[QUOTE=bchaffin;236502]Interesting. I ran msieve for about 28 hours, and got 85 polys. The best was:
# norm 4.742410e-14 alpha -7.760362 e 1.030e-11 rroots 5

How do I know what's good? Batalov says his looks "ok-ish" but I don't know how you make that judgement... so much to learn...[/QUOTE]
Yours is good. Mine was so-so. (I later started another range, and again hardly got one poly with 9e-12; parameters should be relaxed, and the flow of candidate polys should be more abundant. In this particular range of size.)

To make the call you want to have many old results and then put the new dot in the old cloud of dots. There's a theoretical trend, and then there's a practical trend; the practical trend lags behind theoretical in the high range because the search space for polys is visited relatively much less and as a result "sub-optimal" polys serve as "practical optimals".

At last, linear algebra has started on the c145 from 4788.2553. It took 41.5M relations. ETA is about 11 hours from now.

The c145 is done and in the DB:

prp61 factor: 7765498352298376284127611386595578258168693786447736853699799
prp84 factor: 573341045324244764104938059707693492361347345455901260755484457238332565967911859343

The next term decreases slightly, and has a c136. Aliqueit continued running overnight, so I've already done some ecm on it: 74@11e3, 214@5e4, 430@25e4, 904@1e6, 2350@3e6, and 4480@11e6, and starting on @43e6. I've also got about 10 hours of polyfind done, with this best poly found so far:
# norm 3.605655e-13 alpha -7.479655 e 3.198e-11 rroots 5

I'm happy to leave it running but I don't want to hog all the fun, so if anyone wants to take over let me know. Any expert opinions on whether that poly is good enough would be appreciated; by default, polyfind will keep looking for another 24 hours or so.

The c145 took about 4.5 days? Then this one will take 1.5 days (approximation: take the ratio of the e values).

A rule of thumb for polyselecting is to spend no more than 10 (for large numbers) - 25 (at most for small numbers) % of the total time. So, you could switch to sieving by now (e.g. by killing the 'msieve -np' process).

a c141 now, or what?
why the long faces? it actually can flip any iteration now.

[QUOTE=Batalov;237729]a c141 now, or what?
why the long faces? it actually can flip any iteration now.[/QUOTE]

My resources are currently tied up with a GNFS-143 from sequence 10212, so I currently can't help out with this one.

600@11e6 done on c141

Sorry, I've been lax about posting updates... as you can see I added a few more terms, most were quick but there were a c120 and c138 in there. I ecm'ed the current c141 overnight, doing the normal sequence through 11e6, and a fair number at 43e6 (I overwrote the log files accidentally). I just started sieving a few minutes ago.

I've got 24 cores on it now, and I've discovered the joys of multi-client mode in factMsieve, which allows all the other cores to continue sieving while one is collecting the relations. Since the c145 took 4.5 days on 8 cores, I expect this one will be done some time tomorrow.

I finished sieving on the c141 yesterday but while restarting the linear algebra step to run multiple threads I screwed something up and deleted the .dat file so I had to redo it all. :( :( :( Anyway it's (re)done now and in the DB. Next term increased slightly, and still has 2^6, with a c143. I've started ecm and polyfind on it.

[QUOTE=bchaffin;238209]I finished sieving on the c141 yesterday but while restarting the linear algebra step to run multiple threads I screwed something up and deleted the .dat file so I had to redo it all. :( :( :( Anyway it's (re)done now and in the DB. Next term increased slightly, and still has 2^6, with a c143. I've started ecm and polyfind on it.[/QUOTE]Great job. Thanks for the assist.

I did something like that one time: I finished a job on one of my machines and started another. The second job ran a clean up script as the first step. Only thing is, when I went to copy the results of the first one off the thumbdrive, oops! I forgot to copy the summary file first. (Luckily I was able to recover the answer with an undelete utility, after a couple of hours runtime....)

[QUOTE=bchaffin;238209]I finished sieving on the c141 yesterday but while restarting the linear algebra step to run multiple threads I screwed something up and deleted the .dat file so I had to redo it all. :( :( [/QUOTE]
It is a rite of passage, man. Good job moving on, and thanks!