by Konstantin Weiss on October 9, 2018
Here is a thought, that needs validation, so see it as food for thought: When it comes to building an information system, I see it consisting of four layers.
From top to bottom they are:
An #appliance is some function or logic that applies to data with a specific goal. It can be reading, writing, updating, modifying, combining. In short, any kind of data/content display or manipulation. A group of appliances within the same topic or area is what I call a #service. E.g. at #betterdoc, a form asking for a voucher code is an #appliance. This appliance is one of many in the voucher #service, which provides means to administer vouchers, validate them or provide reports about them for further analysis.
Note: I use the term “service” here not to be confused with other technical terms like “micro service”.
A #place a composition of appliances (from different services) that allows an actor to get specific tasks done. E.g. at #betterdoc, the landing page is a place for a patient to register, where the voucher form is just one #appliance next to others like personal data inputs (applications of other services). A careful selection of services and their arrangement is what makes a place useful. (See more on #placemaking)
An #actor is a human or machine who/that is granted certain rights and has a certain role. He/she/it has motivations/goals and usually intends to get tasks done with the help of places. E.g. at #betterdoc a patient wants to use betterdoc and therefore uses the registration place for starting an inquiry.
The four layers are built on top of each other. Every upper layer is depenent on the underlying layer, whereas every underlying layer is not dependent on the upper one. E.g. data is not dependent from the appliance layer, but the appliance layer needs the data layer. Also, every underlying layer has no knowledge of the layers above. Of course, only all four layers together make an information system useful.
Question: Do machines as actors also use/have #places to get their tasks done? And if so, how do they “look” like? Are they also a collection and composition of applications?
Back to Index