[Lula] Bizarre. Firefox can't read a file, but wget can?

[Peter Benjamin]

At 10:05 AM 4/2/2008, Peter Benjamin wrote:
>And the Content-Length might be off.  I count 182 bytes.
>Content-Length: 189

Oh, there are 7 lines in the file, and the difference is 7 bytes.
But that would mean the asc file has DOS EOL, which I do believe
is the HTTP Standard, so maybe my copy and paste and count
was using UNIX EOL.

To figure out if this is a FF bug, here is what I would do.
Write a nph CGI script that supplies the HTTP headers to
Apache, and serve to FF.  I would remove or edit the headers
until FF worked.  Thus, I would know what was causing it.

But first, I would check the official bug list, and forum for such.
It might be known already.

The choose of MIME type x-gzip for the .asc file is disturbing to me,
but it ought to still work.  There is no reason an ascii file can not
be served as a binary file type.  It would preserve the EOL characters.
However, as HTTP standard EOL is two byte, that gets changed...
So, I'm wondering if something is up there.

When I have to serve an extension with a new MIME type, I use
'zip'.  gzip is just as good.  The x- means experimental,
meaning it's not the typical MIME type, and all x- are always
binary downloads.