gpuowl has generated 3 proof files but no results.JSON.txt

[Reed_Young]

Honestly, I had simply hoped to save the next person a similar amount of time to what I spent finding the helpful info you provided.

More than twice that much time has already been wasted since you provided me the solution, so this will be my last try to clarify.

Originally Posted by kriesel

Gpuowl puts json formatted results output in its own results.txt.

That information alone was sufficient to resolve the misunderstanding. That is a fact. A direct logical implication is that the same information also would have been sufficient to prevent it, had the same text been included in any of the documentation. That is a fact. The fact that GPU Owl puts JSON-formatted results into files named 'results.txt" rather than "results.JSON.txt" is not included in any of the documentation. That is a fact.

Whatever you choose to imagine about ulterior motives for stating those facts would not alter their factual nature, even if those speculations were correct. That is also a fact. So that subject is irrelevant and I have ignored it accordingly.

Thanks again for the clarification you have provided me about how to manually upload proof files.

[kriesel]

Originally Posted by Reed_Young

that GPU Owl puts JSON-formatted results into files named 'results.txt" rather than "results.JSON.txt" is not included in any of the documentation. That is a fact.

Nonsense. Please do not again post provably (and already-proven) false statements or claim they are facts.

Quoting again, this time from gpuowl v7.2-129's readme.md:

## Files used by gpuOwl
* `worktodo.txt` : contains exponents to test, one entry per line
* `results.txt` : contains the results
* `N.owl` : the most recent checkpoint for exponent <N>; will resume from here
* `N-prev.owl` : the previous checkpoint, to be used if N.ll is lost or corrupted
* `N.iteration.owl` : a persistent checkpoint at the given iteration

"* `results.txt` : contains the results" That's all result lines from the gpuowl application, whether json formatted or not.
Prime95/mprime uses many filenames that other applications do not; results.bench.txt, <exponent>.bu, .bu1, .bu2, etc. Other applications' authors are under no obligation to tutor users to not expect identical behavior or file names. Different applications are different.
Gpuowl documentation also does not specify that gpuowl does not do ecm computations, or P+1, or PRP-Cofactor computations, or Pepin tests on Fermat numbers, or ... And that is fine, for those that read what is there and understand what is not there is likely not implemented.

[Reed_Young]

You already said it yourself, the documentation "does not contain the string 'json' anywhere."

Originally Posted by kriesel

-results <file>    : name of results file, default 'results.txt'

Note how that does NOT say results.json.txt.
And does not contain the string "json" anywhere.

But the web pages for uploading proof files do "contain the string 'json'" and one of those two is emphatic that a specific file name is the only acceptable file name for upload. I'm not calling for heads to roll. I'm just pointing out that the source of my misunderstanding is one simple fact which is not directly stated anywhere in the documentation.

That is a fact.

Originally Posted by kriesel

"* `results.txt` : contains the results"

Agreed, the readme file says that much. But it does not say the next thing, which for some reason you appended within the same quote tags:

Originally Posted by kriesel

That's all result lines from the application, whether json formatted or not.[/strikethrough]

That is a fact, but it is a fact that is not in any documentation, unless you count forums as documentation. I do not. Do you? That would be an opinion, so we could agree to disagree about that.

[kriesel]

Originally Posted by Reed_Young

You already said it yourself, the documentation "does not contain the string 'json' anywhere."

But the web pages for uploading proof files do "contain the string 'json'" and one of those two is emphatic that a specific file name is the only acceptable file name for upload.

And to attempt to justify your unjustifiable position, you leave out the fact a web page clearly states TWICE that it is specific to mprime/prime95, a DIFFERENT APPLICATION THAN GPUOWL. Which I've previously pointed out to you. See https://mersenneforum.org/showpost.p...62&postcount=8
Just stop already. Stop posting falsehoods. Stop misleading by omission of relevant content.

Agreed, the readme file says that much. But it does not say the next thing.

