🚀 go-pugleaf

In article <876559733snz@isis.demon.co.uk>
    ian@isis.demon.co.uk (Ian Smith) writes:

> On Friday, in article <df3uEJj030n@sktb.demon.co.uk>
>      pla@sktb.demon.co.uk "Paul L. Allen" wrote:
> 
> > In article <876506961snz@isis.demon.co.uk>
> >     ian@isis.demon.co.uk (Ian Smith) writes:
> 
> > It's not that much worse than 1.0.  A load of new things, but no radical
> > changes to the philosophy of sending and accepting requests.
> 
> You're obviously less easily intimidated than I am :)

I didn't say I understood it all. :-)

> [...]
> 
> > Yeah.  At the moment all the homepages sites have their own IP address but
> > they all map to the same server (or cluster of servers).  They can figure
> > out which site you actually wanted by the IP address the request is sent
> > to.  Soon that will stop.  It will be the same IP address for all, and
> > you need the host header to work out which site you were after.
> 
> I don't look at much on Demon (apart from testing :), but it's nice to 
> know that I won't even notice the transition.

You won't as long as you send a host header.

> > However, thinking about it, the proxy should manage to work that out for
> > itself - you're sending it the full URL.
> 
> Hmmm, I wonder why it doesn't.

Ummm, I just tried it and it does.  Started a chat window on port 80,
went to the proxy, told it GET http://sktb.demon.co.uk/ HTTP/1.0 and
the host header appeared.

> > Yeah, I hate that.  I was digging around the clara.net pages so I could
> > advise a friend on how to do something on the clara webspace.  After
> > battling through loads of frames, each of which URLs with no text but
> > just an image to tell you what it was about), I hit their web central which
> > said I didn't have a good enough browser so I could sod off.  Nothing like
> > making important pages accessible to all, eh?  Even those using Lynx.
> 
> Er, I don't see images or frames :)

Neither do I, most times.  Telnet and spool to a file is how I look at
most pages these days.  If I can't spot what I'm looking for and how to
get it easily I figure it's another fluff site which concentrates on
appearances rather than content.  The times I've soldiered on have not
thrown up any exceptions to this rule.

> I've only once come across one form I felt like filling in and there's
> always more than one way to skin a cat...

Yeah, but it's a pain.

> I still do a lot of page/file collection with a script and it's a real 
> bummer to find a collection of rejections waiting for you instead of the 
> expected goodies :(

Nope.  Because if the page isn't accessible easily it's unlikely to have
any goodies on it.

> > That's nice.  If you dig around <URL:http://www.delorie.com/web> for his
> > web stuff you'll find a gigantic text file showing what browsers have hit
> > his site.  Only 3 of the 6 for my platform show up there, and in a limited
> > number of versions at that.  The file is *huge*.  There's a lot of browsers
> > out there, even if idiots like clara.net decide everyone is running the
> > latest Netscape or IE.
> 
> I'm not sure how many DOS browsers there are, but I doubt that it's as 
> many as six and I don't class what I'm using online as a browser.

If it lets you view web pages on-line it's a browser.  Browsers are a
sub-class of user agents.

> There  are several browsers with configurable UA tags, so just because you
> see a different tag may not mean a different browser or even, as in my case, 
> a browser at all.

Yeah, a lot of hits for Netscape will be from browsers pretending to be
Netscape.  But Netscape users don't configure their browser to pretend to
be something else.

> For a while I claimed to be using Netscape 6, but, much to my surprise, I
> got mail begging to know where I'd got it :)

:-)  Whilst it makes sense to pretend to be Netscape for those broken pages
that insist on it, pretending to be Foo version bar does not.  That list
of browsers and user agents may have fake Netscape entries and even fake IE
entries, but the rest are real.

> > > Almost the only things I ever send are search terms to search engines and
> > > they don't need POST.
> > 
> > Some do because of the limitations on length of GET query strings.
> 
> Ah, well, when the need arises...                        
> I haven't hit any limitation yet and I've used some pretty long strings.

Servers are often limited to 1024 characters in the entire GET request.
Some browsers may have even lower limits on the length of a URL they can
send.

--Paul