- From: Scott Jenson <scott@jenson.org>
- Date: Tue, 16 Apr 2013 08:28:29 -0700
- To: Dominique Hazael-Massieux <dom@w3.org>
- Cc: "public-closingthegap@w3.org" <public-closingthegap@w3.org>
- Message-ID: <CACLVYsHQ1dG0ignkjuTqdgY5L=5bZwF+qCnQzqdAkHV=E_M-Tw@mail.gmail.com>
Dom, You've nailed the opportunity, in fact, I strongly feel this is the 'super power' of the mobile web, something that native apps can *never* do. The ability to walk into any location and 'pull up an app' is not only powerful, it provides a tremendous technology base for all sorts of products/companies. I hope we can all agree that magically opening a specific web app in for a specific location/situation/product is useful, what gets sticky is *how* as all of today's technologies are a bit of a compromise. Wifi usually requires you to join the network first. As most wifi networks are protected these days that's a huge hurdle (of course protected wifi *does* give you a modest amount of security to protect your devices from others) Bluetooth is a bit of a mixed bag, it gives you 'drive by' discovery. Bluetooth 4, btw, is much better at this than previous versions. If you had a 'growl style' app on your phone to look for BT4 services that were broadcasting they had a web page, that could work. The only downside is that BT4 wouldn't give these devices access to the internet, they'd need another radio/ethernet. Not a killer, but it makes it harder to make cheap devices that you can interact with. WifiDirect is promising as it solves all of these problems, BT4-like discovery plus full access to the internet through normal wifi. The obvious problem here is that WifiDirect isn't exactly setting the world on fire. I'd be interested in anyone else's thoughts on way's to kick start this 'just in time' discovery. I've come to the conclusion that no one technology will save us here. That 'growl app' I mentioned earlier likely will need to cover a range of wireless technologies (my initial guess is BT4 and WifiDirect but equally likely is a protocol on a router that is advertising nearby services) This is a very confusing space but if we can crack this "lower-level discovery" issue, we can unlock a whole new economy... Scott On Tue, Apr 16, 2013 at 12:38 AM, Dominique Hazael-Massieux <dom@w3.org>wrote: > We discussed in an earlier thread the stakes behind making Web apps more > searchable: > http://lists.w3.org/Archives/Public/public-closingthegap/2013Mar/0007.html > > A related but different topic is how to make Web apps easy to discover > and access from within the user context. > > If a retail store offers its customers a Web app (taking advantage of > the fact they don't need to be installed), how can it make it easy for > them to get to that Web app? > > The existing mechanisms have all drawbacks: > * URLs are hard and slow to type on mobile > * QR codes often require dedicated software, and are probably too > cumbersome to make the pain of using the Web app negligible > * NFC tags require dedicated hardware and software, and still require > the potential user to move close to the tag > > To enable what Scott calls "just-in-time interactions" [1], I think > making the cost of discovering and accessing Web apps as low as possible > would help service providers benefit from this fairly unique > characteristic of Web apps. > > A couple of ideas on what could help make context-relevant Web apps > discovery seamless: > * network-based detection: in our retail store example, let's imagine > the user uses the free WIFI network provided by the store; that network > advertises the Web apps (e.g. via Zerconf/bonjour), and the browser > offers the user a special view of local services; > [ https://dvcs.w3.org/hg/dap/raw-file/tip/discovery-api/Overview.html is > exploring how to expose network services to Web apps more generally ] > > * location-based detection: if Web apps are made easier to search, and > if that search is able to detect and rate the specific value of a given > Web app for a given location, browsers could expose to the user what Web > apps are particularly relevant to her location; one of the usual > challenges is obviously to avoid spammy results > > I'm interested to hear thoughts on the overall importance of this topic, > and ideas on how to make progress on it. > > Thanks! > > Dom > > 1. http://designmind.frogdesign.com/blog/mobile-apps-must-die.html > > > >
Received on Tuesday, 16 April 2013 15:28:59 UTC