Search Haskell Channel Logs

Tuesday, March 7, 2017

#haskell channel featuring monochrom, conal, solrize, subttle, sssilver,

solrize 2017-03-07 18:15:29
those concrete running time calculations are usually done with generating functions
solrize 2017-03-07 18:15:33
it was an eye opener
solrize 2017-03-07 18:16:10
he analyze the running time of "find the maximum" (i.e. the obvious linear algorithm) that way
solrize 2017-03-07 18:16:26
at least in taocp vol 1, not sure if that's also in cm
monochrom 2017-03-07 18:16:28
That might be an overkill.
solrize 2017-03-07 18:16:44
well he used it to show the technique, not to find the running time of find the maximum ;)
monochrom 2017-03-07 18:18:11
P.S. Though I said "types" and "catamorphisms" together, do not think that I meant they were covered by the same book. No, they are yet two different books.
sssilver 2017-03-07 18:19:21
how did you guys come to know and understand all this stuff!
solrize 2017-03-07 18:19:25
hmm dunno
sssilver 2017-03-07 18:19:50
you can push a LOT of JSON through TCP using Haskell and still not have to touch any of it
subttle 2017-03-07 18:19:54
sssilver: https://bartoszmilewski.com/2014/10/28/category-theory-for-programmers-the-preface/
monochrom 2017-03-07 18:19:55
"types" being covered by Types and Programming Languages as said (though I wonder if Practical Foundations for Programming Languages is shorter and cheaper). (Also you don't need to finish the whole book.)
solrize 2017-03-07 18:20:09
PFPL is online for free
solrize 2017-03-07 18:20:17
i liked it
solrize 2017-03-07 18:20:42
it was much less detailed than TAPL, but TAPL has a lot of material on subtype systems
monochrom 2017-03-07 18:20:43
"catamorphism" there are some papers and there are other books with category theory inclinations.
conal 2017-03-07 18:20:53
sssilver: I can't tell why you want to learn the math/theory stuff. Maybe it's out of reach because you don't have an intrinsic motivation.
solrize 2017-03-07 18:20:58
which i thought were unstylish around here
monochrom 2017-03-07 18:21:02
Yeah TAPL is big because subtypes.
sssilver 2017-03-07 18:21:36
conal yes it bores me to death :(( but I consciously understand that I should be interested in it
solrize 2017-03-07 18:21:57
i first saw catamorphism in a program derivation (i.e. bird) context i think
monochrom 2017-03-07 18:21:58
But you don't need subtypes for Haskell. And you kind of learn that subtyping is really difficult, so it might be not such a good idea to use it.
solrize 2017-03-07 18:22:01
sssilver, what kind of theory?
monochrom 2017-03-07 18:22:17
Yeah but Bird has a category theory inclination.
solrize 2017-03-07 18:22:27
yeah
monochrom 2017-03-07 18:22:53
The most extreme being Algebra of Programming (now hard to find and expensive). But many Bird papers are not that extreme.
solrize 2017-03-07 18:22:54
for very basic intro the haskell wikibook page about categories is really nice
solrize 2017-03-07 18:23:01
does't say anything about catamorphisms though :)
conal 2017-03-07 18:23:14
sssilver: maybe it's not true that you should be interested. especially if it's for status or self-image ("the moral right to feel like an accomplished programmer"). it doesn't sound fun to learn without a genuine interest.
monochrom 2017-03-07 18:23:21
Ah yea the wikibook is good.
solrize 2017-03-07 18:23:58
the only way to gain the moral right to feel like an accomplished programmer is to write some really good code. then you'll have accomplished something as a programmer and can feel like you've done that
solrize 2017-03-07 18:24:36
see antirez recent blog
solrize 2017-03-07 18:24:47
http://antirez.com/news/112
conal 2017-03-07 18:26:50
i suspect that most people who are good at math / theory got good at it because they genuinely delight in it, not for status.
monochrom 2017-03-07 18:26:59
I don't worry about people learning useless academic stuff. I do that all the time. I was learning linear algebra (specifically the kind for quantum bits). I am planning to learn Lagrangian and Hamiltonian mechanics soon. Useful for my career? No way. Rabit holes? Yes a multiverse of rabbit holes. It's a hobby, I know that.
conal 2017-03-07 18:27:04
... and there are a lot of other things to enjoy.
sssilver 2017-03-07 18:28:51
:(
monochrom 2017-03-07 18:29:49
Sure. I just enjoyed The Lego Batman Musical today. Nothing to do with Hermitian matrices, which I also enjoy.