[QUOTE=Uncwilly;392807]Does the current server all for manual submissions from Luigi's (ET_) Factor5?

The reason that I am asking, with the minor success that TJAOI has been having with missed factors, I thought about re-running exponents (with no known factors) in the 100M digit range with it up to something like 62 bits.

James has code to ingest said results.[/QUOTE]

You might be able to re-jigger the output from Factor5 to convert the results to the standard Primenet format.

Before charging into battle it might be worth figuring out if it's possible to get a more accurate idea than what's been going on in the other thread. This might require access to the raw data because it's a bit finicky.

You're looking for exponents which have been [B]cleared[/B] by TJAOI. If they cleared an exponent, it means that they found a missing factor because TJAOI is currently nowhere near the current wave front.

AND / OR

You're looking for exponents were cleared but for which a smaller factor has been found by TJAOI.


If you can sift through the data (I have no idea what it looks like, mind) and find where TJAOI found a factor AND his factor is the smallest OR the only factor, I think you've got it.


From there perhaps the easiest way would be to plot each case on a graph where the X-axis is the exponent and the Y-axis is the bit size? My assumption would be that we'd expect a random looking scatter of points if every range of exponents and bit levels is equally vulnerable. Someone more knowledgeable than me would be able to come up with a better "expected" distribution.

Uncwilly has a long history with the 100M digit range. I can't say what his particular interest is, but it is no surprise that he is interested in that range.

[QUOTE=TheMawn;392826]You're looking for exponents which have been [B]cleared[/B] by TJAOI. If they cleared an exponent, it means that they found a missing factor because TJAOI is currently nowhere near the current wave front.[/QUOTE]I am thinking about double check the TF done on exponents that have not been cleared, that are in the 100M digit range, before TJAOI gets to them. My goal would be to clear them sooner, rather than later. There are quite a few people running LL's in that range. I am trying to clear as many as possible before they are assigned. Many workers don't do enough TF before starting on years worth of LLing.

Since Factor5 would be different software it should be a good double check of the TF work.

[QUOTE=ET_;392511]The results from GMP-ECM give errors when manually submitted. Is there a possibility to adjust it?[/QUOTE]There is, but probably not until next week when I get back home. If you can [EMAIL="james@mersenne.ca"]email me[/EMAIL] a sample of the unrecognized line(s) that would be perfect.

[QUOTE=Uncwilly;392807]Does the current server all for manual submissions from Luigi's (ET_) Factor5?
James has code to ingest said results.[/QUOTE][QUOTE=TheMawn;392826]You might be able to re-jigger the output from Factor5 to convert the results to the standard Primenet format.[/QUOTE]You [i]could[/i] fiddle the results into a different format, but it would be better if you didn't, so that the results are recorded accurate -- if it was factored by Factor5 rather than Prime95 or mfkatc, that's probably interesting to know. If you find something isn't processed correctly when submitting manual results, please email me about it. I might have already seen an automated error report about it, but I'd rather know about it twice than not at all.

[QUOTE=pdazzl;393257]Is it possible to have a "recent factors" option as well? It was nice to see all the results minus the "no factor found" lines.[/QUOTE]

Seconded!

Is there anyway to change the recent cleared page to include an option to include all factors and not just first factors?It is very nice the way it is now but seeing expos with newly found factors is of interest to me.

[QUOTE=flagrantflowers;393270]Seconded!

Is there anyway to change the recent cleared page to include an option to include all factors and not just first factors?It is very nice the way it is now but seeing expos with newly found factors is of interest to me.[/QUOTE]

I'm not sure, but I think that info is included in the recent results (as opposed to recent cleared) report. I know one of the problems was that report only showing the past XX results (3000 was it?), and if more than that came in during that 60 minute time period, well, it's FIFO, so you only get the 3000 most recent or whatever.

I got a little busy the past couple weeks with my real job so I haven't had time to work on my project of having daily XML files showing that same set of results that got checked in, but eventually I will have that in place. Then you could grab a single XML, containing everything for that day, and break it down however you like.

A bug: when a factor is found by trial factoring (and possibly other methods), the expired dates for ALL uncompleted LL/DC assignments are displayed as the date the trial factor is found, even if those assignments have expired before.

Here are some examples: [url=http://www.mersenne.org/report_exponent/?exp_lo=M69625627&full=1]M69625627[/url] [url=http://www.mersenne.org/report_exponent/?exp_lo=M69419309&full=1]M69419309[/url] [url=http://www.mersenne.org/report_exponent/?exp_lo=M68978089&full=1]M68978089[/url] [url=http://www.mersenne.org/report_exponent/?exp_lo=M68949029&full=1]M68949029[/url] [url=http://www.mersenne.org/report_exponent/?exp_lo=M68850539&full=1]M68850539[/url]

[QUOTE=Mark Rose;393763]A bug: when a factor is found by trial factoring (and possibly other methods), the expired dates for ALL uncompleted LL/DC assignments are displayed as the date the trial factor is found, even if those assignments have expired before.

Here are some examples: [url=http://www.mersenne.org/report_exponent/?exp_lo=M69625627&full=1]M69625627[/url] [url=http://www.mersenne.org/report_exponent/?exp_lo=M69419309&full=1]M69419309[/url] [url=http://www.mersenne.org/report_exponent/?exp_lo=M68978089&full=1]M68978089[/url] [url=http://www.mersenne.org/report_exponent/?exp_lo=M68949029&full=1]M68949029[/url] [url=http://www.mersenne.org/report_exponent/?exp_lo=M68850539&full=1]M68850539[/url][/QUOTE]

Hmm... yeah. The old exponent report page did that too so I guess that carried over. It might be an easy fix, or maybe it's actually doing that in the database when that factor gets checked in, in which case that's outside the parts I would tinker with.

EDIT: Ah, just checked... those are the dates in the database itself, so yeah, it's doing that when the exponent is checked in with the factor. That'll require a, hopefully easy, fix. Something to the effect of only updating/expiring other assignments if they weren't already expired. But yeah, that'd be something I'd let George or James tackle because I wouldn't trust myself to not mess it up entirely.

So I just got my first email about a suspicious result. I really love that functionality.

How hard would it be to have a follow-up email when the result is verified telling you if your result was correct or incorrect? I looks like one of Chris's machines grabbed the exponent in question [URL="http://www.mersenne.org/report_exponent/?exp_lo=33899219&exp_hi=&full=1&ecmhist=1"]33899219[/URL] almost immediately so in a few days if I remember I can check back just to see. On the other hand, if it was given to someone who needs a bit more time (month) to do it, it might be nice to be reminded.

Thoughts?

[QUOTE=TheMawn;394841][URL="http://www.mersenne.org/report_exponent/?exp_lo=33899219&exp_hi=&full=1&ecmhist=1"]33899219[/URL][/QUOTE]

Evidently the question of good or bad has been resolved.
I too have puzzled over the notion maintaining a list of LL results to later inquire as to any departure from nominal.
I reckoned that it was not worth the effort to bother about it.
Did you receive any followup message, declaring your innocence for this instance?

[QUOTE=TheMawn;394841]So I just got my first email about a suspicious result. I really love that functionality.[/QUOTE]

I didn't know Primenet offered such notification. Cool. :smile:

As someone who uses mprime to ensure the computers I'm responsible for are "sane" (which is why I only do DCs), it would be great if there could be a report (and, optionally, emails) which details which results are "Suspect" vs. which only didn't match a previous LL (or both).

Important metrics for those who are responsible for 24/7/36[56].