on either a Mac or a PC. How about giving the email address you registered with seti so we can see for ourselves?
I run it as a screensaver.
Of course, my email was at the top of the message I'd posted and others quickly pointed this out to Edwin, who then looked like the fool he was. Again. So why didn't we do more work than you? Because we weren't running the client on more than one machine and lying about it. But in any case, somebody else was also speculating that something wasn't kosher with the number of results that Edwin was getting. So I went and got a copy of Edwin's results myself... February 16, 2000: I post a message with results for Edwin pulled from the Seti@home website at that time. I do the arithmetic and see for the first time that Edwin's results are using more "Total CPU time" in any interval than there are hours available. Total CPU time: 6068.5
In this interval: Elapsed time between messages: 6 days Hours available: 144 CPU Time used in the interval: 6068.5-5884.25 = 184.25 As a percentage of elapsed time: 130% Same old, same old. February 20, 2000: Edwin posts a reply to Znu where he makes no mention of any problem he today claims to have with the calculations. He just claims that they only reflect a change to more than one computer made recently.
Failure to live within the assumptions made by others does not constitute "lying" on my part. The collection of wackos here have their own unique form of 'logic'. Don't blame me for leading them to where they go. Exactly when did I tell anyone that I'd never use more than one computer to run the seti client? When did I say that I wouldn't add to my efforts after posting my results?
You led everybody to believe that you were only running SETI@Home on a single computer, when in fact you are not.
You even bragged about how because of PMT, you were able to get large numbers of units done on that single machine.Yes indeed. Does that mean I'm not allowed to allocate more resources to my efforts? When did I say I wouldn't?
The fact that conditions have changed after I made my orginal post doesn't make me a "liar." You have failed to point out any "lies" made by me. All you have is obfuscation. The fact that your assumptions are proven wrong doesn't make me a "liar."
What was the lie I'm being accused of telling? Your clear, concise, rational answer should follow below (I should live so long).
Note: he has no problem agreeing that the figures show more than one computer was used, he just claims it isn't relevant to the times when he was bragging about using one computer to get all this work done. Well we all can see now that he was using more than one computer from the *very beginning*. And Edwin: you just "lived so long". February 21, 2000: Edwin announces (*again*!) the he's leaving CSMA forever. Anyone want to speculate on why that date? His post's penultimate paragraph reads as follows:
I've left this group before, but this time it's for good. It's a promise this time, and I expect to be held to it.
May 9, 2000: Less than three months after announcing that he's left forever, gone beyond, whatever... ...he's back. August 13, 2000: Once again, Edwin states that the only reason the figures originally posted showed what they showed is that he *had* at that point started using more than one PC for Seti. And this time he specifically states that whatever he'd said about his results earlier was from when he was using a single PC. He even reiterates his claims that his still hoping that no one will check the older figures, I guess.
Not through most of my efforts, and not when I posted my challenge here.
What happened is that Mac advocates let many months go by, until I added more computers to my efforts. Then they came in with the lie that I'd always been using several computers. I'd only been using one computer for months, and averaging one block every 24 hours, while they claimed a Mac could do a block in 8 hours. Yet no Mac advocate could show more than 5 to 15 blocks finished, while I had hundreds done.
Using multiple computers, obviously.
So he's claiming that it was only later on that he'd added additional PCs -- still not disputing that the figures indicate anything else, mind! -- but as we've seen, that was just more Edwin bull. Since then, he's tried to obfuscate matters by claiming that by running Setiathome as a screensaver it somehow makes the CPU hours reported higher; not realizing that CPU hours are a finite resource! There is simply no way that a machine can perform more CPU hours in a given interval than there is *time*! But Edwin still hoped to obfuscate. But I'm willing to let everyone here decide; I wonder if Edwin is.