OFFICIAL "SERVER PROBLEMS" THREAD

[chalsall]

Originally Posted by kriesel

Well that seems rude and uncalled for. Perhaps I misunderstand. I can read, and sometimes remember what I read.

Dude! My sincere apologies for typing something that you interpreted as rude and uncalled for.

"Take responsibility for the listening into which you are speaking" was very good advice a good friend of mine gave to me many years ago.

Please do note the language you quoted is the Primenet API documentation from the site. First draft.

"As-built" might be slightly different...

[kriesel]

Originally Posted by chalsall

Please do note the language you quoted is the Primenet API documentation from the site. First draft.

I understood the version notes at the top to indicate it is the SEVENTEENTH listed, spanning nearly 4 years, and as far as I know, still the latest publicly available.

"RELEASE CANDIDATE VERSION 0.97(c)".

The extensions since this document version are in need of documentation, in one comprehensive document, available to developers at least, outside of what can be inferred from prime95/mprime source code, or https://www.mersenneforum.org/showthread.php?t=23992. Yes the mersenne.org public version is very stale, from late 2007.

I also note that in employing curl to send to the server manual GPU updates on occasion, it refuses v0.97, but accepts v0.95 in my attempts.
If there is nothing more current available, what are we would-be developers or users to go by?
I feel it badly needs updating, and repeat now my earlier offer to help do that.

[chalsall]

Originally Posted by kriesel

If there is nothing more current available, what are we would-be developers or users to go by? I feel it badly needs updating, and repeat now my earlier offer to help do that.

I hear you. And agree. When I work with the API I use the documentation largely as guidance, rather than definitive.

There are still many things I don't understand. And, frankly, there are many, many things that are not currently documented.

JSON, anyone? Already passing over the wire...

[kriesel]

666666667 false TF reports to 128, 1024 bits
987654323 to 128
811451933 to 80
811451947 to 74
333339953 to 2
333339907 to 2
333339887 to 2
333339883 to 2
333339847 to 2
333339823 to 2
333339701 to 2
333339647 to 2
333339619 to 2
666666323 to 32

[LaurV]

Originally Posted by kriesel

666666667

Wow! Some idiot messed with my exponent? I will find his computer and stab its wheels!...

[kriesel]

Originally Posted by LaurV

Wow! Some idiot messed with my exponent? I will find his computer and ...

You're a little late to the party. See also
https://mersenneforum.org/showpost.p...postcount=2420
https://mersenneforum.org/showpost.p...postcount=2424
and a couple other related threads.

Also, I have a large-bounds P-1 under way for M666666667, originally launched as a possible (~5% chance) way of refuting the miscreant's forged TF no-factor reports. Its previous P-1 runs have lower than optimal bounds, no stage 2, perhaps because back then enough RAM could have been costly or scarce.
And surely you know the numbers belong to no one.
(M666666667 is not reserved for P-1, since the manual assignment page refused to issue it.)
Not the most effective use of a GPU for a few days, but hey, it's all just for fun anyway, right?

[PhilF]

Originally Posted by kriesel

And surely you know the numbers belong to no one.

Oh no, I must disagree. That number is LaurV's. Now your computer is also at risk of getting its wheels stabbed.

[Uncwilly]

Originally Posted by PhilF

Oh no, I must disagree. That number is LaurV's. Now your computer is also at risk of getting its wheels stabbed.

[kriesel]

Originally Posted by PhilF

Oh no, I must disagree. That number is LaurV's. Now your computer is also at risk of getting its wheels stabbed.

Some of my systems have feet. None are on wheeled carts or stands. Anything here is subject to surveillance, and more.

[VBCurtis]

Originally Posted by PhilF

Oh no, I must disagree. That number is LaurV's. Now your computer is also at risk of getting its wheels stabbed.

[LaurV]

Originally Posted by kriesel

Also, I have a large-bounds P-1 under way for M666666667, originally launched as a possible (~5% chance) way of refuting the miscreant's forged TF no-factor reports. Its previous P-1 runs have lower than optimal bounds, no stage 2...

Somehow I totally agree with that. And, your resources, your time, your money. I would really like to see a huge factor for that monster.
Just for records, my TF and P-1 are not bogus, I did the job, and I am almost sure the results are right. I know you are not accusing me, but just to be clear. Your computer wheels are safe
Edit: and if somebody feels the urgent need to triple check my self-DC, the residues to keep the LL test on track, are posted here. Now, enough off-topic (and sorry for it!), this thread is about server bugs.