On Thu, 24 Feb 2005 10:02:07 -0800, Kesava Mallela <kesava(a)gmail.com> wrote:
1. Wikipedia shows only the last edit of an
article in the watch list.
This leaves the writer no clue of the edits that took place from
his/her base lined version. (Base lined version will generally be the
user's last edit on the article.). Some kind of visualization
technique on all the edits only can give the writer a sense of the
course taken by the article.
I believe work related to the "e-notif" patch has implemented a link
which shows the difference between the current version and the version
that you last viewed, which goes some way towards this. That doesn't,
of course, give you the list of *summaries* for those edits, but you
can get that out of the page history easily enough anyway, I guess.
As for visualising a whole sequence of edits, that's certainly a
laudable aim, but in practice it's not so easy. For the moment, you
can page through the diffs (with the new "next" and "previous"
links).
Some people have suggested a "blame" feature (like CVS repositories
provide) which shows in which edit each part of an article was
introduced; but it has been pointed out that unlike source code,
encyclopedia articles [or other content likely to be contained in a
MediaWiki] doesn't divide neatly into lines, so the resulting display
would be that much coarser, and thus less useful - I might correct one
typo, and the whole paragraph would end up annotated as my edit. And
then there was IBM's "history flow" experiment
[
http://www.research.ibm.com/history/index.htm] which is certainly
fascinating, but whether or not it can be put to generic and practical
use I'm not sure...
Meanwhile, some overall container which could provide standard
features and interface to Special:Recentchanges,
Special:Recentchangeslinked and Special:Watchlist would certainly be
nice - for instance, the global recent changes is filterable and has
collapsible groupings for multiple edits to the same article.
Obviously, some things are harder in the database than others, but it
would certainly be useful for all three pages to have this and more...
--
Rowan Collins BSc
[IMSoP]