Feature: Check the version nearby to jump over vandalism

Brought to you by: flominator, powtac, siebrand

#64 Feature: Check the version nearby to jump over vandalism

Milestone: None

Status: closed-fixed

Owner: nobody

Labels: None

Priority: 5

Updated: 2017-10-31

Created: 2013-08-14

Creator: Fg68at

Private: No

Often seen, here an actual example:
http://wikipedia.ramselehof.de/wikiblame.php?article=Kondom&user_lang=de&skipversions=2&ignorefirst=3&limit=250&searchmethod=int
Search: "Ab 1930"
Search before "14:14, 5. Jun. 2008"
Search before "01:45, 3. Apr. 2006"
Insert found between 23:05, 13. Dez. 2005 und 23:05, 13. Dez. 2005
But: 23:05, 13. Dez. 2005 was a vandalism: http://de.wikipedia.org/w/index.php?title=Kondom&oldid=11611821
The real insert version is: Version vom 2005-09-16, 09:31:23 Uhr and there is also the source.
http://de.wikipedia.org/w/index.php?title=Kondom&diff=next&oldid=9190480

Simplest help: A "Search from here" also at the result page
http://wikipedia.ramselehof.de/wikiblame.php?project=wikipedia&article=Kondom&needle=Ab+1930&lang=de&limit=250&offjahr=2006&offmon=4&offtag=3&offhour=23&offmin=55&searchmethod=int&order=desc&force_wikitags=off&user_lang=de&ignorefirst=&binary_search_inverse=false

More complex help: A check of the versions 1 until propably 3 (some make 3 vandalism edits in a row, more are very rare, not check version 2 in case of edit war) before or after the initial found version to ckeck if there was a vandalism or edit-war and the real initial version is before.

Discussion

Flominator - 2013-09-15

status: open --> pending

Group: -->
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Flominator - 2013-09-15

Detecting vandalism is quite complex imo.

When I click http://wikipedia.ramselehof.de/wikiblame.php?project=wikipedia&article=Kondom&needle=Ab+1930&lang=de&limit=250&offjahr=2006&offmon=4&offtag=3&offhour=23&offmin=55&searchmethod=int&order=desc&force_wikitags=off&user_lang=de&ignorefirst=&binary_search_inverse=false it doesn't find a vandalism and shows a lot of "search from here" links. What am I doing wrong?

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Fg68at - 2013-09-15

Oh, now the Dataset for the binary search is another and it goes in the second search to the correct version.

There are many "search from here", but only if the string is in both versions found (green OO) or not found (red XX).

If an insertion is found (XO)(now on the second resultpage), there is no "serch from here". But this insertion can be a revert of an vandalism and the relevant insertion in a version before.
I would like to have an "search from here", even if a difference is found.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Flominator - 2017-10-31

status: pending --> closed-fixed
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Flominator - 2017-10-31

I've been able to reproduce the problem by picking an earlier day, I think.

In case it gets stuck, WikiBlame now throws away revisions where it is sure, that the term was already inserted: https://github.com/FlominatorTM/wikiblame/commit/e4aad37993e4b8f720366d4af19f7ab1ef5ee69b#diff-d7aa6e2e161a9837cc49fb8d725fd5de

Also "search from here"in case the hit is vandalism works meanwhile:
http://wikipedia.ramselehof.de/wikiblame.php?lang=en&project=wikipedia&article=Human+trafficking&needle=one+place+to+another&skipversions=0&ignorefirst=0&limit=500&offtag=22&offmon=10&offjahr=2017&searchmethod=int&order=desc

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Log in to post a comment.