See also
http://bugzilla.wikimedia.org/buglist.cgi?bug_id=471,472,943
- 472, for instance, discusses turning individual page histories into
feeds, for a more fine-grained kind of "subscription" which wouldn't
suffer the privacy problem.
Ta for the pointers :)
I don't think any of the bug reports quite get there, but "individual
page histories into feeds" is exactly what I was thinking about [1]
To be all fashionable about it, it's the "long tail" thing. There are
a lot of people who, between them, are interested in a lot of
articles. But on average there is only going to be one or two people
passionate about one given article.
At the moment, we are assuming either
- a (relatively) few people monitor lots of pages
- lots of people are going to login and check their (small) watchlists
fairly regularly
Imagine I am interested in the "random weird topic" page. I'd like to
be notified of updates to this page, but I won't be checking in daily
just in case.
Email alerts would be perfect, and I note that it *could* be a
wonderfully seperated (ad sponsered) service (low traffic for the
user, relevent ads, useful service, entirely independent)
Ok, why the WP-tech list?
[1]
This only happens if it is easy (and or massively popular, which it
isn't right now)
It could be done in a couple of ways.
WP offering RSS feeds of page histories is one. It's not optimally
efficient use of server resources or bandwidth. WP sending out emails
to users itself is another, but ... core business ... focus ...
Anyone could, given the interest (ad revenue), set up a site/service
that would easily let me "bookmark" a wikipedia page, and deliver this
service. It still presents problems about such an organisation
polling a bunch of pages every day or so. Yes, you could block them,
but why block a service that is mutually usefull all round?
Someone like Bloglines could (with arrangement) download the DB
nightly and then email out daily updates. More efficient all round.
Bloglines (or similar) is a commercial company, so I'm sure they might
help with the logistics of a daily DB dump. This then benefits
everyone in the community.
I'm in no way associated with bloglines, other than as a end user of
the service. I'd also be quite happy with any other solution.
There are a bunch of articles I am interested in, but I won't be
logging in to watchlist them. That might mean I am lazy, and so be
it. But some form of RSS feed of page histories would be a really
neat feature to have.