Grumbles, Glory and All Your Off Topic Discussions

Message boards : The Lounge : Grumbles, Glory and All Your Off Topic Discussions
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 72 · 73 · 74 · 75 · 76 · 77 · 78 . . . 117 · Next

AuthorMessage
Richard Haselgrove
Volunteer tester
Help desk expert

Send message
Joined: 5 Oct 06
Posts: 5077
United Kingdom
Message 108832 - Posted: 1 Jul 2022, 12:14:15 UTC - in response to Message 108831.  

Fair comment. Largely because of Collatz, other projects went their own way, too, and broke the original plan of cross-project comparability.

And also broke any semblance of honesty in the reverse calculation of work done in petaflops across BOINC as a whole. that seems to have vanished from this website at long last, but David used to use a broken global estimate in his applications for money from the NSF.
ID: 108832 · Report as offensive     Reply Quote
betreger
Volunteer tester
Help desk expert

Send message
Joined: 18 Oct 14
Posts: 1472
United States
Message 108833 - Posted: 1 Jul 2022, 12:16:24 UTC - in response to Message 108831.  

The idea of credit new is good. The implementation is not. Collatz giving credit faster than everyone else is stupid and should be stopped.

Amen
ID: 108833 · Report as offensive     Reply Quote
Profile Dave
Help desk expert

Send message
Joined: 28 Jun 10
Posts: 2518
United Kingdom
Message 108834 - Posted: 1 Jul 2022, 12:26:29 UTC

Makes me glad I only look at credit as a way of spotting other problems.
ID: 108834 · Report as offensive     Reply Quote
Ian&Steve C.

Send message
Joined: 24 Dec 19
Posts: 228
United States
Message 108835 - Posted: 1 Jul 2022, 13:22:54 UTC - in response to Message 108831.  
Last modified: 1 Jul 2022, 13:24:34 UTC

The idea of credit new is good. The implementation is not. Collatz giving credit faster than everyone else is stupid and should be stopped.


the fact that Collatz has an impossible goal combined with its disproportionate credit award makes it especially egregious. useless project wasting computational resources crunching numbers for the sake of crunching numbers.

and to top it off, their results aren't even valid. their code/method is flawed.
ID: 108835 · Report as offensive     Reply Quote
Sirius B
Avatar

Send message
Joined: 12 Jun 09
Posts: 2098
Ireland
Message 108836 - Posted: 1 Jul 2022, 15:07:48 UTC

Glory (well (partially).
After being patient enough, called council.
CT overpayment should be in bank approx. Tues week as will the energy rebate.
Reason given for so long is refunds are only done on Weds.

Grumble.
Amazing how "automatically" becomes manual.
Informed that it must be requested.
Asked about the energy rebate. Got asked questions.
After giving date of application & app number, got told "Oh, you should have received it by now".
Got given an e-mail address to chase it up.
WAS very calm & collected as I said "thank you, have a nice weekend" & hung up.
As for what I was actually thinking, ask me no questions & I'll tell you no lies.
ID: 108836 · Report as offensive     Reply Quote
robsmith
Volunteer tester
Help desk expert

Send message
Joined: 25 May 09
Posts: 1283
United Kingdom
Message 108837 - Posted: 1 Jul 2022, 15:48:14 UTC - in response to Message 108835.  

Well said. I would rename that project as "Collatz Con-trick".
ID: 108837 · Report as offensive     Reply Quote
betreger
Volunteer tester
Help desk expert

Send message
Joined: 18 Oct 14
Posts: 1472
United States
Message 108838 - Posted: 1 Jul 2022, 17:00:35 UTC

I'm glad that I am not the only one who feels strongly about collatz,
ID: 108838 · Report as offensive     Reply Quote
robsmith
Volunteer tester
Help desk expert

Send message
Joined: 25 May 09
Posts: 1283
United Kingdom
Message 108841 - Posted: 2 Jul 2022, 8:04:58 UTC - in response to Message 108840.  
Last modified: 2 Jul 2022, 8:05:36 UTC

Neither have I got a maths degree, but I could follow the logical arguments outlined in another thread:
https://boinc.berkeley.edu/dev/forum_thread.php?id=14159


Put simply, the Collatz Conjecture at Home project uses a sieve to remove a lot of data points, which is fine if only "invalid" data points are removed. However the sieve used throws out a lot of valid data points (and possibly misses a few of the invalid ones?). This is not "good science". In the project's defence it has stated that it is not interested in the absolute number of valid data points, but some subset of them, this ignores the act that part of the required proof of the conjecture requires that all the valid data points be listed in sequence, along with the actual number of data points.

