The Pew Research Center asked a set of professors, businessmen, and readers of "technology-oriented listservs" to imagine the Internet of 2025. (Source in PDF and HTML.)
Some respondents speculated that there would be amplification of known trends: "ambient" networks (sensors, cameras, phones/tracking devices) that are increasingly integrated into work and social life, ongoing disruption of traditional "content" industries, and the continued growth of analytics/surveillance ("tagging, databasing, and intelligent analytical mapping of the physical and social realms.") Of course, networks "accurately predict[ing] our interests and weaknesses" implies the loss of personal privacy, first to governments and corporations, but eventually to any interested party or social engineer.
Others predicted decentralization and fragmentation self-forming mesh networks, darknets, and proliferating incompatible national/corporate algorithms. Your freedom would be circumscribed by the ideology of your network's owner.
I put the question to you, O People of Soylent. What futures do you foresee? What trends or pathologies does the Pew report minimize or neglect? How can or should *we* influence the Internet's direction in the next decade?
This strikes me as a very small obstacle. On my homescreen I have several bookmarks, right next to app-shortcuts. Some even have the icon of the website instead of the icon of the web-browser (it depends on which browser I used to make the shortcut).
This process of sending a bookmark to the homescreen (on android at least), is relatively easy (easier than installing an app) but not particularly obvious, at the moment. It would be fairly easy to make it more obvious, though.
I think both performance and the UI are holding back the mobile web to some extent, for now. But there are software and hardware improvements which continue to occur, and which will almost certainly persist for the foreseeable future. For example hardware getting faster for the same price, js engines getting more efficient, webgl and the canvas element in general becoming more popular and well-supported.
Ten years is plenty of time for these things to combine to make the mobile web fast and smooth and pretty enough for almost any application.