On Sun, Sep 12, 2010 at 1:25 PM, MZMcBride z@mzmcbride.com wrote:
Rob Lanphier wrote:
I've drafted this up here:
http://mediawiki.org/wiki/Pending_Changes_enwiki_trial/Reversion_collapsing
Would this require storing the checksums in the database or would this be done dynamically on page history views? There's a related bug about implementing checksums of page text into MediaWiki. Some people aren't thrilled with the idea.[1] [1] https://bugzilla.wikimedia.org/show_bug.cgi?id=21860
We'll probably need to store the checksums. My read of the objections in bug 21860 looked like objections based on not having a clear use case (which this provides), fear that developers will start querying on the new field, and a refuted concern about possible MD5 collisions.
There's a broader question about whether page histories should be "pure" or not. The history of what happened to a page might be unsightly, but tampering with it (or the public's view of it) can be dangerous.
I don't think this is really tampering with the history; just with the presentation of the history.
It may be that, at first, the feature would need to be enabled on pages configured for Pending Changes, since that's where the need is the most acute. There's a lot of clamoring for "proper" rejection of a revision, where "proper" is "the revision doesn't show up in the revision history". If we implemented the request literally, there would be other people who would complain that we're destroying good faith edits, so we need them to show up somewhere. This would be a compromise; the revisions are still there in the db, but they aren't in everyone's face.
Rob