www-apache-bugdb mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Howard Chu <...@highlandsun.com>
Subject Re: protocol/1538: First chunk body is too big, by one byte
Date Wed, 10 Dec 1997 02:23:56 GMT
I've attached a gzip'd tar file with the sample. It was obtained with
the command I gave before:
    telnet www.looksmart.com 80 > z1
    GET / HTTP/1.1
    Host: www.looksmart.com

The file "z1" is the unedited text captured from telnet. "z2" is the
same file
with the first 14 lines deleted (up to the chunk header "eca\r\n"). The
character of this file is the '<' in the "<html>" preamble. The file
"z3" is a
hex dump of z2, showing the byte offsets (16 bytes per line). In this
with the chunk header "eca", the byte at offset 0xeca ought to be the
CR. It
is not, it is a data byte of the actual html page. (Bear in mind that
the byte
offsets in the dump are zero-based, but the chunk header starts counting

from one, not zero!) There are no extraneous CRs or LFs in the
text to account for the extra byte, and presumably there are no added
or tabs either, but I cannot actually verify that case since I'm on the
looking in.

I have deleted apache-bugdb from the cc: list, presumably you don't want

binary attachments showing up in your bug database.
  -- Howard

marc@hyperreal.org wrote:

> Synopsis: First chunk body is too big, by one byte
> State-Changed-From-To: open-analyzed
> State-Changed-By: marc
> State-Changed-When: Tue Dec  9 17:55:33 PST 1997
> State-Changed-Why:
> I'm afraid I can't see where this is happening.  Are
> you sure you are removing the last CR and LF properly?
> Remember, if it ends in the middle of the line the
> last CR and LF need to be removed.  Some editors,
> such as vi, do not allow you to do that.
> If you are certain that isn't happening, perhaps send
> an example?

View raw message