(setq a 2) => 2
(setq 'a 3) => 3
(eval a) => 2
(eval 'a) => 2
(eval ''a) => a
(setq a 2) equivalent to (set 'a 2)
(setq 'a 3) equivalent to (set ''a 3)?
but result is:
(set ''a 3) =>
ERR: symbol expected in function set : ''a
Hello,
I would expect that (setq 'a 3) should throw the same error as with (set ''a 3).
(Autocad's autolisp does so)
Regards
> (setq 'a 2)
2
> a
nil
> (define (foo) (setq 'a 2))
(lambda () (setq 'a 2))
> (foo)
2
> foo
(lambda () (setq '2 2))
>
Quote from: "psilwen"
Yes!^h^h^h^h
No. See Lutz's comment above. (Surprised me, btw. :))
Quote from: "psilwen"
(set ''a 3) =>
ERR: symbol expected in function set : ''a
Yes, this is also correct, because that code induces a type error.
My head hurts.
Quote
Whilst
Indeed, the explanation is clear and valid and sensible, but maybe a bit liberal on the use of the word "type", because
how's the head now?
Quote from: "ralph.ronnquist"
Yes, I did not expect this.
true
> (quote? (quote (quote x)))
nil
Quote from: "ralph.ronnquist"
Head hurty. Tommy no likey. :)
The thing I like the most about newLISP is that it has the least amount of "surprise" compared to other Lisps. And my ideas of "surprise" come from my C/shell/Unix background.
So, with the recent addition to the get-string function, I found myself again "surprised"... because the "chunk" size changes if you put the magic values of ' 00 00' or ' 00 00 00 00' in as delimiters. I feel like the less surprising thing would be to have get-bytes be raw data, and get-string be for getting strings of characters, no matter what the format.
Principle of Least Surprise. Makes my head feel clear and powerful.
Lately, I find myself getting more and more surprised. Every surprise means a gotcha, means yet another bit of state to hold in my head when programming.
yes, sometimes some features of newlisp make me confusion.
for instance, in-place modification
(define (myinc)
(inc 0))
(myinc) => 1
myinc = > (lambda () inc 1)
(myinc) => 2
myinc => (lambda () inc 2)
of course, this feature be really good. Just I've never seen in other languages.
i am learning newlisp in effort.
thanks all.
by the way,
1) i knowns win32demo.lsp. but is there win32gui module written using newlisp FOOP?
deal with BUTTON,TEXTBOX,LISTBOX,MENU,LISTVIEW,TREEVIEW control,etc
2) how to uncompress http gzip stream in memory with zlib? who has complete example of it. code of newlisp, not c/c++
i want to extract HTTP headers, such as Cookie..., and HTTP Body, but "get-url" cannot do it for me. I had to write my own one.
but i don't know how to do :(
in xmlhttp,we can do
getResponseHeader
getAllResponseHeaders
make things so simple.
Does any suggestion? more detail more well.
thanks in advance.
Before evaluation
; one is a quoted symbol the other a list:
(quote? ''x) => true
(list? '(quote x)) => true
; on the top level both evaluate to the same:
(= (quote 'x) ''x) => true
(= (quote (quote x)) '(quote x)) => true
; and
(quote? (quote 'x)) => true
; but
(= '(quote x) ''x) => nil
; and
(quote? '(quote x)) => nil
On the level of evaluation both do the same, which is creating a do-nothing evaluation shield.
That sounds a lot clearer. Guess there was some Paul Graham macro tutorial stuff still rattling around in my brain. The way Paul Graham taught it, ' actually is a macro that expands to (quote). And in newLisp, it behaves a little bit like a macro, even though there are no Common Lisp style macros.
psilwen wrote:
Quote
Maybe you can use gunzip?