3 ms·

It’s funny reading this, I can tell he’s struggled to solve a problem I used to explain on the first day to my Intro to Computers students: a lot of this techni

It’s funny reading this, I can tell he’s struggled to solve a problem I used to explain on the first day to my Intro to Computers students: a lot of this technical information is interrelated in a web, rather than one concept leading to the next. Yet we are forced to learn it linearly. It took a lot of effort and experience teaching to resolve that conflict.

dboreham 4y ago
Same thing is true for Mathematics.
- qsort 4y ago
  Very true. One of the most effective approaches for learning math is reviewing the same material multiple times, at different levels of depth. This is what is usually done for math competition training and it's night and day compared to the usual "multiplication tables => polynomials => precalc => calc..." A similar approach to teaching CS isn't too far fetched.
  - mixmastamyk 4y ago
    Yes, aka "progressive encoding." Like a JPEG file that gets clearer and clearer as each chunk is transferred. Start with a introduction and summary of each related discipline. Go back to the beginning and build another layer on top, with as many iterations are necessary. Has much in common with the Spiral methodology of software development: https://en.wikipedia.org/wiki/Spiral_model https://en.wikipedia.org/wiki/Spiral_model
allturtles 4y ago
I think you can generalize this problem to just about any subject, or any kind of story telling. e.g. Herodotus' Histories has these digressions (I've sometimes seen them called 'pendants', which I really like), which can also be nested. i.e. time is not mapped linearly to the text, the text jumps around in time (and space) to provide context to the main narrative (and sometimes to provide context to the context). Digressions also appear in Homer. So you can think of the story elements as a kind of graph that have to be flattened into a linear structure, and to perform this flattening well is the art of good story-telling.
- 2-718-281-828 4y ago
  interesting perspective to consider writing as graph traversal.
pessimizer 4y ago
I think I accidentally got a good order because of the ad hoc learning I did as a child. I first encountered Commodore/Apple II/IBM through text-mode games/Wordperfect/typing programs in from magazines, then > learned C which gave me an idea of how they worked > learned microcontrollers then understood how the keyboard and monitor worked and how to drive things like that > then logic, electronics and transistors to figure out how microcontrollers worked (and how to make sense of digital voltages) > finally the D-latch. I do have to admit I had a multiuser OS sized hole for a while. The computers in my head were Harvard architecture RISC, generally programmed in assembly, and were driving pin voltages. But it was close enough. So an accidental process of being introduced to things as black boxes, then opening the black boxes. Still linear, but backwards. Needless to say, I loved Code when it came out and gave away at least four copies. I would tell people who thought I knew a lot about computers "this book contains pretty much everything I know about computers." I actually think computing lends itself well to that because it is built in layers. You can talk about how a layer behaves before talking about how it was made to behave that way. The problem I think people run into is that some come to think lower levels of abstraction are a waste of time from a combination of pragmatism ("I don't care how it works") and sour-grapesism ("I don't know how it works so it doesn't matter.")

dboreham 4y ago

Same thing is true for Mathematics.

qsort 4y ago
Very true. One of the most effective approaches for learning math is reviewing the same material multiple times, at different levels of depth. This is what is usually done for math competition training and it's night and day compared to the usual "multiplication tables => polynomials => precalc => calc..." A similar approach to teaching CS isn't too far fetched.
- mixmastamyk 4y ago
  Yes, aka "progressive encoding." Like a JPEG file that gets clearer and clearer as each chunk is transferred. Start with a introduction and summary of each related discipline. Go back to the beginning and build another layer on top, with as many iterations are necessary. Has much in common with the Spiral methodology of software development: https://en.wikipedia.org/wiki/Spiral_model https://en.wikipedia.org/wiki/Spiral_model

allturtles 4y ago

I think you can generalize this problem to just about any subject, or any kind of story telling. e.g. Herodotus' Histories has these digressions (I've sometimes seen them called 'pendants', which I really like), which can also be nested. i.e. time is not mapped linearly to the text, the text jumps around in time (and space) to provide context to the main narrative (and sometimes to provide context to the context). Digressions also appear in Homer. So you can think of the story elements as a kind of graph that have to be flattened into a linear structure, and to perform this flattening well is the art of good story-telling.

2-718-281-828 4y ago
interesting perspective to consider writing as graph traversal.

pessimizer 4y ago

I think I accidentally got a good order because of the ad hoc learning I did as a child. I first encountered Commodore/Apple II/IBM through text-mode games/Wordperfect/typing programs in from magazines, then > learned C which gave me an idea of how they worked > learned microcontrollers then understood how the keyboard and monitor worked and how to drive things like that > then logic, electronics and transistors to figure out how microcontrollers worked (and how to make sense of digital voltages) > finally the D-latch. I do have to admit I had a multiuser OS sized hole for a while. The computers in my head were Harvard architecture RISC, generally programmed in assembly, and were driving pin voltages. But it was close enough. So an accidental process of being introduced to things as black boxes, then opening the black boxes. Still linear, but backwards. Needless to say, I loved Code when it came out and gave away at least four copies. I would tell people who thought I knew a lot about computers "this book contains pretty much everything I know about computers." I actually think computing lends itself well to that because it is built in layers. You can talk about how a layer behaves before talking about how it was made to behave that way. The problem I think people run into is that some come to think lower levels of abstraction are a waste of time from a combination of pragmatism ("I don't care how it works") and sour-grapesism ("I don't know how it works so it doesn't matter.")