Primenet web design - Page 32

snme2pm1 · 2014-11-02, 01:39

Quote:

Originally Posted by garo

large volume of results in the TF-LMH range that means one usually cannot see more than a couple of hours of recent results.

A couple of hours sounds like a generous description compared to many instances that I've seen.
Not that I'm complaining.
This awkwardly used term LMH defies usefulness in contemplating fitters.
I found the recently introduced provisions for sorting recent results useful, and I have now tended to sort by exponent, then look at my region of interest.
Yes, that doesn't detract from potential masses of outer range material that bumps information of more near period interest.

snme2pm1 · 2014-11-02, 04:11

Quote:

Originally Posted by Prime95

I'm on a cruise ship

I expect that you know we are all doing our utmost.
Should we expect to see any of your leisure time pictures from cruise posted any time soon?
Or do we need to wait an appropriate time?
Or you might not particularly want further exposure, of one kind or another.
Regardless your photo is "out there", as it was at some time past.
p.s. You might consider a new photo stance.

Uncwilly · 2014-11-02, 06:37

Quote:

Originally Posted by Prime95

Yes, you folks have failed to find a new Mersenne Prime while I'm on vacation -- most disappointing. You have a week left to rectify the situation

Didn't know that you where away. I have one all teed up and ready to submit.

Madpoo · 2014-11-02, 06:46

Quote:

Originally Posted by Prime95

I can see value in this idea. It is trivial to create two (or more) canned hourly reports. We could also output the hourly data to a temporary table rather than an ugly text report. From the temporary table, we could then produce pretty HTML tables.

Also, I don't see why a dynamic (instead of hourly) recent results report should create a terrible server load. There is an index on date-received, so SQLServer should be able to execute any recent results query efficiently (even a non-LMH report that tosses 90% of result rows because they are LMH results) -- we should run some sample queries to test that out.

I had a little nugget of an idea floating around in my brain today as a result of this topic...

What I came up with so far was the idea of spitting out a *daily* file of some kind that has all of the results checked in for that 24 hour period. People could load that up and have fun with it.

Too much data though? This past October, for example, has 2.24MM results over it's 31 days. Yesterday (Halloween) was 80,147 results, for instance, but if we look at Oct as a whole it averaged 72.3K results per day.

Those 2.24MM results for October were for 2.08MM exponents, so presumably some of that were multiple check-ins like "no factor from bit xx to bit xx".

Again, I guess it kind of comes down to what people are doing with this info, as far as what to include, etc.

Like, if you're checking this to look for some particular exponent, then of course the exponent report is more suitable. But if you're trying to snag the hourly copy of this data to import into your own data format, then yeah, if you miss an hour here or there because that job didn't run, then you'll miss some. In fact, since it's limited to the 3000 most recent, if there are more than that many in any one hour period (or certainly more than 72,000 per day, like yesterday had) then you'll miss some.

That's why I thought exporting the data in 24-hour blocks of raw data (csv, or XML) seemed good. You won't miss anything, historical copies of the file would be there (we could go back and generate these files using the existing data, as far back as that table allows), and because it's canned, it won't cause much server load beyond bandwidth.

If I had to estimate (well, a REALLY GOOD estimate since I just looked at Oct 31 data), it'd be roughly 11 MB of raw text, but of course it compresses really well to a 2.1 GB gzip or zip.

XML would add some markup but also compress really well... anyone doing offline analysis for whatever could just grab the raw details, and tease out the stuff they're interested in. I have a hunch that's what people are doing already using the less than optimal solution of pulling it from a table.

Of course if there's something that people in general would like to see on the website itself, we should try to show that somehow, but I'm just thinking that the "feed" of results being checked in is kind of a lot of stuff due to it's ephemeral nature, and people might be more interested in the exponent-centric instead of date-centric reports?

I could be wrong though. I often am. :)

James Heinrich · 2014-11-02, 12:35

Quote:

Originally Posted by Madpoo

That's why I thought exporting the data in 24-hour blocks of raw data (csv, or XML) seemed good. You won't miss anything, historical copies of the file would be there (we could go back and generate these files using the existing data, as far back as that table allows)

Make it so.
XML in 24-hour blocks would be great. Archiving reports by date (e.g. /report_recent_cleared/archive/20141101.xml) would be tremendous too, so that if either my server or mersenne.org is dysfunctional at some point I can always come back the next day to get the data I missed.

TheMawn · 2014-11-02, 19:17

Quote:

Originally Posted by Prime95

Yes, you folks have failed to find a new Mersenne Prime while I'm on vacation -- most disappointing. You have a week left to rectify the situation

We're on it, boss.

TheMawn · 2014-11-03, 04:07

Quote:

Originally Posted by TheMawn

We're on it, boss.

Well, my only chance at finding a prime in the next week turned out to be composite.

Honestly, I'd be finding loads of primes except that the server keeps giving me duds.

Mark Rose · 2014-11-03, 04:24

Quote:

Originally Posted by TheMawn

Well, my only chance at finding a prime in the next week turned out to be composite.

Honestly, I'd be finding loads of primes except that the server keeps giving me duds.

That's why I stick to trial factoring.

Madpoo · 2014-11-03, 21:00

Quote:

Originally Posted by Mark Rose

That's why I stick to trial factoring.

Well, your odds of finding the next Mersenne Prime by trial factoring is pretty slim, I'd have to say.

retina · 2014-11-08, 07:43

I think the Recent cleared report can be enhanced by aligning all the numeric columns to the right, i.e. the Exponent, Days, GHz-days and Result columns.

And the same for the recent results report.

Madpoo · 2014-11-08, 20:36

Quote:

Originally Posted by retina

I think the Recent cleared report can be enhanced by aligning all the numeric columns to the right, i.e. the Exponent, Days, GHz-days and Result columns.

And the same for the recent results report.

Okay, I tweaked a few things now in the reports for column alignments.

I also added handy links on the exponents in the recent results/cleared to go to the detailed stats for that exponent.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Database design	xilman	Astronomy	1	2017-04-30 22:25
Theoretical Experiment Design	c10ck3r	Homework Help	7	2015-02-03 08:54
Digital Logic Design	henryzz	Puzzles	9	2014-12-04 20:56
new intel design	tha	Hardware	5	2007-04-19 11:38
design factoring algorithms	koders333	Factoring	14	2006-01-25 14:08

2014-11-08, 07:43	#351
retina Undefined "The unspeakable one" Jun 2006 My evil lair 2²·3²·173 Posts	I think the Recent cleared report can be enhanced by aligning all the numeric columns to the right, i.e. the Exponent, Days, GHz-days and Result columns. And the same for the recent results report.