Factoring database

[henryzz]

Originally Posted by henryzz

edit: are the workers down or something?

i still cant assign work to workers

edit: now can thanks
edit2: trial factoring and primality proving seem to work but the rest just say waiting for worker
there seems to be a ridiculous amount of qs work assigned on k*2^n-1 possibly creating a backlog
i have done a little but i have done nowhere near as much as is assigned

[mdettweiler]

Originally Posted by henryzz

i still cant assign work to workers

edit: now can thanks
edit2: trial factoring and primality proving seem to work but the rest just say waiting for worker
there seems to be a ridiculous amount of qs work assigned on k*2^n-1 possibly creating a backlog
i have done a little but i have done nowhere near as much as is assigned

According to http://factorization.ath.cx/workerstatus.php the worker queue is completely empty and idle now.

[henryzz]

Originally Posted by mdettweiler

According to http://factorization.ath.cx/workerstatus.php the worker queue is completely empty and idle now.

I didn't know about that page.
Where is the link to it from the main pages?

[kar_bon]

Originally Posted by henryzz

I didn't know about that page.
Where is the link to it from the main pages?

when you're in the display of a special number say "3^789-1" there're listed the factors and the composites. at the bottom of the page the composites are listed with the choices of factoring. the title says "Assign cofactors to workers (?)" -> click the '?' and you got those infos!

[kar_bon]

great enhancements in the last weeks for your Factoring Database!

very nice work!

go on!

[10metreh]

Markus, could you add a "very high limits" option?

[mdettweiler]

Originally Posted by 10metreh

Markus, could you add a "very high limits" option?

Seconded--a "ECM very high limits" (for 30 digit factors) would be great. Even though these are heftier jobs, it shouldn't be too big a deal since the workers are idle most of the time anyway.

Oh, this reminded me of a suggestion I had thought of regarding the workers: I've noticed that, even though there are dedicated workers for TF to 1e6, TF to 1e7, Miller-Rabin probabilistic primality test, and primality proof, all four of them are idle practically all of the time (and even when they are in use, their tasks are so tiny that they're done in less than a few seconds). Thus, I was thinking that it might be more efficient to combine some of these lightly-used workers. Firstly, I think TF to 1e6 and TF to 1e7 should be combined into one worker; even with both tasks on the same machine, it will most likely still be idle 99% of the time. So, I was also thinking that primality proof jobs, and possibly even PRP tests, could be also combined into this same worker. Again, most likely it will still be idle most of the time (unless some low-priority jobs such as massive primality proving are inserted on a regular basis), but the main benefit would be that 3 additional workers could be moved to ECM/sieve/P+1/P-1, which is in much, much higher demand than any of the others.

Anyway, just a suggestion.

[Syd]

Originally Posted by 10metreh

Markus, could you add a "very high limits" option?

Something i thought of, too. My plan was to start two steps, first: Assign as "high limits" to one worker to see if there are small factors. Then: Assign with B1 ~ 7E5 to all idle workers at once and run, maybe, until another job comes in. That would be >20Ghz at once!

The workers on the X5335 machine: Actually its not idle, it runs my sieving jobs in background
But you're right, the dedicated workers for TF and so on are not really necessary.

[schickel]

Syd,
Could you add one more option for the aliquot output? It would be useful for me to get the output of a sequence as straight text without the vaious info tags. Would that be too difficult to add?

Thanks,
Frank

[Syd]

Originally Posted by schickel

Syd,
Could you add one more option for the aliquot output? It would be useful for me to get the output of a sequence as straight text without the vaious info tags. Would that be too difficult to add?

Thanks,
Frank

Hi Frank,

I've added a simple compact text output, is it what you expected?

Btw: The very high limits option is working now

[10metreh]

Originally Posted by Syd

Hi Frank,

I've added a simple compact text output, is it what you expected?

Btw: The very high limits option is working now

The text format:

1. Could you make it write the full numbers instead of "P86" and the like?
2. Sometimes the factors are written backwards!
3. A big bug: when there is a power of 2 (or another number), the text format will only give "2", instead of "2^2".

The very high limits option is good!