An easy offline read-only version of a fed wiki site?

[coevolving]

On the weekly federated wiki hangout, a question arose: what might make federated wiki more attractive on mobile devices (and maybe particularly on tablets)? The adoption of mobile devices is rising so quickly, we shouldn't presume that everyone will have a keyboard at hand.

Mobile devices are more read-mostly platforms rather than write-mostly platforms. A while back, when I was going to a cottage that didn't have mobile service, I figured out the way to move a Dokuwiki to a laptop was to install an HTTP server. (That's not likely to be done by many people!)

If there was a "download this site" button on a federated wiki, would people use it? A tablet could be connected to wifi while at home, and downloading an entire federated wiki as a read-only site could make reading a lot easier. There could be other benefits, such as archiving snapshots of the site. (We surfed over to archive.org, and noted that fed.wiki.org is actually crawled, although there are only one-page views because the Internet Archive doesn't support Javascript.

Is this a feature that is worth prioritizing?

A digest of the Nov. 12, 2014 federated wiki hangout is posted at http://fed.coevolving.com/view/digests-from-sfw-meetings/view/digest-2014-11-12 .

[WardCunningham]

Perhaps every client should read-ahead and cache a breadth-first traversal rooted from any page you touch. Some priority scheme could balance the available storage space agains many indicators of possible interest such as nature of touch (read or fork), distance from touch (some page rank formula), and even newer versions of any page uncovered in this traversal. The cache could be improved any time you happen to catch a bit of internet while grabbing a latte.

[WardCunningham]

There are uses for a user-initiated while site clone. This could come in two versions, one that adds a fork to every journal and one that doesn't. The latter would be most useful when moving sites between servers. For this I currently use scp or rsync.