Mailinglist Archive: opensuse (4020 mails)

[Allen <gorebofh@xxxxxxxxxxx>]

On Friday 08 October 2004 12:51, Danny Sauer wrote:
> Anders wrote regarding 'Re: [SLE] new v9.2 is out' on Wed, Oct 06 at 18:44:
> > On Thursday, 7 October 2004 01.27, Allen wrote:
> > > > > One thing I'd like to see, is the same Apache RedHat and Fedora
> > > > > come with. I can set up Apache with those easy on my LAN, and it
> > > > > won't try loading my hostname when the page  have running has a
> > > > > link to something in the index.html directory. I'm not sure what it
> > > > > is but RedHat and Fedora are then only two that seem to allow me to
> > > > > set up Apache like that. Slackware and SUSE, my two favs, when I
> > > > > click on Links on my page, it tries loading the hostname.
> > > >
> > > > I'm pretty sure it's the browser that decides how a link is handled,
> > > > based on what it says in the html
> > >
> > > The HTML doesn't have anything to do with it. I've used the same
> > > browser to test my pages as I always do, (Links and Galeon), and it
> > > works different on RedHat, it's how they have Apache set up.
> >
> > Well perhaps you could explain what you mean by "loading my hostname"
> > then? As far as I know, there are three types of links, relative,
> > absolute and URL. if you have a relative link (one that doesn't start
> > with a /) the browser will prepend the current URL up to the last / and
> > submit the request. If you have an absolute link (one that starts with a
> > /) then the browser will prepend the current hostname up to the first /
> > and submit the request. And if you have an URL (a link that starts with
> > http:// or other protocol) then the browser won't alter it at all, it
> > will just open it. Maybe I've missed something, but I really don't see
> > where the web server comes into all this.
>
> It's just a guess, but he's probably referring to the hostname set in
> httpd.conf.  It'd be relevent when you go to a URL that's actually a
> directory, and the server generates a redirect to the index document.
> Apache can be set to either use the servername/client-provided host header
> to generate a "real" redirect, or just do the redirect internally (like
> it does with the ProxyPass directive).  If it's doesn't do the redirect
> internally, or if ServerName is set, then the client can request
> http://alias/path/ but get redirected to http://realname/path/index.html.
> The browser will display the new URL in that case, wich will lead to
> a situation similar to what Allen's describing.
>
> Allen - it'll take some time, but you'll probably benefit from reading
> up the description of all the stuff Apache can do.  Just go to apache.org,
> click on "HTTP Server", and start reading documentation for the version
> you prefer to run (which should really be 2.0 unless you *have* to use
> 1.x).  Knowledge is power, and Apache takes lots of power to use
> effectively.
>
> :)

Heh, I got it now. I used Apache 2 and it worked fine. I don't like reading 
manuals for server software, it makes me feel weak and un-adventure style. 
I'm testing t for now, and if it passes I'll put one on my FTP server. which 
is another server I never read documentation for. ;)