Article delegate-en/1272 of [1-5169] on the server localhost:119
  upper oldest olders older1 this newer1 newers latest
[Top/Up] [oldest] - [Older+chunk] - [Newer+chunk] - [newest + Check]
Newsgroups: mail-lists.delegate-en

[DeleGate-En] Re: rewriting problem with cgi program
01 Aug 2001 07:55:16 GMT (Yutaka Sato)

Hi Dennis,

In message <_A1270@delegate-en.ML_>
on 08/01/01(06:32:14) you Dennis <> wrote:
 |> What versions of NNs do you use and are there problems with that?
 |> >From my site, NN6 and MSIE5.0 work with your HTTPS-DeleGate without
 |> problem. NN4.75 works except it does not support 303 response code
 |> which was not in HTTP/1.0 on which NN4.75 based.  I suppose the
 |> problem will be escaped by adding a parameter HTTPCONF="svver:1.0"
 |> to the configuration of your DeleGate.
 |Well, this computer has NN 4.51, but I'm not so concerned about it working
 |with this version. I almost always use MSIE 5.5.
 |So this whole webmail thing has me stumped. Here is something the author
 |of sqwebmail said:

I'm a little stumped because I could not be informed of the concrete
results of your browsers.  Here is a list of results tested by myself.

  NN   4.75 OK except 303 code (which may be escaped with HTTPCONF=svver:1.0)
  NN   6    OK
  MSIE 5.0  OK
  MSIE 5.5  OK

So I think the problem was fixed (and will be fixed) generally
by modifying DeleGate as follows:

  - don't forward Accept-Encoding to the server, to disable content encoding
    in the server which will disable the response rewriting by DeleGate.
    (by the last patch or with CFI script)

  - don't return 303 code to HTTP/1.0 clients, by forwarding request
    in HTTP/1.0 or converting response from 303 to 302
    (by HTTPCONF=server:1.0 or future modification)

  - rewrite not only Location but URI header too (by the first patch)...
    but it might be unnecessary since the URI header was something
    experimental and obsolete

All of these are generic problems to be considered independently
from your problem.  These will be adopted in DeleGate in near future.

 |Subject: [courier-users] Re: webmail non-relative urls

You seem to think it problem that non-relative URLs are sent in
a response from a CGI.  I think it is not problem at least from
the viewpoint of HTTP/1.X standard, although it certainly reduces
its portabilty and narrows its application.
But the URL rewriting mechanism of DeleGate (MOUNT) does not care
whether URLs in response are relative or absolute, nor whether
they are in header or body.

  @ @ Yutaka Sato <>
 ( - ) National Institute of Advanced Industrial Science and Technology (AIST)
_<   >_ 1-1-4 Umezono, Tsukuba, Ibaraki, 305-8568 Japan

  admin search upper oldest olders older1 this newer1 newers latest
[Top/Up] [oldest] - [Older+chunk] - [Newer+chunk] - [newest + Check]