Factoring database

[richs]

I have run into an unusual corruption in the database in Aliqueit sequence 845642 which I have been working under the subproject.

I completed my subproject work to C103 at index 985 and transferred the factors back to the database by the Aliqueit program. The elf file is correct and will be posted in the subproject thread.

Index 775 shows the two large prime factors with a superscript 2 (indicated squared I assume) and index 776 then shows a C165. No further sequences are shown.

[EdH]

Originally Posted by richs

I have run into an unusual corruption in the database in Aliqueit sequence 845642 which I have been working under the subproject. . .

I think that should be 845652. (Maybe you still have edit time, like I do?)

Sorry, if I should have kept out...

An interesting iteration...

[10metreh]

That's what "Repair Sequence" is for. One click and it's fine.

[Mini-Geek]

Syd, please see the latest discussion regarding the mysterious aliquot sequences worker:
http://www.mersenneforum.org/showthr...697#post214697
Could you add the notice we're talking about please? (this:)

Originally Posted by EdH

I would still like to see a message on the sequence page that invites reviewers to the forum. . .

Also, has all this (all aliquot sequences from 850k-880k, and other work by user(s) unknown) been done by a logged-in or an anonymous user? Do you know who it is? Do you know how we might be able to contact this person, besides a notice on the DB?

[Syd]

Originally Posted by Mini-Geek

Syd, please see the latest discussion regarding the mysterious aliquot sequences worker:
http://www.mersenneforum.org/showthr...697#post214697
Could you add the notice we're talking about please? (this:)

Also, has all this (all aliquot sequences from 850k-880k, and other work by user(s) unknown) been done by a logged-in or an anonymous user? Do you know who it is? Do you know how we might be able to contact this person, besides a notice on the DB?

The user was not logged on, all I have is an IP, if thats helps.

[mdettweiler]

Originally Posted by Syd

The user was not logged on, all I have is an IP, if thats helps.

That might still be somewhat useful--an IP can be traced to a specific region, which may give some clue as to the mystery worker's identity, and whether he's truly not heard of this forum, or actually someone already present here who's going incognito. Given the mystery worker's propensity for striking sequences right after they're reserved, I'm somewhat inclined to believe the latter. (Either that, or there's two mystery workers, one benign and one not, but either way an IP will be helpful to clear things up.)

Edit: ah, I just saw this post which clears things up a bit. Never mind.

[EdH]

I have not as yet registered with the db. Would I gain extra abilities? Should I be consistent with this ID?

On the Sequence Overview page for Aliquot Sequences, the current size is not provided. Is there a way to request this for the group or must you review each individually? (Possibly available if I register?)

Thanks.

[schickel]

Originally Posted by EdH

I have not as yet registered with the db. Would I gain extra abilities?

When Syd first set up the registration process it wasn't automatic to be granted "extra permissions". He gave all (most?) of the members of Mersenneforum the extra permissions. It allowed you to set the status of a number as P or PrP if you knew which it was. This helped out the DB by causing it to skip doing the test itself. This did in fact help out until someone started (randomly? maliciously? Only the perpetrator knows for sure.) marking numbers as P. Needless to say, that wreaked havoc with pretty much everything, but especially sequences that depend on the primality of a particular number (Aliquot and Home Prime sequences especially.)

Should I be consistent with this ID?

It doesn't matter, really, whatever is most convenient for you.

On the Sequence Overview page for Aliquot Sequences, the current size is not provided. Is there a way to request this for the group or must you review each individually?

That has been a frequently requested feature around here.

(Possibly available if I register?)

No, it's not in there, but now maybe now that Syd is back online (and, most importantly, said that he has some free time.....)

Thanks.

You're welcome.

The other advantage to registering was the ability to request some higher levels of ECM work be done by the DB. Points were required to do this which were earned by running a "worker" script on your local computer to do ECM and/or msieve work. Unfortunately, the Windows incarnation of the script didn't work as expected.....work was sent out, statuses (statusi?) were returned, but no actual work was done....not real helpful there.

[Mini-Geek]

Originally Posted by Syd

Now as I have Internet connection & lots of time again, I decided to rewrite some/most parts of the scripts and finally make it thread-safe. I'll report back here when its ready for use, maybe with the worker-script again, I dont know yet.

To start off from a clean state I'll reset all numbers to the state "U" for Unknown and check them all again, so dont worry if it behaves strange the next days. Just in case, made a backup on external disk. 9.5GB bzipped!

Thanks to everyone reporting all those bugs, collecting broken sequences and offering help!!

Syd (Markus)

The composite bugs still exist (as this list that Batalov prepared still shows). The reset to U either never happened, or didn't help. Can you please fix these problems? Even if you don't have a permanent solution ready yet, please at least fix the bugged composites below 100 (preferably 1000).

[EdH]

Has anyone heard from Syd since he posted about being back on line with lots of time to put into the db?

I've not noticed any repairs to the db and still miss things like the graphs.

I wonder if the time suddenly evaporated, the connection wasn't permanent or the rewrites have been slower than expected. I thought perhaps the 57 and 159, etc. bugs would have been an easy fix, but I have no knowledge on which to base that assumption.

[Andi47]

Originally Posted by EdH

Has anyone heard from Syd since he posted about being back on line with lots of time to put into the db?

I've not noticed any repairs to the db and still miss things like the graphs.

I wonder if the time suddenly evaporated, the connection wasn't permanent or the rewrites have been slower than expected. I thought perhaps the 57 and 159, etc. bugs would have been an easy fix, but I have no knowledge on which to base that assumption.

Hmmmm... According to his user statistics, he seems to look into mersenneforum every few days (last visit was yesterday).

@Syd: I know it takes quite some time to completely revamp and debug a program, but I want to just ask some questions here:

* Do you still have time to work on the DB?
* What is the current status? (even if you did much work in the background, very little is visible in the foreground yet, so we have almost no info)
* Can you please manually fix the "small composites" (57, etc.) which are blocking several aliquot sequences in the DB? (e.g. sequence 10212 and many more...)