mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   mersenne.ca (https://www.mersenneforum.org/forumdisplay.php?f=173)
-   -   mersenne.ca (https://www.mersenneforum.org/showthread.php?t=23051)

James Heinrich 2020-06-26 19:51

[QUOTE=kruoli;549148]I looked up [url=https://www.mersenne.ca/exponent/3643]M3643[/url] and it shows an actual TF limit of 2[SUP]209[/SUP][/QUOTE]Ah, that's something different. My previous fix was for the spurious entry under Trial Factoring Results.
This was an entirely different bit of archaic code doing unexpected things. Which has also now been pulled aside and told to stop its nonsense.
Did I miss anything else? :smile:

kruoli 2020-06-26 20:32

At least I can confirm my problem is now fixed (after Ctrl+F5, really necessary!). Thanks! :smile:

snme2pm1 2020-06-30 02:26

[QUOTE=snme2pm1;546789]I guess that we're not connecting on this issue.
[/QUOTE]

Another instance confirming the prior mentioned behaviour.
Consider 118078879 factor found and lodged 2020-06-28 23:12.
The .ca information reflects a 75 bit result, whereas in reality it would have better been disposed as factor found.
Some order of daily operations might be at issue.

Uncwilly 2020-06-30 02:34

CTRL-F5.

I see it here. (Factor for 118078879, not 118078879 as a factor)

James Heinrich 2020-06-30 02:45

[QUOTE=snme2pm1;549390]Another instance confirming the prior mentioned behaviour.
Consider 118078879 factor found and lodged 2020-06-28 23:12.
The .ca information reflects a 75 bit result, whereas in reality it would have better been disposed as factor found.
Some order of daily operations might be at issue.[/QUOTE]I'm not sure what your question is or what you think should be happening that's not?
[url]https://www.mersenne.ca/exponent/118078879[/url]
It shows your factor (found, as you say, at 2020-06-28 23:12) under both Known Prime Factors and Trial Factoring Results.
Although the time you posted your message (approx 100 minutes ago) was slightly before the server even started processing the daily export record, and it takes anywhere from 30-90 minutes to process (depending on how many results were submitted that day). So you're clearly checking too early. Assume as rule that results will not show up on mersenne.ca until 0600h UTC the next day.

snme2pm1 2020-06-30 03:04

[QUOTE=James Heinrich;549393]So you're clearly checking too early. Assume as rule that results will not show up on mersenne.ca until 0600h UTC the next day.[/QUOTE]

Yes I have noticed that the .ca view is updated at some time like that.
Time zones are a nuisance! But I'm talking about the prior period.
Yet I an vacant for an idea as to why 118078879 was regarded at .ca as 75 bit result rather than factor found, according to the view of [url]https://www.mersenne.ca/status/tf/20200628/0/4/11800[/url]

James Heinrich 2020-06-30 03:14

[QUOTE=snme2pm1;549394]Yet I an vacant for an idea as to why [url=https://www.mersenne.ca/M118078879]118078879[/url] was regarded at .ca as 75 bit result rather than factor found, according to the view of [url]https://www.mersenne.ca/status/tf/20200628/0/4/11800[/url][/QUOTE]Because, at the time the report is produced (immediately after the mersenne.org data is imported) the day's factors will likely not all have finished being imported into mersenne.ca -- when the mersenne.org results are parsed the day's new factors are inserted into the processing queue and ingested over the next minutes or hours (depending how many factors are in the queue and how many factors were discovered that day). Processing factors (checking that they're prime, checking they belong to the claimed exponent, etc) takes some time. Not much time, but when there are thousands (sometimes tens of thousands) of new factors it adds up, so they're processed on an as-soon-as-feasible basis in the factor queue. This may lead to the situation you describe where a factor found late in the day is not yet imported by the time the status report is generated. But it will be by the next day.

snme2pm1 2020-06-30 03:33

[QUOTE=James Heinrich;549396]This may lead to the situation you describe where a factor found late in the day is not yet imported by the time the status report is generated. But it will be by the next day.[/QUOTE]

I doubt that Arron would willingly agree as to a lapse in processing capacity, and the lodging of results was not particularly late in the day (not last minute as previously).

snme2pm1 2020-06-30 06:21

[QUOTE=James Heinrich;549396]Because...imported into mersenne.ca -- when the mersenne.org results are parsed the day's new factors are inserted into the processing queue[/QUOTE]
Are you suggesting that your input is from multiple channels?
As a programmer I always hate the potential for multiple definitions for anything.

James Heinrich 2020-06-30 13:48

[QUOTE=snme2pm1;549397]lodging of results was not particularly late in the day (not last minute as previously).[/QUOTE]23h12m is later than 95.8% of the rest of the day's results, therefore firmly to the end of the process.[QUOTE=snme2pm1;549397]I doubt that Arron would willingly agree as to a lapse in processing capacity, and the lodging of results was not particularly late in the day (not last minute as previously).[/QUOTE]Not sure what Aaron has to do with anything?
[QUOTE=snme2pm1;549402]Are you suggesting that your input is from multiple channels?[/QUOTE]What multiple channels? Data <1000M comes from mersenne.org, >1000M is submitted by users directly. So I suppose two non-overlapping channels, yes.

James Heinrich 2020-06-30 17:30

[QUOTE=James Heinrich;549396]Because, at the time the report is produced (immediately after the mersenne.org data is imported) the day's factors will likely not all have finished being imported into mersenne.ca[/QUOTE]I have added some complexity and separated out the report generation into a separate job that will attempt to run 1-2 hours after the data import job, but it will check the factor queue and wait until all of yesterday's factors are processed before compiling the report.


All times are UTC. The time now is 22:47.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.