How does LMH work now?

[JHagerson]

Maybe this was discussed at length somewhere else, and if so, I apologize for failing to find it.

In the past, I download a file and use the NOFACTOR file and extract factoring assignments from it. However, I now understand that NOFACTOR is no longer available.

I do have a utility that processes my results.txt file into a worktodo file to take the unfactored exponents in the range I have been testing to one additional bit of factor depth. However, what do I do if I just want to answer the call and factor a range of exponents I haven't been working on already?

Thank you very much for your assistance.
JHagerson

[markr]

Primenet v5's Factoring Effort page provides factor depth info. It can report up to 50000 exponents with the simple text option, which also makes the report closer to worktodo format.

You can also submit results to the v5 server at manual reporting page and they get into the v5 database automatically - nice! It doesn't seem to matter if the results are from the v25 client or an earlier version. They give v5 credit if you're logged in. You can't get LMH assignments from the v5 manual tests page, though.

You can do LMH work with the v25 client, too, either primenet or not. If you let the client report to primenet v5, there is a way to add your own assignments. Or just choose the LMH work-type (for example, it's currently giving 162M exponents from 2^57 to 2^60) but of course that's not the same as working through a block you've picked yourself.

Wow - that's much more than you asked for!

[lycorn]

Can you confirm that results obtained with v24 (in this case ,"classical" LMH range 74-74.5M to 64 bits) may as well be uploaded to the v5 server via the manual form and will get credited to my v5 account?
Thx

[tha]

Originally Posted by lycorn

Can you confirm that results obtained with v24 (in this case ,"classical" LMH range 74-74.5M to 64 bits) may as well be uploaded to the v5 server via the manual form and will get credited to my v5 account?
Thx

I think it should work. I use v25.6 but I believe the output format is the same anyway. Just try it, if it doesn't work you can still use any other way.

[markr]

It has worked for me, so far. I've submitted tf, p-1 & LL, and received the credit.

[lycorn]

Thk you both for the answers.
In a day or two I´ll have a large batch of results to submit, so instead of mailing it to GW as usual, will try to upload the file.

[Graff]

Originally Posted by lycorn

In a day or two I´ll have a large batch of results to submit, so instead of mailing it to GW as usual, will try to upload the file.

Be careful of the file size. I used to mail batches to GW every 5 days or
so. That would be over 5000 results. He reported that it often took him
several attempts to submit those batches via the manual reporting
form due to the script timing out. Since I moved to reporting the
results myself, I do it every day, so that means only 1000 results at a time.
It takes the script about 60s to digest that number of entries.

[lycorn]

Originally Posted by Graff

Be careful of the file size. (...)
It takes the script about 60s to digest that number of entries.

Thanks for the hint. I have around 12K exponents to submit and will take that into consideration. Do you think it´s better to upload files or to paste results in the window?
BTW, are you still crunching for LMH? You used to be very active in the forum but I haven´t seen you around for a while.

[Graff]

Originally Posted by lycorn

BTW, are you still crunching for LMH? You used to be very active in the forum but I haven´t seen you around for a while.

When all the available ranges to 2^63 under 79.3M were
assigned, I moved to higher ranges. I started at 79.3M, working
upwards in 1M increments and am currently doing 87.0-88.0M
(about 60% done).

A problem is looming. The V5 server indicates that there are
2000+ exponents in my intended next range (88.0-89.0M) assigned.
There are also 1800+ exponents in the 90.0-91.0M range assigned.
These are presumably not being assigned automatically by the V5
server, but are being requested by LMHers. But I don't see an
easy way to request a large range. In addition, I do my LMH
runs on a non-PC/non-Prime95 setup, so that will probably mean
that I can't use the method used by the 88/89 and 90/91 range
LMHers.

I could just move to the 89.0-90.0M range, but I would be worried
that exponents in that range would start being assigned by the
server.

What are other LMHers doing?

[ckdo]

Originally Posted by Graff

The V5 server indicates that there are
2000+ exponents in my intended next range (88.0-89.0M) assigned.
There are also 1800+ exponents in the 90.0-91.0M range assigned.
These are presumably not being assigned automatically by the V5
server, but are being requested by LMHers.

As for the 90-91M range... that's me; I'm currently taking the range to 2^64. I'll go deeper later. Please skip that range.

As for 88-89M... Chris Halsall is taking the range to 2^63. I don't know if he wants any help, but you should probably skip that range as well.

Since you don't use Prime95, the best you can do is announce what range you are taking and hope noone decides to take the same range. It might be helpful to reserve some exponents (say, the first 100) using a Prime95 machine so that the project summary shows someone (and who) is working there.

I don't think there's any good reason to be worried about Primenet assigning exponents in the 79.3-100M area other than by explicit request, yet...

Cheers,

Carsten

[Graff]

Originally Posted by ckdo

As for the 90-91M range... that's me; I'm currently taking the range to 2^64. I'll go deeper later. Please skip that range.

As for 88-89M... Chris Halsall is taking the range to 2^63. I don't know if he wants any help, but you should probably skip that range as well.

Will skip those two ranges and take 89-90M to 2^63.

Originally Posted by ckdo

Since you don't use Prime95, the best you can do is announce what range you are taking and hope noone decides to take the same range. It might be helpful to reserve some exponents (say, the first 100) using a Prime95 machine so that the project summary shows someone (and who) is working there.

Can we get these > 79.3 M ranges incorporated into the "All Available
Ranges" thread? Listing spoken-for ranges there will probably give better
visibility than posting in the LMH > 79.3M sub-forum. Alternatively, can
we create a sticky thread in the sub-forum?

Graff