World: r3wp

Basically, what I want is this in raw binary I/O:

   print send-to-proc tcp-port input

I can write a CGI to do that, but it will be slower because it requires 
an extra CALL from cheyenne.

rsp doesn't use an extra call

it uses one of the existing rebol processes to deal with it

Does rsp run in binary port mode?

It must do if it accepts binary uploads

Ok, so via RSP, I can use request/posted to get the input cotent. 
 That's good.

So, I guess a short RSP should do what I need.

if you want to let users login, then you need to track a session 
cookie

And then you can run everything as a Cheyenne webapp

This is for R3 Chat, not web.  Session management is already there. 
I'm simply using a port 80 tunnel.

webapp just gives you session management with authentication, doesn't 
mean it is web

Ok... got a one line RSP that works for this:


>> to-string second load write http://host4.altme.com/echo.rsp"testing"
== "testing"

>> r: second load read/custom http://host4.altme.com/echo.rsp[ POST 
"testing" ]
connecting to: host4.altme.com
== testing

Got it!

>> to-string write http://host4.altme.com/echo.rsp"testing"
== "data: testing^/^/"

me too

The "magic" RSP line is:
<% insert tail response/buffer request/posted %>

so, now, need to throw some binary at it....

>> write http://host4.altme.com/echo.rsp#{0209103040508898af}
== #{0209103040508898AF0A0A}

Good to go.

Ok... basically got it working, but it's very slow due to hitting 
a timeout on each packet.

Why is it timing out?

And is this a http timeout or a cheyenne timeout?

The core of the problem is that SET-MODES port [close: false] is 
not supported in 2.7.  Without it, we cannot close half of the duplex 
socket.

In other words, there is no socket shutdown() function.

so the client times out ...

The server's input request packet times out.

Ok, no problem... will add content-length field. Shouldn't take much 
time.

So, R3 Chat server has been moved to Dallas TX running in Cheyenne 
on a cloud.

Not logging in though

working for me.

--- Note: login initiated

*** RebDev Error: server connection failed, is server down? (server-failed)

probably a DNS issue... he has changed servers... so the IP surely 
has changed?

POST /cgi-bin/rebdev HTTP/1.0
Accept: */*
Accept-Charset: utf-8
Host: host.rebol.net
User-Agent: REBOL
Content-Type: application/x-www-form-urlencoded; charset=utf-8
Content-Length: 56


[0.4.0 "Graham" password  login]HTTP/1.1 500 Internal error in REBOL 
CGI Proxy
Date: Sun, 14 Feb 2010 02:05:45 GMT
Server: Apache/2.0.53 (Fedora)
Content-Type: text/plain; charset=UTF-8
Via: 1.1 bc7
Connection: close

Cannot connect (3)

I believe it has shifted from rebol.net .

Found the issue ... http://www.rebol.com/r3/chat.rwas being cached. 
 Refreshed it in my browser until I got the new version and now it 
is working again.

SVN revision 72: File uploading improvements (changes might break 
older scripts)


FEAT: File upload management improved, in-memory uploading removed 
for consistency.

FEAT: request/store specifications changed. Now it renames the temp 
file to its original name by default. Use the new /as refinement 
to move the file (see Changelog).

FEAT: New complete example of file uploading (client and server-side) 
in www/ folder.

FEAT: New config keyword: 'incoming-dir <%/path/> to specify a custom 
incoming folder per domain or per webapp.

Btw, thanks to Will for improvements in the upload.html demo and 
to Oldes for its frequent feedbacks.

its => his

Maybe a silly question, but why Carl uses:
  <% insert tail response/buffer request/posted %>
instead of:
  <% print request/posted %>

or just <%=request/posted%> :-)

He's using <% insert tail response/buffer send-chat request/posted 
%>. So the shorter version would be :
<%= send-chat request/posted %>

Yes... already posted as a comment to his blog.

Anyway.. I thought that APPEND is already a native in R2 as well. 
It's not.

Strange, your blog comment code line appears empty here both in Chrome 
and IE8, but ok in FF??

Might be an XP related issue, will try with Seven.

It looks that FF converts < > chars automatically inside PRE. It 
should be converted on the server side I think.

Can someone give me some guidance on the following I keep getting 
a decoding error.

<%
 	fname: request/content/fname
   	lname: request/content/lname
   	spouse: request/content/spouse
	address: request/content/address
	city: request/content/city
	state: request/content/state
	zip: request/content/zip
	h_phone: request/content/h_phone
	c_phone: request/content/c_phone
	email: request/content/email

 do-sql 'bugs ["INSERT INTO names VALUES (?,?,?,?,?,?,?,?,?,?)"fname 
 lname spouse address city state zip h_phone c_phone email ]
%>

What does trace.log say?

BTW, I normally do it this way

fname: any [ select request/content 'fname copy "" ]

so that the variable is always a string ...

Also, I specify the field names ... so I've never used your syntax


sql: do-sql 'bugs [ {insert into names ( fname, lname, spouse .... 
) values (?, ?, .... ) } fname lname .. ]