mersenneforum.org Prime Gap News
 Register FAQ Search Today's Posts Mark Forums Read

2020-10-15, 20:56   #177
mart_r

Dec 2008
you know...around...

32·71 Posts

Quote:
 Originally Posted by SethTro Can someone validate I understand the discoverer line? "PDG (- 8 maa 2016) | ET (+ 10/2012 vb PDG 6 nov 2012)" I take this to mean that PDG (Patrick De Geest) found the lower prime on 8th march(?) (may?) 2016 and ET (Edward Trice) found the next prime 10/2012 and it was verified by PDG 6th nov 2012. if next and previous discoverer are the same person it's obvious what to do (use their name at the later date). Any suggestions for how to credit it when the record was found by two people?

Like in the example in your last post, they should both be credited. So, looks good to me (that's what the acronym stands for, right?)
Also, "maa" = maart = March, "mei" = May.

 2020-10-17, 08:12 #178 MJansen   Jan 2018 43 Posts Goedemorgen Mart, good to see I am not the only one that knows their Dutch ;-) Mvgr michiel
2020-10-17, 08:19   #179
SethTro

"Seth"
Apr 2019

3·89 Posts

Quote:
 Originally Posted by mart_r Like in the example in your last post, they should both be credited. So, looks good to me (that's what the acronym stands for, right?) Also, "maa" = maart = March, "mei" = May.
How should I mark credit for the pair in allgaps.sql?
We currently have a 8 letter discoverer for each row. And I think it would be silly to create a new 'discoverer' for each pair (as some pairs have only a single discoverer together)

Imagining that RobSmith and I worked together on some MEGAGAPS would these be
1 'RobSmith -, SethTroisi +'
2. 'RobSmith&S.Troisi'
3. 'RobSmith,S.Troisi'
4. 'RobSmith S.Troisi'
5. 'RobSmith and S.Troisi'
6. 'RobSmith working together productively with S.Troisi for the finding of megagaps'

Last fiddled with by SethTro on 2020-10-17 at 08:22

2020-10-17, 13:26   #180
mart_r

Dec 2008
you know...around...

32×71 Posts

Quote:
 Originally Posted by SethTro How should I mark credit for the pair in allgaps.sql? We currently have a 8 letter discoverer for each row. And I think it would be silly to create a new 'discoverer' for each pair (as some pairs have only a single discoverer together) Imagining that RobSmith and I worked together on some MEGAGAPS would these be 1 'RobSmith -, SethTroisi +' 2. 'RobSmith&S.Troisi' 3. 'RobSmith,S.Troisi' 4. 'RobSmith S.Troisi' 5. 'RobSmith and S.Troisi' 6. 'RobSmith working together productively with S.Troisi for the finding of megagaps'

If my quick-and-dirty analysis was somewhere near correct, there wasn't a single case for a gap not yet in the list where two different discoverers other than Patrick were listed. So for example MLB-/MLB+ and PDG-/MLB+ could both be listed as MLBrown as discoverer in short and with mention of Patrick De Geest/BorderPRP in long text. Or, maybe "PDG__MLB" for the second example and "RS____ST" in your example above. But I'm running out of ideas...

Quote:
 Originally Posted by MJansen Goedemorgen Mart, good to see I am not the only one that knows their Dutch ;-) Mvgr michiel

I have it from some of the calendar notebooks I used to scribble into when I was a kid.
I'm not fluent in Dutch, but, Germany being a bordering country to Belgium and The Netherlands, I do recognize a good deal of words (like "met vriendelijke groeten" and such:).

2020-10-17, 22:47   #181
Bobby Jacobs

May 2018

22×53 Posts

Quote:
 Originally Posted by robert44444uk Code: 49896 Jacobsen 12243 RobSmith 10341 Rosnthal 8970 Toni_Key 5477 M.Jansen 2385 MJPC&JKA 1114 Gapcoin 898 MJansen 851 Spielaur 752 PierCami 509 S.Troisi 136 TorAlmJA 120 RP.Brent 116 Be.Nyman 113 MrtnRaab 94 M 80 TRNicely 72 LndrPrkn 71 Yng&Ptlr 70 TOeSilva 52 HrzogTOS 43 Glaisher 42 Andersen 38 DHLehmer 13 Ritschel 12 GABandAR 9 PardiTOS 8 LLnhardy 8 RosntlJA 6 DonKnuth 5 ColeStev 3 AEWestrn 3 JLGPardo 3 PDeGeest 2 CKernTOS 2 JamesFry 2 JRdrgTOS 2 MJandJKA 2 ML.Brown 2 Pinho_MF 2 RosntlJF 2 Weslwski 1 ANair_MF 1 APinhTOS 1 ATeixTOS 1 CBastTOS 1 Euclid 1 H.Dubner 1 JFNSTOeS 1 LMorelli 1 RFischer 1 TAlmFMJA 1 YPPauloR
It is funny that Euclid is in this table. I am pretty sure that he discovered more than 1 prime gap. lol

Last fiddled with by Bobby Jacobs on 2020-10-17 at 22:50

 2020-10-18, 07:28 #182 robert44444uk     Jun 2003 Oxford, UK 193210 Posts This graph just goes to show how puny our efforts are, and how many really big (40+ merit) gaps there are out there! Attached Thumbnails
2020-10-19, 07:45   #183
SethTro

"Seth"
Apr 2019

3·89 Posts

Quote:
 Originally Posted by robert44444uk This graph just goes to show how puny our efforts are, and how many really big (40+ merit) gaps there are out there!
I coded up a dynamic version of this with credit

https://primegaps.cloudygo.com/graphs.html

2020-10-19, 08:29   #184
robert44444uk

Jun 2003
Oxford, UK

22·3·7·23 Posts

Quote:
 Originally Posted by SethTro I coded up a dynamic version of this with credit https://primegaps.cloudygo.com/graphs.html
Very pretty - I would take the top of the CFC line and code as "Mersenne Prime Gap Search" (PGS) given the efforts of this group to get the search to 2^64 a couple of years back.

