Hi all,
I have been trying ;
(get-url "http://www.foo...bar.com") which works beautifully but;
(get-url "https://www.foo...bar.com/") just seems to spin forever without returning anything. I usually have to kill newLisp (on winXP) as it stops responding.
I haven't tried it on my linux box yet, does anybody know if get-url supports https out of the box, I couldn't find anything in the help doco.
Regards, Simon Cusack.
get-url supports the HTTP GET protocol.
I do not think that it does support the secure protocol.
No SSL protocol support yet but you can specify a timeout in milli seconds:
> (get-url "https://www.foo...bar.com/") 2000)
"ERR: timeout"
>
because https is not implemented, newLISP will still try port 80, but you can force it to the SSL default of 443:
> (get-url "https://www.foo...bar.com:443" 2000)
"ERR: invalid response from server"
>
now it will return right away, but again these are all games because https:// support is not implemented.
As a work around use the newLISP 'exec' funtion with the 'curl' UNIX utility:
(exec "curl https:// ......")
This will return all standart out from the curl utility in a list of strings to newLISP.
Lutz
Quote from: "Lutz"
No SSL protocol support yet
Any plans for this? Would be rather useful if get/put/post-url functions supported this.
Most Unix systems have libcurl on it. It is possible to write a module to do SSL via https and other stuff. Until such a module exists you could use the method, Cormullion uses to post to blogspot.com:
http://newlisper.blogspot.com/2007/06/posting-to-google-blogger-using-newlisp_16.html
Then there is also the Amazon module to communicate with there ec2 data services via https:
http://www.newlisp.org/modules/various/amazon.lsp.src.html
The module assumes you already have keys.
Yeah I'm currently using 'curl' right now, was just wondering if there were any plans for building this into newlisp.