I just ordered
Mathsemantics for a hefty $2.10.
Your article's were an enjoyable read and very informative. I'll dig more into you blog tonight.
I've read the Great Good book, Haskell school of music, and I'm working my way through Real World Haskell. I've also read countless blog articles on Haskell.
With a great deal read and understood about Haskell I have no confidence that I can make anything in it at all.
Kurt Vonnegut retyped James Joyce's work to feel a great novel under his fingers before writing his own.
Webster knew English better than Shakespeare. Shakespeare was a master of creation.
To be able to create from a small core and then extend those intuitions with knowledge over time is to me far more effective than mastering language and then attempting creation.
While not rigorous, getting hands on with high level practical libraries and working by example would have built my intuitions far faster than all of the countless reading and toy examples I've done. The problem is, for that approach, there isn't any material for a book or insightful blog post to be written. Mimetics are mundane and unnecessary to those in the know. The teachers seem to be unaware of how their own intuitions were formed.
While learning the fundamentals my mind struggles to imagine how these basic concepts play into the larger picture - how would they use foldr to build persistent? I don't have real answers to those questions but it's a constant distraction.
I am certain that sitting down with a few simple examples of how to use a library like Persistent, without any concern as to how it works, will surely take me from a useless Haskeller to being able to make useful tools that I can use in my career as a composer.
In learning Do notation the books took me through three ways of expressing the same thing before arriving at the sugary syntax that I will likely use for the next ten projects. I don't see that as building a core towards creation, but rather the elevation of a fetishy obsession with language. Children learn the most critical words before grammar - only in language studies does grammar come before vocabulary.
The question is what is the core knowledge that facilitates creation?
That core is a mutating form. It works from the high level downward as it needs to, not from the low level upward because it is thought that it should. There are thousands of articles on how to use raw C++ pointers. One in the know knows to use smart pointers because they facilitate creation.
I constantly read authors of blog posts say things like "I wish I had learned monad transformers sooner." What is a rigorous way to prioritize learning the full scope of Haskell so that creative intuition is maximized? How can I know that Arrows will be generally more effective than Category-Extras for creating things?
If data mining Hackage to find the practical reality of how Haskell is actually being used by people who are creating complete and useful things is not an effective way to learn, what approach is better?
Best,
Casey