Programming fits both of those categories: consider the number of programming languages in vogue at the moment. These languages however are composed of the same components which comprise the machine.
Programming languages are software, no differently than iMessage or Windows Live Mail. It's just another way to control the machine.
NewChat(tm)'s best bet is to have the new interface mimic and borrow the metaphors of the older interface.
In a sense, what makes an interface "good" is that it uses what is already known.
As for the original question regarding the distribution of user's "computer skills"... if you can figure out how designers think about how you should use systems, then it's much less about having "computer skills" than having the capacity to figure out the psychology of the application designer - if you can type and use a mouse and can either Google or read a book describing the paradigms of an application or eco-system and figure out what they mean, you've got all the computer skills you need.
As programmers, I think we tend to forget what exactly it is we do... we don't just program, we figure shit out at the most technical levels. Other people are quite adequate at this too - we have lawyers, doctors, surgeons, mechanics and any number of other professions that solve way more complex problems than we do - we're not islands of all the world's intelligence. If people appear to lack "computer skills" it's because the paradigms we use to define our user interfaces and solve the problems we are attempting to solve are shit - and I say this as a pretty well seasoned computer programmer who spends most of my life trying to figure out how to get other people's shit user interfaces to get my own shit done.
So... that's really what I've got to say about that. People will figure out what they have time and inclination to figure out. If they don't have the time or inclination to figure it out, then it's your problem: Your software isn't solving big enough pain points for the learning curve involved. This means that either this user is not your target audience, or you don't understand your target audience well enough.
Either way, it's not on the end user, it's on you as the software designer.
Lawsuit.
wouldn't that be knowing how to program?
Regardless of their future career, people should probably be taught how to use a word processor (use styles, hierarchically outline your document, basic sensible typography) and spreadsheet software at school, along with basic internet skills (particularly with regards to finding reliable information and understanding the pros and cons of social media).
None of these skills require any programming knowledge.
> wouldn't that be knowing how to program?
Nearly all programming languages or instruction sets for assembly languages are also coincidental interfaces (this time for programming), often (though not always) at least originally designed by some "megacorp":
- Assembly languages: typically the respective processor vendor
- C: Bell Labs (AT&T)
- C#: Microsoft
- Go: Google
- Java: Sun (now Oracle)
- Javascript: Netscape
- Objective-C: NeXT (now Apple)
- R: A reimplementation of S (originally by Bell Labs; later commercially offered by TIBCO Software (another megacorp) under the name "S-PLUS").
- Swift: Apple
So what I would consider as more general computer skills is rather the mathematical or the electrical engineering side of computing.
Nearly all the mainstream, massively successful ones, yes. But the vast bulk of languages are not, and there is little evidence to be provided that there is some sort of massive advantage to be gained by "normal users" for any of them. There are better ones than those for some purposes, yes, but after many thousands of languages it's not like there's one that is clearly better for non-programmers, and this after at least several dozen and probably several hundred attempts explicitly aimed at that purpose.
Asking for people to understand the math is an even larger ask than asking them to understand the programming, which I base on the fact you can still find a very significant percentage of professional programmers, possibly even the majority, who have disdain for the mathematical elements of computing.
It is also not an unfriendly reading when my very first words (which I have not edited) acknowledge the alternative readings.
Yes, there are programming languages aimed at non-programmers. Many of these are also "commercial" languages, or at least were commercial in origin (like C). Smalltalk (Xerox), Hypertalk (Apple), the efforts towards 5GLs in the 1980s, Visual Basic (Microsoft), etc. We agree that these are relevant, I'm just not sure what point you're trying to make with that relevance, and I'd love it if you could elaborate.
I'll make an attempt to explain why jerf's response looks out of place.
Basically, wolfgke and jerf looked at the 2 groupings of programming languages (mainstream megacorps vs unknown) as evidence for 2 different goals.
wolfgke: generalization path -- dominance of megacorps languages means you must keep going one level lower in hierarchy of computer science concepts to learn universal concepts instead of proprietary syntax.
jerf: pedagogy ease of use -- megacorps languages are not provably any harder to learn than specialized toy languages
How did those two end up talking about 2 different things?!?
If you look at the comment chain from posters threatofrain-->bryanrasmussen-->wolfgke ... they started a dialogue that keeps diving lower and lower in underlying principles. It's a variation of the XKCD comic about "purity".[1]
jerf's response doesn't continue that purity dissection. Instead, his emphasis on pedagogy seems to point back to the original article by Jakob Nielsen which prompted this thread. That article says that elite users (like HN readers) can use complicated computer software and we forget that most others can't. The communication breakdown was assuming that wolfgke listed the megacorps language as a (ease-of-use) response to J Nielsen instead of a specific (purity) reply to bryanrasmussen.
But then again, I might have misunderstood everybody and I have no idea what people were trying to say.
Correct, with the additional (IMHO important) fact that threatofrain criticized "coincidental interfaces made by a few mega corporations" and looked for more general skills, bryanrasmussen meant that "knowing how to program" is such a skill, but I analyzed that most popular programming languages are also just "coincidental interfaces made by a few mega corporations" (this time for programming instead of the general user), so that we have nothing won concerning the original problem of "coincidental interfaces made by a few mega corporations" - we are just some layers deeper. So I suggested that if you look for more general skills, you probably have to look even deeper into the mathematical or the electrical engineering side of computing.
Are there computer skills more general than the ability to use iMessage or Windows Live Mail? Because if using iMessage means you now know how to chat, then what happens when NewChat(tm) comes out with a new brand experience?
What happens when Microsoft goes Metro or Windows 10?
As a society, should we be paying for user training for specific familiarity with proprietary interfaces? Shouldn't Microsoft or Apple be paying for this stuff? This is why I'm generally very suspicious of what people mean by "computer skills" in schools. They generally mean user training for Apple, Microsoft, or Google.