freenode/#lisp - IRC Chatlog
Search
18:11:20
phoe
"there's only three people on the Internet: you, me, and this guy who types really fast"
18:23:42
makomo
phoe: hey that's pretty neat. i saw that the reddit 1.0 code was out but didn't know someone made it work
18:24:10
phoe
pfdietz: some code changes to make it work on non-CMUCL, reconstructing the database schema
18:48:46
Shinmera
Though actually the current 1.1 release was erroneously built on 10.12 and thus doesn't run on 10.11. There's a 1.1b release though that does run on 10.11: https://github.com/portacle/portacle/releases/tag/1.1b
18:49:18
Shinmera
I did build it with travis on 10.10, which works, but 10.9 gives errors from brew about it no longer being supported and I didn't want to care to try and make it work.
19:08:39
pfdietz
(scrolls back in log) That unicode's case mapping is not 1-1 doesn't prevent it from working with a comformant Common Lisp, I think. Those beyond-standard characters just aren't treated as having Common Lisp case.
20:11:58
phoe
Okay. You'll hear from me again when I'm waiting for my plane in Warsaw. Over and out.
21:40:43
drunk_foxx[m]
Lexical scope: let, defun, lambda, let over lambda, let over lambda over let over lambda, etc.
21:40:44
drunk_foxx[m]
Dynamic scope - special variables (defined with defvar), and their value can be "overwritten" within the lexical scope
21:41:40
drunk_foxx[m]
First 6 chapters are for free on his website (including the one about lexical and dynamic scoping)
21:45:29
phoe
iqubic: they do not need to. *FOO* is a special variable, as proclaimed by DEFVAR. it means that now all bindings of that symbol are dynamic.
21:45:56
pfdietz
Lexical shadowing is shadowing according to position in the code. Dynamic variable shadowing is according to position in the stack.
21:49:23
pfdietz
In the lexical case the symbol 'x' goes away entirely, typically, except if the compiler has kept the name around for debugging and such.
21:51:12
pfdietz
I believe they are also typically thread-local, although that's not in the standard.
22:26:37
comborico
Can't find the answer to this: 9 / 4 . nine is the dividend, four is the divisor, 2 is the quotient, and one is the remainder. But what is it called when you carry the remainder over to as in 2 1/4?
23:38:10
aeth
comborico: 2 1/4 is a mixed fraction or mixed number. https://en.wikipedia.org/wiki/Fraction_(mathematics)#Mixed_numbers
23:38:56
aeth
Three names on Wikipedia. The one I learned (a very long time ago) was "mixed fraction" iirc.
0:06:43
krwq
are there any good libraries for 2d matrices (not expecting larger matrices than 1000x1000), only need multiplication, addition and some kind of elementiwise map? I've tried magicl but it seems super unfinished, and supporting only complex matrices although seems like it's under development and calls into lapack
0:22:19
pierpa
I'm not sure that for 1000x1000 matrices, sophisticated methods are faster. Are you better informed than me about this?
0:24:24
pierpa
it looks like that for 1000x1000 Strassen may be useful, according to https://stackoverflow.com/questions/22807065/where-is-strassens-matrix-multiplication-useful
0:27:38
krwq
pierpa: seems like n^2.373 is the fastest: https://en.wikipedia.org/wiki/Coppersmith%E2%80%93Winograd_algorithm
0:29:06
krwq
I think naive might do in this case unless there is some ready solution then I'd go with that
0:29:31
pfdietz
Yeah, you have to be careful with algorithms. Constants matter, sometimes more than non-constant factors.
0:31:03
pfdietz
For matrices the bigger wins are in exploiting sparsity, and in block-structuring the computation to take advantage of caches (and also pipelining).
0:34:48
pierpa
krwq: how fast is at multiplying 1000x1000 something that must be optimized for this, say, matlab or mathematica?
0:35:48
krwq
pierpa: do not have matlab at this point, I remember back when I used it it was fairly instant, it's a good point
0:39:36
pfdietz
There was some excitement in combinatorics a couple of years ago related to a problem that had implications for matrix multiplicaton.
0:40:05
pfdietz
If the conjecture (the Erdos-Szemeredi Sunflower Conjecture) had been false, it could have been possible to multiply matrices in O(n^2) operations.
4:01:48
pfdietz
A 3-sunflower is a triple of distinct sets such that the pairwise intersection of each of the three pairs equals the intersection of all three.
4:02:27
pfdietz
E-S conjecture said that a sunflower-free collection of subsets of {1,...,n} has size O(c^n) for some constant c strictly less than 2.
4:27:48
krwq
could someone explain this? (array-element-type (make-array '(2 3) :element-type 'float)) => T
4:35:49
pfdietz
Ok, there is no specialized array for element type float. There IS for single-float or double-float. But float is a union of the various float types.
4:36:28
pfdietz
If you tell SBCL you want arrays that can store any float type, it will give you an array that can store any values whatsoever.
4:37:04
pfdietz
upgraded-array-element-type gives the 'least' element type that an implementation has that will support a given type.
4:37:49
krwq
pfdietz: and how do you know what implementation supports (or spec requires)? is it possible to add your own specialization?
4:39:20
pfdietz
A conforming implmentation must support arrays of certains types (bits, chars, base-chars, T), but can support more. There are some constraints.
4:42:31
pfdietz
No conforming implementation is required to support arrays specialized to hold single-floats.
4:46:33
pierpa
though, I think all modern implementations implements single-floats and double-floats as their corresponding IEEE sizes
4:48:12
loke
This one is a little more surprising: (upgraded-array-element-type '(integer 0 300)) ⇒ (UNSIGNED-BYTE 15)
4:50:39
pfdietz
If T1 is a (recognizable) subtype of T2, then (u-a-e-t T1) must be a subtype of (u-a-e-t T2).
4:52:04
pfdietz
In other words, SBCL must also support (unsigned-byte 15) as an actual array element type.
4:52:40
pfdietz
This also has the bizarre consequence that conforming lisps must have arrays with element type NIL.
4:53:17
pfdietz
They are required to support actual array elememt types of char and bit, and the intersection of those types is nil.
4:54:09
pfdietz
Yes. This is required by the standard, at least for recognizable subtypes. Obviously it can't be done in general for SATISFIES types without solving the halting problem.