It also does not define common English words used in the documentation. Or teach reading comprehension. An understanding of the language, and good faith participation, is assumed. Which may be an invalid assumption regarding some users.

which for some reason you appended within the same quote tags:

That, also, is just not so. See https://mersenneforum.org/showpost.p...9&postcount=13

Moderators: how much longer are you going to allow of Reed's misleading posts?

As for me, Reed has now earned a place on my ignore list.
And he should really stop trying to run LLDC. His bad residue rate is triple or more the project average, at 4/63 ~6.35% of attempts. https://www.mersenne.org/report_LL/?user_id=reed_Young
Also, do adequate TF and adequate-bounds P-1 before PRP, especially 100Mdigit PRP attempts!

[Mark Rose]

I've always wondered why json results are not in a file called results.json and why the extraneous .txt is present.

[Reed_Young]

Originally Posted by kriesel

... very persistent error. (An exceedingly rare one, I don't recall seeing before in YEARS of forum activity.)

There is some ambiguity about what exactly you are assuming about the number of times an error is encountered based on the number of times it is reported, if anything. But if that was not the point of that remark, then what was?

In any case, interpreting absence of evidence as evidence of absence is nothing to be ashamed of in the general case, whether that is what you did there, or not. It is a very common, very easy logical fallacy to commit without intention, especially when emoting based on a narrative that was constructed (even in part) based on prior experience, rather than taking the current interaction at face value and on its own terms, and thinking carefully about it without bias.

As I see it, the bias that would best fit your errors with me in this conversation is exceedingly common among IT support staff, and has also been promulgated literally for decades via pop culture trash like Dilbert and Internet memes like the one about the CD-ROM tray being used as a cup holder, so I don't take it personally. Being so alert to bias is nobody's job, so almost nobody bothers and I'm not bothered by that absence of effort any more. Just disappointed.

Originally Posted by Reed_Young

Honestly, I had simply hoped to save the next person a similar amount of time to what I spent finding the helpful info you provided.

More than twice that much time has already been wasted since you provided me the solution, so this will be my last try to clarify.

As promised, I am no longer trying to clarify my original message, just pointing out errors in your reasoning now, to show you that it was your own incorrect assumptions that led you to waste so much time after I informed you that the problem was solved.

Originally Posted by kriesel

You had MANY opportunities to detect that the error was yours, and still seem to be denying it.

You are "fixated" on that. I am not, and if you think about the next comment that I quote below, half as much as you have emoted on the subject of your fixation, you will see that I did acknowledge that - immediately, even. Then I moved on, and provided additional information, in an attempt to be helpful about a topic that's actually relevant: the text which, when read literally and without bias or prior knowledge of software other than the software package being used, also contributed to the error.

But perhaps multiple causality, which scientists must consider in order to achieve successful understanding of many interesting real world phenomena, is too subtle a concept for some mathematicians. That would completely explain the fixation on the matter of blame.

Originally Posted by Reed_Young

Oh, I see. Applying that information has solved the problem. Thanks.

Since the solution was achieved by adjusting my actions, what you claim I was denying, in fact I admitted right away. That acknowledgement was only implicit, but then it is also directly implied by logic, and as you consider yourself skilled enough at drawing inferences from text that you presume to tell me my own motives, then surely, implicit acknowledgement was more than sufficient for you to have detected that acknowledgment.

Originally Posted by kriesel

Ego defense is a stubborn thing.

Or was that a confession that ego defense has been your own motivation throughout this conversation? I think so, albeit probably unintentional.

[Reed_Young]

Originally Posted by kriesel

... LLDC. His bad residue rate is triple or more the project average, at 4/63 ~6.35% of attempts. https://www.mersenne.org/report_LL/?user_id=reed_Young

Only 1 of those is more recent than 2010, and the 1 that occurred 2023-01-04 was due to a cooling issue that has been resolved. Still, thanks for trying to be helpful!