Luckily we can find out who was thanked too, it takes the same time to run. Data below and link for updated report (which is on wikitech due to WMF server outage for main projects at the time of writing). * https://wikitech.wikimedia.org/w/index.php?title=User:F%C3%A6/sandbox&o… Extending to other wikis is not a problem, we just use the relevant wiki-name of the tables, or call on it when we go into MySQL (and then the script does not need to be edited). It should be possible to do a nice report that does this for all active wikis; once a month is enough. I have published the SQL below for those technically minded. If someone wants to pick this up, that's cool. If left to me it may take a couple of months to polish off due to my backlog. My other maintained reports are mostly listed at https://commons.wikimedia.org/wiki/User:F%C3%A6/reports Top 10 *most thanked* users in Jan 2015: English Wikipedia: 351, Ryulong 212, Niceguyedc 151, Ssven2 119, Materialscientist 111, HJ Mitchell 100, Ser Amantio di Nicolao 89, GoingBatty 89, Drmies 87, John of Reading 79, Rocketrod1960 Wikimedia Commons: 47, Steinsplitter 44, INeverCry 44, Thibaut120094 44, 1989 37, Yann 32, Medium69 27, Be..anyone 26, Brackenheim 22, ArionEstar 21, Marcus_Cyron SQL: // Top 10 thanks givers in a month SELECT log_user_text, count(log_action) FROM logging_userindex WHERE log_action="thank" AND LEFT(log_timestamp,6)="201501" GROUP BY log_user_text ORDER BY count(log_action) DESC LIMIT 10; // Top 10 thanked people in a month SELECT log_title, count(log_action) FROM logging_userindex WHERE log_action="thank" AND LEFT(log_timestamp,6)="201501" GROUP BY log_title ORDER BY count(log_action) DESC LIMIT 10; P.S. I would be cautious about interpretation. I am tempted to automate some thanking, just for my name to float in the top 10 and surprise everyone. ;-) On 5 February 2015 at 16:18, Richard Farmbrough <richard(a)farmbrough.co.uk> wrote: Fae

Thanks Fæ, those reports are very interesting from my perspective. However, even though the number of thanks received may be more reliable than the usual edit count, none of them could ever measure the invaluable impact of our contributors :-) Il 23/02/2015 16:57, Fæ ha scritto:

Excellent!! Great initiative, congratulations! Thank you for answering my request for inclusion of pt.wikipedia! Best regards Rodrigo Padula 2015-02-23 12:57 GMT-03:00, Fæ <faewik(a)gmail.com>om>:

Fæ wrote: Hi. Hmm, I'm not sure how you're measuring largest, but I imagine on most wikis there are more rows in the revision table than there are in the logging table. For example, on the English Wikipedia, there are approximately 598,859,006 rows in the revision table and 62,731,285 rows in the logging table. I suspect on most wikis, revision, text, and maybe archive would typically be larger than logging, except in weird cases such as loginwiki[_p]. And then of course there are the *links tables. But it depends on whether you're comparing size on disk or number of rows. You probably want to use logging_userindex instead of logging. The former is typically significantly faster due to the way we use table views. I have a bit of experience with database reports. Off-hand I'd say it should be possible to query all of this information in under an hour. With the index on logging[_userindex].log_action, even a large table such as logging shouldn't be too awful to query for this information. If you have queries that are taking a very long time, we should probably investigate. MZMcBride

Hello Fae, Are you generating json somewhere or thiking to create an API for that information? I have some ideas that I started to develop last year, I will ping you in private. Best regards and thanks by helping us with the important information. Rodrigo Padula 2015-02-26 14:01 GMT-03:00 Fæ <faewik(a)gmail.com>om>: