I currently use three of the desktop computing platforms (Windows, macOS and GNU/Linux) and one of the mobile computing platforms (Samsung-flavoured Android); I currently get paid to develop software for “the web”, an amorphous non-platform that acts in many ways like a vendor platform orthogonal to those just mentioned. This is not because I want to ally myself to any or all of those platforms but because I want to be practically independent of all of them. This is where I try to understand why.
A code of ethics
As a member of the ACM I have committed to act in accordance with the ACM Code of Ethics and Professional Conduct, or to revoke my membership. I believe that a lot of my unwillingness to buy into these platforms (and by extension, as a software professional, my unwillingness to push them on others by building upon them) can be expressed in terms of these codes.
Of course, this is expression within my interpretation of the terms. I don’t believe that Google’s data harvesting activities are consistent with the ACM Code of Ethics, but a previous ACM president was Vint Cerf, a Google employee. Clearly we disagree on our application of the Code to Google’s activity. Neither of us is, without further analysis, [in]correct; ethical decisions will be situated in a cultural and experiential context.
This, in general, is why American technology companies find it hard to expand their operations to Europe; the framework in which their actions will be viewed and judged is different.
App stores and the introduction of dependence
The iOS platform is locked down to most people, in three ways: technologically (a cryptographic system prevents all but specific native software and arbitrary JavaScript from being run); market control (only Apple can add software to the approved list; only paid members of their developer program can propose software for approval) and socially (the norms among developers, both native and web software, follow the “don’t make me think” principle in which applications are simplistic and unextensible, as discoverability and ease of use are valued over flexibility or composition).
The same argument applies to Google’s Chrome OS and Microsoft’s Windows 10 S.
This situation promotes an “Apple knows best” effect in which use of a computer is a passive consumption activity where the only applications available are those deemed fit to publish by the central actor, much like broadcast television.
In a professional community that previously gave us Mindstorms and Smalltalk in the Classroom, it is hard to accept that such a prescriptive model of computing is considered tolerable. Indeed, it seems at odds with an ethical imperative to improve public understanding of computing (section 2.7 in the ACM Code). It also seems to produce a three-tier system, in which those who have (the platform operators) are at the top, those who have means (ISVs who can buy into the approval system) are a rung down – albeit in a feudalistic vassal state that seems akin to coal miners buying their picks from the mine owners – and everyone else is below them. This would not seem consistent with an imperative to be fair (section 1.4 in the Code). Indeed I would go as far as to say that putting most of the people interacting with a software system into subject positions does not contribute to society or human well-being (section 1.1).
The web and “cloud” as protection rackets
If I use a web-based software application, it will probably offer to store all data on its developer’s servers (or more accurately on virtual machines run on the developer’s behalf by some service provider). It may not, indeed probably won’t, offer an alternative. My ongoing use of the application is predicated on my ongoing acceptance of the developer’s terms and pricing structure. The collection of individuals and organisations with whom my data is shared are also subject to change at any time, and I either demur or stop using the service.
But because their “service” also includes exclusively providing access, even to me, of what I created using their application, even accessing the things I already created is subject to their licence and my acceptance of their future changes, which can’t be known (and will probably be hidden up front, even where they are already being planned). This does not seem honest nor trustworthy (section 1.3 of the Code), nor to provide comprehensive and thorough evaluations of the impacts of the system (section 2.5), nor to credit my intellectual property in creating the things that are ransomed by their service (1.6).
