httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Brian Behlendorf <>
Subject Re: MIME types
Date Tue, 19 Mar 1996 23:22:03 GMT
On Tue, 19 Mar 1996, Paul Richards wrote:
> In reply to Brian Behlendorf who said
> > 
> > 1) You can compile with -DXBITHACK, which means that all files with the 
> > user "x" bit set will be parsed for server-side includes, *except* for 
> > CGI files, which are executed anyways.  Thus, file.html will be parsed, 
> > file.pdf will be parsed, et cetera, and no matter what the original MIME 
> > type, it will be given the MIME type of "text/html".
> > 
> > As you can see, if you have XBITHACK enabled, the only files which should 
> > have the X bit set are CGI scripts and files with server-side includes in 
> > them.  Otherwise, they will be parsed.  So, there's nothing particularly 
> > "flaky" about XBITHACK, it's just that the decision tree can be somewhat 
> > confusing sometimes.
> > 
> Well, I disagree, I think it is flakey. If I set a MIME type mapping for
> a file suffix I think it should override the XBITHACK option. I really don't
> see any point in my pdf files being parsed and I really don't want them
> being returned as text. XBITHACK was a hack to allow html files to get parsed
> and I think that would still be reasonable behaviour.
> I agree it's all a hacky mess but I think my version of the mess makes more
> sense!

I totally agree, I would like to see a server compiled with XBITHACK 
check first to make sure a file is content-type text/html before going in 
and parsing it.  However, "flakiness" implies inconsistant behavior, and 
that's not the case here.  So, someone with CVS access fix it already. :)


--=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--  |  We're hiring!

View raw message