Re:Non-determinism: a sublanguage rather than a monad
by alt2501 on Thursday September 21, 2017 @07:55AM
(#15302429)
So the 'stupid linalg book' was the turning point? What am I missing here?
Re:Non-determinism: a sublanguage rather than a monad
by agumonkey on Thursday September 21, 2017 @08:54AM
(#15302817)
It was the final nail in the coffin. Linalg was a limit cas of something that was "hard", foreign and serious. The book itself was not all, that would be dishonest, I must say that reading FP for years also brought my mind up. It took me off the material side of things, no more thinking about setting memory with bits, and looking at abstractions instead (doesn't really matter how functions and types are done below, the semantics and algebra is the focus). That said the difference in mindset and perspective in that book (Gareth Williams Linalg 5th edition) told me that a topic can be approached in many different ways. What group A say about a domain may not speak to your brain, while group B perspective will fed your brain just the right way. What is hard is mostly "impedance mismatch", not a limit in the reader.
Re:Non-determinism: a sublanguage rather than a monad
by alt2501 on Thursday September 21, 2017 @03:57PM
(#15306816)
Makes sense, thanks. Hard thing for me is finding the texts that match my impedance.
Re:Non-determinism: a sublanguage rather than a monad
by agumonkey on Friday September 22, 2017 @01:01PM
(#15313876)
Slightly related, when doing the first coursera proglang mooc, I remember newbies crying on IRC because meta evaluation made no sense to their mind. Some of the dudes that had gone through this before and could discuss it without trouble. I remember the odd feeling of being completely incapable of explaining that "simple" thing to the newcomers. Some actual mathematicians say that math makes no sense, you just get used to that, so maybe don't run from one book to another and try osmotic learning. Read some, think about it. At times I just note what eludes me without overthinking it and I come back to it later. Maybe my lower brain made sense of that overnight. One thing is that newbies were very impatient that's another thing to avoid. It's just odd to balance between patience and boredom vs impatient excitment and pain.
For instance I failed linear algebra in college. Even though I loved CGI and 4 dimension matrices are bread and butter of 3D transforms. So I was eager to master it. I tried for 10 years reading a math text book and made no progress. I did lots of programming which helped me get a feel of what notation could be used to and also bought a stupid linalg book on amazon. Their wording and approach was different. All of a sudden everything became trivial. When I go back to the older book I understand what they mean now, but it hid so much meaning. I needed to get a culture of mathematical notation. Some kids might develop it naturally, or be able to grow with the help of teachers, I wasnt able at the time so I was left in the dark.
Now most advanced abstract subjects feel different. It's gone from mental cramp to tiny things that MUST HAVE meaning, and I just have to change my perspective and I'll see too. And weirdly it applies to CS fields too. After that I could finally read fat compilers book full of succint notation. Combinatorics became reachable, same for fancy physics. The symbols carry almost no fear no. And in a way I'm still a bit sad that I didn't know that before, feels like a huge waste, for me but for other kids too.
Lastly, going back to programming, haskell and the likes. After struggling to grok continuations and non determinism, I got a very very different picture of computing altogether. And I realize how crazy bad imperative first programming courses are. I didn't forget how impatients students can be and how efficient a simple statement sequence with visual output can be .. but gosh how far it sets you back in trying to approach different and vastly more useful ideas (say prolog).