The GIMPS Status page is going (gone?) to V5...

[Prime95]

Originally Posted by lycorn

If I understand it correctly, this report is now much more often updated (actually, in near real time) than lucas_v.txt was, isn´t it?

True - once the the v4-to-v5 bridge is in place.

One nice feature would be to filter records using the name of the tester as one of the criteria.

There is such a feature if you've logged in with your v5 id. However, it does not return any results returned while you were operating under v4. Needs work.

[Prime95]

Originally Posted by Graff

There seems to be a lot of factors that are missing from the results
returned by http://v5www.mersenne.org/report_factors/.

I was checking that the results that I'd reported by e-mail to GW
while doing lone Mersenne hunting were included. When I had reported
a new first factor and an additional factor for an exponent (e.g.,
M 6147139), both factors are returned. When I had reported
additional factors for an exponent with a previously-known small
factor (e.g., M 8002297), the additional factors are not listed.

The old database only recorded the smallest factor. The new SQL database records all known factors. Most additional factors reported prior to a month ago are likely lost - or at least too difficult to retrieve from my logfiles.

If you find any missing factors, please resend them!

As to M8002297, I don't think I got that email. Please resend.

[Graff]

Originally Posted by Prime95

The old database only recorded the smallest factor. The new SQL database records all known factors. Most additional factors reported prior to a month ago are likely lost - or at least too difficult to retrieve from my logfiles.

If you find any missing factors, please resend them!

As to M8002297, I don't think I got that email. Please resend.

I've just sent you two files via e-mail . The first contains
3641 first factors. The second contains 22052 additional
factors for 18376 exponents. All were reported previously.

A series of checks on factoring limits for exponents
without any known factors seems to show that my
upper limits have been included. In the new V5 is there
a way to get the upper limit for trial-factoring for
exponents with known factors?

Graff

[Prime95]

Originally Posted by Graff

In the new V5 is there a way to get the upper limit for trial-factoring for exponents with known factors?

Nope.

[Batalov]

In order to leave no visitor/potential contributor behind, I'd add the http://mersenne.org/freesoft.htm and maybe other links to similar pages to the left sidebar of the http://v5www.mersenne.org/ (right under "Getting started" somewhere).

[cheesehead]

Originally Posted by Prime95

The new SQL database records all known factors. Most additional factors reported prior to a month ago are likely lost - or at least too difficult to retrieve from my logfiles.

If you find any missing factors, please resend them!

On March 12, I made a manual report of a new factor using http://mersenne.org/ips/manualtests.html. As of when I compose this, that factor hasn't yet shown up on the report for that exponent range using http://v5www.mersenne.org/report_factors/. I'm not complaining, but just wondering about how long you expect it would take to show up in the V5 report?

Or is there not yet a connection of current V4 manual reports to the V5 database?

[Prime95]

Originally Posted by cheesehead

Or is there not yet a connection of current V4 manual reports to the V5 database?

Currently their is no automatic connection from v4 to v5. This requires the "bridge" software Scott is writing/debugging.

The v5 pages are updated with the latest v4 results when I manually download the v4 results and run them through my program that sends them to the v5 server via a backdoor. This is very similar to the roughly bi-weekly process I used to update my text databases and the old status page.

[dswanson]

Any chance of adding an option to the new query page to dump flat text files in the same format as the old lucas_v.txt, hrf3.txt, pminus1.txt, etc? I have programs written that ingest those files, and they choke on the new output format. At a minimum, something tab-delimited would be helpful.

[garo]

I love the new date feature! It means I won't have to download the entire set of results ever again. Some of the requests I was going to make have been listed above but I'll repeat them nevertheless and add some of my own. In decreasing order of preference:
1. As dswanson said can we please please have a text-only format. Preferably, the comma separated style from v4 so our parsing/analysis programs can work unchanged.

2. The html could do with a bit of cleaning up. Currently it looks something like this:

tr><td><font face="Tahoma" size="2">16266829</font></td><td><font face="Tahoma" size="2">Northwestern College</font></td><td><font face="Tahoma" size="2">C378DEB61</font></td><td><font face="Tahoma" size="2">74E59C7D53C557__</font></td></tr><tr><td><font face="Tahoma" size="2">16320053</font></td><td><font face="Tahoma" size="2">Aaron Zod</font></td><td><font face="Tahoma" size="2">lcat-2</font></td><td><font face="Tahoma" size="2">200BA3C7B04DA3__</font></td></tr><tr><td><font face="Tahoma" size="2">16523869</font></td><td><font face="Tahoma" size="2">Stephan Grupp</font></td><td><font face="Tahoma" size="2">fish</font></td><td><font face="Tahoma" size="2">90B1FD0F56DA5F__</font></td></tr><tr><td><font

it would be nice if you could put a newline after each </td> and two new lines after each </tr>
Also, you can get rid of the font specification for every single cell by putting one at the start of the table. It will save a lot of bandwidth.

3. While only results from the logged in user are nice, can we have a simple text search as well?

4. Have you considered the security implications of an open form that places such a massive load on the database? If you or Scott haven't already done so, it may be an idea to limit request rates from the same IP etc.

[Prime95]

Originally Posted by garo

1. As dswanson said can we please please have a text-only format. Preferably, the comma separated style from v4 so our parsing/analysis programs can work unchanged.

2. The html could do with a bit of cleaning up. Currently it looks something like this:

it would be nice if you could put a newline after each </td> and two new lines after each </tr>
Also, you can get rid of the font specification for every single cell by putting one at the start of the table. It will save a lot of bandwidth.

3. While only results from the logged in user are nice, can we have a simple text search as well?

4. Have you considered the security implications of an open form that places such a massive load on the database? If you or Scott haven't already done so, it may be an idea to limit request rates from the same IP etc.

1. Working on that now for the LL report
2. I'll output newlines if I can figure out how to do that in PHP. You'd think it would be easy - I must be missing something.
I tried moving the font specification and it did not work. Any clues as to why?
3. I'll see if I can add a userid/username search. You didn't want pattern matching, did you?
4. We'll drive off that bridge when we come to it. The queries are limited to 10000 rows. Scott is the expert on these kinds of server issues.

[Prime95]

Got newlines working. Test page for you to try:

http://v5www.mersenne.org/testbed/

The csv text portion is between <pre> and </pre>. Comments?