…when will the rediscovery of plaintext and the applied Unix philosophy make waves in the “tech industry”, and how will it be repackaged to sell back to us?
It’s already happened. We’re on like, Wave 3.
One word: “Services”.
To apply the author’s argument: Text should be queryable, too; specifically, it should be readable. But readability is not something that we can measure in the absolute or the objective. You say that text should be “plain”, but even plain old text files are actually built out of arbitrarily chosen alphabets. Do we use Unicode or ASCII for text?
If we consider the human-computer-interface point-of-view, then text should also be writeable, and it is here that we see the programs-as-prisons metaphor on full display: We can only write using the alphabets available for text encoding!
Every ASCII file is a UTF-8 file. Just use UTF-8 :)
I’m appalled by the lack of historical knowledge. Sure, everyone has heard of Alan Kay and Smalltalk, but there have been other attempts at these ideas over the years. Microsoft was pushing object linking and embedding (OLE) for several years but it went nowhere. Did anyone think to maybe look into that and try to find out why it failed? Was it too early? Was it too confusing a concept? Did it morph into COM? (And I say this as someone who almost never used Microsoft Windows)
Yes, this article needs a lot more analysis (or links) on what has been tried in the past, what’s failed, and why. The problem is real but the article is vague.
Some Java GUI IDEs have something like this: you can create a component, add various events handlers, then copy paste the whole component into another menu and it will recreate all the handlers and layout. Excel will also let you copy paste rows/columns/grids within a spreadsheet. It walould be really cool to make that a standardized format so you can use it across applications.
Open Source is obviously great but the next step would be something like “Open Applications”, as this article presents.
Since setting these features up requires additional energy and work, it would be similar to how companies open source to get “bonus points”.
The standardization of aspects of technology is both extremely powerful and extremely dangerous. And I think that is what would be needed here. Standardization, i.e, on these Image objects, and interfaces for reading and writing them.