mersenneforum.org  

Go Back   mersenneforum.org > Factoring Projects > FactorDB

Reply
 
Thread Tools
Old 2019-03-08, 13:59   #1
wpolly
 
wpolly's Avatar
 
Sep 2002
Vienna, Austria

3×73 Posts
Default Spammers in FactorDB

As we know, lots of nonsensical composites are being added to the database every hour, overwhelming the factorisation workers and has pushed the total number of <120 digit composites up to 900K. I've noticed some trends among these abominations, and it looks like the spammer(s) are unaware of the stress they put on the database, and are merely abusing it in order to find primes.

Here are some common types of the useless composites they produce:
  • Nonsense formulas such as a*b##+c or a^b-concat(a,b). It looks like they're using the "only show primes" option in the database search function, unaware of the fact that each query adds up to 500 new composites to the database.
  • Substrings of some long decimal string (likely from the expansions of certain irrational numbers), as shown in this screenshot:
    Click image for larger version

Name:	Screenshot_20190308_143241.png
Views:	265
Size:	41.0 KB
ID:	20005
    It looks like they're skipping substrings that ends in 0,2,4,5,6 or 8, further confirming that their main purpose is to look for primes.
  • (some known prime)+n where n goes up to 10^5 - the worst offender is probably (10^77*28-1)/9+n, for which I have seen n~90000. Like the first case, they are using the "only show primes" option to only search for primes without caring for the mess they left behind.
  • Last and the worst, decimal strings containing certain "birthdays" and other random substrings. Whoever doing this was probably born on 29 August 1983.......

The database search function should really be fixed so that it does not produce unwanted composites.
wpolly is offline   Reply With Quote
Old 2019-03-08, 16:50   #2
chris2be8
 
chris2be8's Avatar
 
Sep 2009

32·227 Posts
Default

Do any of the other substrings look like a telephone number, zip code, SSN, etc? Or a name when converted to ASCII?

It would be nice to be able to phone them and tell them what you really think of them (preferably at about 5am their time).

Chris
chris2be8 is offline   Reply With Quote
Old 2019-03-08, 18:32   #3
wpolly
 
wpolly's Avatar
 
Sep 2002
Vienna, Austria

110110112 Posts
Default

Quote:
Originally Posted by chris2be8 View Post
Do any of the other substrings look like a telephone number, zip code, SSN, etc? Or a name when converted to ASCII?

It would be nice to be able to phone them and tell them what you really think of them (preferably at about 5am their time).

Chris
I can identify two further birthdays (both in the 60s), presumably their parents....
wpolly is offline   Reply With Quote
Old 2019-03-09, 13:14   #4
Dr Sardonicus
 
Dr Sardonicus's Avatar
 
Feb 2017
Nowhere

2·2,273 Posts
Default

Quote:
Originally Posted by chris2be8 View Post
It would be nice to be able to phone them and tell them what you really think of them (preferably at about 5am their time).
This brings to mind the following Paul Erdos anecdote:
Quote:
When the name of a colleague in California came up at breakfast in New Jersey, Erdos remembered a mathematical result he wanted to share with him. He headed toward the phone and started to dial. His host interrupted him, pointing out that it was 5:00 A.M. on the West Coast. "Good," Erdos said, "that means he'll be home."
Dr Sardonicus is offline   Reply With Quote
Old 2019-04-15, 14:04   #5
DukeBG
 
Mar 2018

3×43 Posts
Default

Somebody submitted about 11 thousand 300-digit PRPs that are most likely literally consectutive primes. Example.

From the ones I checked, only the last 7 digits are varying. With the average gap at that size 11000 * 300*ln(10) ≈ 7.6e6, thus I'm saying they're very likely just consectutive.

Sigh. I'ld wish there was a "notable" field for a number in FactorDB (essentially a text comment; autofilled for numbers in the tables, sequences and their factors, +/-1 for PRPs) and not notable numbers would be auto-purged after some time.

Last fiddled with by DukeBG on 2019-04-15 at 14:06
DukeBG is offline   Reply With Quote
Old 2019-04-16, 11:19   #6
DukeBG
 
Mar 2018

3×43 Posts
Default

Quote:
Originally Posted by DukeBG View Post
Somebody submitted about 11 thousand 300-digit PRPs that are most likely literally consectutive primes. Example.

From the ones I checked, only the last 7 digits are varying. With the average gap at that size 11000 * 300*ln(10) ≈ 7.6e6, thus I'm saying they're very likely just consectutive.

Sigh. I'ld wish there was a "notable" field for a number in FactorDB (essentially a text comment; autofilled for numbers in the tables, sequences and their factors, +/-1 for PRPs) and not notable numbers would be auto-purged after some time.
Most of these thousands are now proven by certificates by the user "apcnc" (the one I linked – proven by anonymous).
DukeBG is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
Other Factordb Problems wblipp FactorDB 465 2021-05-14 15:57
A suggestion for factordb. enzocreti FactorDB 10 2021-01-05 19:49
Extending Factordb carpetpool FactorDB 6 2017-01-23 11:04
FactorDB PRP's smh FactorDB 231 2015-07-28 02:30
Closing Older Threads to Thwart Spammers Prime Monster Forum Feedback 0 2004-10-24 23:47

All times are UTC. The time now is 00:53.

Mon May 17 00:53:44 UTC 2021 up 38 days, 19:34, 0 users, load averages: 1.73, 1.59, 1.57

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

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.
A copy of the license is included in the FAQ.