Talk:Small and large wikis/Statistics

From Meta, a Wikimedia project coordination wiki

Stats and sorting[edit]

The stats for these sites sort by alphanumeric of the leading number, not as complete numbers. Is there a/the means to either populate with leading zeros or to get them to sort as complete numbers? If not, then maybe we should just make certain columns unsortable rather than as an meaningless feature. billinghurst sDrewth 03:41, 10 February 2012 (UTC)Reply[reply]

It seems to sort correctly now, no? Which column are you talking about? πr2 (tc) 05:20, 8 December 2012 (UTC)Reply[reply]

Dutch/Swedish Wikivoyages[edit]

Can anyone explain why nl.wikivoyage and sv.wikivoyage are listed as having 18446744073709551579 and 18446744073709551615 articles, respectively? According to nl:voy:Special:Statistics there are fewer than 4000 content pages, and fewer than 10000 pages total. "18,446,744,073,709,551,615" is the total number of grains in the wheat and chessboard problem (i.e., 2^0 + 2^1 + 2^2 + 2^3 + ... + 2^63). This is very odd. Thanks, πr2 (tc) 22:11, 3 December 2012 (UTC)Reply[reply]

Also, the other Wikivoyage versions have much lower numbers than the actual amount of articles (it.wikivoyage has 2,400 page according to voy:it:special:statistics, but the table says it has 3). sumone10154(talk) 02:27, 4 December 2012 (UTC)Reply[reply]
I'll ask the ts admins if that an replication failure or something wrong at wmf. Merlissimo (talk) 20:35, 31 January 2013 (UTC)Reply[reply]
mysql -hsql-s3 -e "select * from site_stats;" nlwikivoyage_p
| ss_row_id | ss_total_views | ss_total_edits | ss_good_articles     | ss_total_pages | ss_users | ss_active_users | ss_images |
|         1 |              0 |          15901 | 18446744073709551596 |            635 |     3149 |             168 |      -291 |
Looks like the integration of the Wikivoyages caused some problems. I copied over the values from the master for nlwikivoyage, the other Wikivoyages showed better values. The values of nlwikivoyage may be off a little bit now, but it is surely better than before. In the end only a complete re-import of the s3-cluster can really solve this. --DaB. (talk) 12:37, 1 February 2013 (UTC)Reply[reply]