What I miss here is the perspective of the categories of end-users vis-a-vis required types of input.

If we assume that all consumers of agriculture related information, raw data or data sets use a handheld device (GSM, smart phone, tablet, netbook  or other device) or use a notebook or other computer to request, access or send data, we must have a grip on generic ICT system variables that define the user and his enabling ICT technologies.

The formatting of data sets, raw data and the way these are stored, requested and retrieved from repositories, and the applications driving the exchange, interchange and transmission of data must be shielded as much as possible from the end-user.

This inevitably leads to requirements of semantics, aggregators and logical data definition, storage, retrieval, and the applications that aggregate user requested content to combine mature semantic web technologies and open access, open repository standards for the widest range of user generated digital content possible.

Visualization of end user categories in their untilization environments and their enabling technologies would give us a baseline for defining what types of information need to be shared.

And in this context it is important to define the public domain (open access) versus private domains (subscriber access) defined by online publishers of content which already have widespread use or acceptance, whether they be online publishers of journals, social networking sites, news repositories, data sets, or raw data from remote sensors and input devices.