[EMAIL PROTECTED] wrote:
>
> OK. Two scripts have been posted to REBOL.org though not yet linked
> from the main page. Go to:
>
> http://www.rebol.org/unsupported/
>
> for a brief text blurb and the files. These are the FTP and HTTP
> protocols modified so that a wisely formed read/custom call will make
> use of the partial read functionality of many web browsers and the
> restart command that exists in some ftp servers.
>
> If these changes make it into a full release version of REBOL then
> thes files will be removed from this location never to be seen from
> again as they perish into /dev/null.
>
> Sterling
>
> If I totally messed up and the files fail all over the place for you
> I'll fix them quickly so they work but they are unsupported so nobody
> is going to constantly be updating them with bug fixes as you find
> servers they don't work on. If somebody wants to take it on as their
> script, then just grab it and post it back into the REBOL.org script
> library. Then you can update it all you like.
>
> Party on Wayne! Party on Garth!
Also, remember that the first byte is 0...
read/custom http://www.ibm.com/index.html [restart 1 reend 2]
This reads the *second* byte and the third byte, so the length
is (3 - 2) + 1 = 2
read/custom http://www.ibm.com/index.html [restart 10 reend 20]
Read from the eleventh byte upto and including the twenty-first byte, a
length of
(21 - 11) + 1 = 21
Rebol will also translate line terminators, which is why (on a Windows
machine) -
>> length? read/custom http://www.ibm.com/index.html [restart 100 reend 1000]
== 899
The length that was actually read is (1001 - 101) + 1 = 901, but Rebol
removed two #"^M"'s.
Using the binary refinement is a wise choice -
>> length? read/binary/custom http://www.ibm.com/index.html [restart 100 reend 1000]
== 901
The same applies to ftp, [restart 1] will start reading from the second
byte, and you'll probably want to use the binary refinement to avoid
confusion.
Julian Kinraid