On a selfish note, I think any colour similar to Dana's (i.e. mine) need to be changed to accentuate. Maybe swap out some red yellows and green to allow current searchers to see their results more clearly.

Another selfish note - on another list - https://primegap-list-project.github...merit-over-10/ where I am noted as cooperating with Dana for the large gaps - in fact for all gaps over 650k I did this work with my own software and approach. The problem can be traced back to the simplicity of the Nicely site that largely listed the submitter, and not the approach.

My records come in four flavours - those less than 1500 were my finds under the PGS collaboration, those over 650k, which were mine alone, and those in between which were largely my running Dana's perl program but also contain quite a lot of records in the 1500 to 5000 range relating to CRT offsets, which was a novel approach I worked on.

2020-10-19, 08:52   #185
SethTro

"Seth"
Apr 2019

4138 Posts

Quote:
 Originally Posted by robert44444uk Very pretty - I would take the top of the CFC line and code as "Mersenne Prime Gap Search" (PGS) given the efforts of this group to get the search to 2^64 a couple of years back. On a selfish note, I think any colour similar to Dana's (i.e. mine) need to be changed to accentuate. Maybe swap out some red yellows and green to allow current searchers to see their results more clearly. Another selfish note - on another list - https://primegap-list-project.github...merit-over-10/ where I am noted as cooperating with Dana for the large gaps - in fact for all gaps over 650k I did this work with my own software and approach. The problem can be traced back to the simplicity of the Nicely site that largely listed the submitter, and not the approach. My records come in four flavours - those less than 1500 were my finds under the PGS collaboration, those over 650k, which were mine alone, and those in between which were largely my running Dana's perl program but also contain quite a lot of records in the 1500 to 5000 range relating to CRT offsets, which was a novel approach I worked on.
These are all reasonable suggestions.

1. It's moderately hard to do this. Maybe a good balance would be to add a line with text "Mersenne Prime Gap Search Group" under that area.

2. You can toggle Dana's (or anyone's) records off by clicking on their label. I certainly agree that you don't want to be near Dana's color, but it's hard to select many different colors. You can also toggle the range with ?max=20,000 and hope the order rejiggers so that you get a different color.

3. This can be tracked to the credit line, the site uses the text from the final "display" column verbatum
Quote:
 INSERT INTO credits VALUES('RobSmith','Robert W. Smith','Coordinator, PGS. Some results contributed after 2014 employ Perl code written by Dana Jacobsen.','Robert W. Smith [Dana Jacobsen]');
4. One solution would be to split your contributions to multiple names and credit Dana for only the 1500- 5000 range records?

Sidenote: The code that runs this site is at https://github.com/sethtroisi/prime-gap-record-server, It's fairly easy to run locally and I'd happily accept pull requests.

Code:
$git clone https://github.com/sethtroisi/prime-gap-record-server$ cd prime-gap-record-server
$git clone https://github.com/primegap-list-project/prime-gap-list.git$ sqlite3 prime-gap-list/gaps.db < prime-gap-list/allgaps.sql
\$ python flask_serve.py

2020-10-24, 06:56   #186
robert44444uk

Jun 2003
Oxford, UK

78C16 Posts

Quote:
 Originally Posted by robert44444uk Most wanted: No gap known: Code: ... 121366 .... 127346 .... 127544 ...
Now found

Last fiddled with by robert44444uk on 2020-10-25 at 07:43

 2020-10-24, 14:41 #187 robert44444uk     Jun 2003 Oxford, UK 36148 Posts I just found the second largest merit 30 gap ever: The list will look like this when I post it - top 20 30-merit gaps Code: 309030 31.7822 Rosnthal 214090 30.5885 RobSmith 203890 35.6402 RobSmith 189078 35.5311 M.Jansen 183860 32.0045 M.Jansen 157194 30.3829 RobSmith 157178 31.8285 Jacobsen 129326 30.2817 RobSmith 128826 31.1432 RobSmith 125078 30.4885 Jacobsen 120664 32.3245 MJPC&JKA 116676 30.3614 Toni_Key 109776 32.2502 Jacobsen 101616 30.8124 Jacobsen 99160 30.5264 Jacobsen 89028 30.7589 RobSmith 86676 31.2247 RobSmith 85686 30.093 MJPC&JKA 84766 30.711 Jacobsen 82404 31.0414 M.Jansen 78260 30.0127 RobSmith

 Similar Threads Thread Thread Starter Forum Replies Last Post gd_barnes Conjectures 'R Us 299 2021-02-19 09:30 gd_barnes No Prime Left Behind 251 2021-02-15 03:00 willmore Computer Science & Computational Number Theory 48 2010-09-19 08:30 NBtarheel_33 Hardware 17 2009-05-04 15:52 MoZ Factoring 6 2006-02-28 12:02

All times are UTC. The time now is 08:12.

Thu Apr 15 08:12:23 UTC 2021 up 7 days, 2:53, 0 users, load averages: 1.84, 1.87, 1.98