factordb seems to be having problems with a 50 digit composite. [url]http://factordb.com/listtype.php?t=3[/url] returns nothing unless you ask for numbers over 50 digits (or skip the first 50 digit number).

And it's not automatically factoring numbers over 50 digits. I've updated my script to factor small numbers to start at 51 digits to stop them building up.

Chris

Does this seem to be a continuing problem? Permanent change? What happened to the page that listed the autofactor boundary? I know it's existed in the past, but I can't seem to find that info anywhere

[QUOTE=Dubslow;486438]Does this seem to be a continuing problem? Permanent change? What happened to the page that listed the autofactor boundary? I know it's existed in the past, but I can't seem to find that info anywhere[/QUOTE]


Sometimes it works, sometimes not.
Yesterday I uploaded two ~110 digit factors (composite), booth got instand FF status. They had max 7 digit factors.

FDB seem also to struggle with even numbers. They are in the c-list, if you quarry them you will see that FDB adds some factor (even 50 digits factors) for them. The remaining cofactor was mostly generated a few months/years ago.

I also got some FF results while searching for even numbers in the composite list (notice degit size <1500 digits.)
They also occour at higher digit ranges, but way less.

Strange thinks happening.
I took a look into the PRP-list and opened the (atm) smallest Canidate.

It is listed as a factor of a 4361 digit number. I clicked on that number an it was shown as C without a factor!
I added the PRP as factor which worked...BUT.
[URL="http://www.factordb.com/index.php?id=1100000001125414468"]This number[/URL] is shown as factor from [URL="http://www.factordb.com/index.php?id=1100000001125414455"]this number. [/URL][URL="http://www.mersenneforum.org/(1881^4833+4833^1881)/3^5643/5894262665081118517795918"]
[/URL]
Surprising the last number 1881^4833+4833^1881/factors is listed as C without factor.

What is going on?
Also in the status tap I´m missing "smallest composite number without factors".

[QUOTE=chris2be8;486108]factordb seems to be having problems with a 50 digit composite. [URL]http://factordb.com/listtype.php?t=3[/URL] returns nothing unless you ask for numbers over 50 digits (or skip the first 50 digit number).

And it's not automatically factoring numbers over 50 digits. I've updated my script to factor small numbers to start at 51 digits to stop them building up.

Chris[/QUOTE]
They still seem to be building up. Would you like me to move my lower bound down as well? I won't unless/until you reply with a value.

I'm not limited by my CPU power, but I'm hitting the limit on how much CPU time factordb uses to process the factors I'm submitting. So I don't need help.

The problem is that someone is submitting a load of numbers like ((2^25-1)^1095-1)/(2^25-2) and (8191^2105-1)/8190 (the power of two and exponent vary). But they havn't submitted the algebraic factors for the numbers. My scripts are adding algebraic factors and finishing off small composites factordb found by ECM. I've tweaked my scripts to work a bit better so I should catch up once the flood ends.

More helpful would be whoever is adding them to add the algebraic factors as well. Or just pace the rate they submit them.

Chris

[QUOTE=chris2be8;486958]I'm not limited by my CPU power, but I'm hitting the limit on how much CPU time factordb uses to process the factors I'm submitting. So I don't need help.

The problem is that someone is submitting a load of numbers like ((2^25-1)^1095-1)/(2^25-2) and (8191^2105-1)/8190 (the power of two and exponent vary). But they havn't submitted the algebraic factors for the numbers. My scripts are adding algebraic factors and finishing off small composites factordb found by ECM. I've tweaked my scripts to work a bit better so I should catch up once the flood ends.

More helpful would be whoever is adding them to add the algebraic factors as well. Or just pace the rate they submit them.

Chris[/QUOTE]
OK, I'll stay at 79 dd and up, then.

[QUOTE=chris2be8;486108]factordb seems to be having problems with a 50 digit composite. [url]http://factordb.com/listtype.php?t=3[/url] returns nothing unless you ask for numbers over 50 digits (or skip the first 50 digit number).

And it's not automatically factoring numbers over 50 digits. I've updated my script to factor small numbers to start at 51 digits to stop them building up.

Chris[/QUOTE]
There's also one like that at 98 digits too.

I'ld wish there was a way to see them and help FDB get over it, but there don't seem to be. Further numbers not being factored seems to indicate that the factoring-small-numbers worker is stuck on an infinite loop with those bugged [I]hurdle[/I] numbers. Whenever a hurdle number is the smallest, that row is not present in the status page.

As for the flood. About 5 hours ago I've helped factor few hundreds of those stacking up small composites in 61-62 range. Maybe will do some more right now (not starting at the beginning to do same work less... halfway? halfway). Seems like there are about 0.9K numbers up to 70 digits right now.

Also, the worker that checks Unknown numbers started lagging behind for whatever reason. Load averages are highish, but I've seen them higher. Anyway, it's working, so whatever.

[QUOTE=DukeBG;486982]There's also one like that at 98 digits too.
[/QUOTE]

I think there are others higher up as well. My efforts to sweep downwards from 120 digit numbers also periodically hit them. Unfortunately, I haven't been able to isolate any of the higher ones to exact ids.

[QUOTE=DukeBG]Seems like there are about 0.9K numbers up to 70 digits right now.[/QUOTE]
70-digit ones start at page 1, so I consider the flood (at the moment) dealt with and stop my (minor) help right now.

[QUOTE]I think there are others higher up as well. My efforts to sweep downwards from 120 digit numbers also periodically hit them. Unfortunately, I haven't been able to isolate any of the higher ones to exact ids.[/QUOTE]
They likely have much older ids than their "digit-neighbors", there's not much hope to find them.

[QUOTE=chris2be8;486108]factordb seems to be having problems with a 50 digit composite. [url]http://factordb.com/listtype.php?t=3[/url] returns nothing unless you ask for numbers over 50 digits (or skip the first 50 digit number).

And it's not automatically factoring numbers over 50 digits. I've updated my script to factor small numbers to start at 51 digits to stop them building up.

Chris[/QUOTE]

The hanging C50 is fixed now.