I hope we don't go onto repeat the debate that took place in the thread I linked to, as that got diverted into a couple of (interesting?) sidelines, and with the Collatz Conjecture project "slumbering" we don't have ready access to any of their debates (if any?) on this matter.

(* Before anyone screams, I know that I've used "data points" instead of the more correct "computational steps".)
ID: 108841 · Report as offensive     Reply Quote
Profile Dave
Help desk expert

Send message
Joined: 28 Jun 10
Posts: 2518
United Kingdom
Message 108843 - Posted: 2 Jul 2022, 12:27:44 UTC - in response to Message 108842.  

I'm sure there are many that fully understand it and could make a rival Collatz project that works correctly. Presumably it's not that hard to set up a Boinc server. I only have 32Mbit down, 7Mbit up here, but there must be someone with decent broadband that could host it.


Fully understanding the conjecture is not the problem as you suggest. However, there are many things I consider more worthwhile to have my computer crunching even were the project's method of testing the conjecture not flawed.
ID: 108843 · Report as offensive     Reply Quote
Profile Dave
Help desk expert

Send message
Joined: 28 Jun 10
Posts: 2518
United Kingdom
Message 108845 - Posted: 2 Jul 2022, 15:07:23 UTC - in response to Message 108844.  
Last modified: 2 Jul 2022, 15:09:02 UTC

Perhaps so, but we might as well solve the problem properly.
There are mathematical conjectures where it has been proved that no proof or refutation is possible! I suspect an actual proof or otherwise of the conjecture if it ever comes will be through mathematical rigour rather than crunching which looks to me like it only checks that the conjecture is true for increasing values of n.

Edit: Obviously not as simple as proving if true for n then true for n+1 as in some more elementary proofs or it would have been done by now.
ID: 108845 · Report as offensive     Reply Quote
Profile Jord
Volunteer tester
Help desk expert
Avatar

Send message
Joined: 29 Aug 05
Posts: 15477
Netherlands
Message 108846 - Posted: 2 Jul 2022, 15:51:54 UTC

Pathetic Lewis Hamilton fans who need to boo Max Verstappen at Silverstone.
ID: 108846 · Report as offensive     Reply Quote
Profile Dave
Help desk expert

Send message
Joined: 28 Jun 10
Posts: 2518
United Kingdom
Message 108847 - Posted: 2 Jul 2022, 16:09:41 UTC - in response to Message 108846.  

Pathetic Lewis Hamilton fans who need to boo Max Verstappen at Silverstone.
Agreed, it wasn't as if he was the one that made the decisions that were questionable.
ID: 108847 · Report as offensive     Reply Quote
Profile Dave
Help desk expert

Send message
Joined: 28 Jun 10
Posts: 2518
United Kingdom
Message 108849 - Posted: 2 Jul 2022, 16:45:39 UTC

This is taken far too seriously.
There are people who take any sport too seriously. My personal view is that the proportion of people who do so is directly proportional to the amount of money involved.
ID: 108849 · Report as offensive     Reply Quote
Profile Jan Henrik
Avatar

Send message
Joined: 5 Jul 20
Posts: 33
Message 108855 - Posted: 3 Jul 2022, 13:35:01 UTC - in response to Message 108848.  

F1 is racing for OCD folk.


yikes! I ‘d rather avoid any “race condition”
ID: 108855 · Report as offensive     Reply Quote
Previous · 1 . . . 72 · 73 · 74 · 75 · 76 · 77 · 78 . . . 117 · Next

Message boards : The Lounge : Grumbles, Glory and All Your Off Topic Discussions

Copyright © 2024 University of California.
Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.