[MUD-Dev] The impact of the web on muds

[Travis Casey]

Greg Munt <greg at uni-corn.demon.co.uk> wrote:


>[Please ignore historical inaccuracies]


Ignoring historical ones, but there are some non-historical ones I'd like
to point out...

>In 1990, the Internet was a significantly different beast from what it is
>today. It was largely unknown outside of academic circles, and was
>entirely a textually-based medium. You had your hacked-up telnet programs,
>your ftp and mail programs, and you had lynx and gopher. The World Wide Web
>wasn't even a dream. All of this was running from a unix operating system
of
>some kind.
>
>It is little wonder that The Big Three (DIKU, LP and Tiny) were all
>text-based games which all supported the TELNET protocol. These three
>still dominate the mudding field, although the Internet itself has
>changed drastically: gopher is little more than a memory now, lynx
>surviving through sheer strength of will, of the unix die-hards. The
>Internet has become a graphical medium. The text-based tools, such as
>mail, ftp and telnet, these are still around - albeit with a GUI bolted
>on top of them.

FTP is text-based?  It's always been able to transfer binaries... if one
can say that FTP is text-based, then it's equally true that the Web is a
text-based interface "with a GUI bolted on top."  After all, the first
web browsers were text-only, and the HTTP protocol is entirely text-based,
just as the FTP protocol is.

>What happened to MS-DOS has happened to the Internet.

It's become a bloated monstrosity?  (Sorry, couldn't resist that one.  :-)

>So what does all of this mean for muds?
>
>If you look at the mainstream, not too much; the old text-based games are
>still popular - and the majority of people use what they know, rather
>than what might be better, or of a higher quality, or easier to use. What
>I am saying, is that the medium is stuck in the eighties.
>
>Many of us are writing scratch muds. This, then, is an ideal opportunity
>to explore the new interface. This is an opportunity to lead muds into
>the arena of the GUI, where the rest of the Internet has been for years.

Of course, it's also possible to build GUI clients on top of existing
muds, just as has been done for FTP, the Web, and other Internet services.
The real obstacle is not the fact that muds are text-based; it's the fact
that there is not, and never has been, a real standard for how muds
describe objects, rooms, etc.

>It would be hard to disagree with the assertion that The Mud Connector
>(www.mudconnect.com) would not have become as popular as it is, within a
>text-based medium.

Huh?  I just took a look at the URL you reference, and everything on those
pages that is of substance seems to be text.  There's no reason that I
couldn't access it via lynx or another text-mode web browser and make
full use of it.  The ability to point-and-click is not unique to graphical
interfaces; there are and have been many text interfaces which support it.

>Count the number of Internet users on a global scale,
>and compare that figure to the number of mud users. I would be surprised
>if there was not a massive difference. The simple fact is that the
>majority of Internet users do not know what a mud is - and if they know
>of a talker, it will probably be something web-based, like PowWow. A
>large proportion of people sign up with an ISP considering the Internet
>and the Web to be synonymous. This attitude will not be changed - so it
>must be taken advantage of, by us.

I'll agree that most people signing up with an ISP today consider the
Web and the Internet to be synonymous.  I don't necessarily agree that
we *must* take advantage of this -- each of us has our own ideas about
who we want on our muds.  It's quite possible that some of us don't want
people who would want a graphical interface.  (Please note; I am *not*
saying that this is my attitude -- I'm simply pointing out that saying
that a group of people *must* do something is almost always an
overgeneralization.)

[much cut]

>So let us now turn our attentions to the GUI. Ignoring the initial
>problem of creating an attractive and intuitive interface, we have the
>problem of increased ease of use resulting in decreased functionality.
>Having an icon for each of 60 verbs would definitely result in
>unwieldiness. We need to consider the extent of the conversion from text
>to graphics. At one end of the scale we have text I/O planted onto a web
>page - at the other, we have removed text altogether, and are left with -
>perhaps - something which looks a little like Doom, with extensions.
>
>Can we really remove the interface for text input? Or would doing so
>decrease the possible functionality too much? Perhaps those involved on a
>professional level could shed some light on this issue.

IMHO, the need is not necessarily for a "web interface" -- at least, not in
the sense of having the mud show up in a web browser window.  What's needed
is a simple way for people to get started with muds.  Interface enhancements
are a separate issue, which I've just discussed in another post, so I won't
discuss them here.

>Is TELNET dead? Is the text interface to muds dead? Will muds catch up
>with the rest of the now-graphical Internet? Should they? Will new
>JavaScript clients evolve, to ease the transition? Which of us here have
>been involved with mud GUIs? Perhaps you can give your opinion on this.

Telnet (it's not an acronym) is far from dead; the protocol doesn't
specify much about the client end, so it can be used for much more than
just emulating a dumb terminal.  Telnet isn't limited to sending just text
characters either... but even if it were, that wouldn't mean much.  After
all, a PostScript file is nothing but text, but it contains the
*instructions*
for creating graphics.

IMHO, something like the telnet protocol is more useful for muds than a
true web-based interface, with JavaScript and CGI, would be.  One of the
biggest problems with developing web interfaces is the lack of a persistent
connection.  HTTP/1.1 does some work on this; it might even do enough, but
I'm not sure, not having looked at the changes closely.  Most mud commands
are short; opening a new connection for each command is very wasteful.
Further, the client would need to send an identifier or status information
to the server with each command, so the mud would know *who* the command
was coming from (you can't even guarantee a stable IP for a web connection;
it might be coming through a firewall which assigns IPs dynamically or
through a proxy with multiple connections and IPs.

I agree that a move toward a GUI is a good thing; I'm not sure that a
first-person animated 3-D view is the ideal GUI for muds, though, and I
feel that the interface will need a persistent connection.

The ideal thing to do would be to establish some sort of standard for how
muds can interface to (possibly graphical) clients; a protocol which would
support text, graphics, and possibly even sound.  There have been efforts
like this made in the past -- however, none of them have been widely
adopted by muds.
--
      |\      _,,,---,,_        Travis S. Casey  <efindel at io.com>
 ZZzz  /,`.-'`'    -.  ;-;;,_   No one agrees with me.  Not even me.
     |,4-  ) )-,_..;\ (  `'-'        rec.games.design FAQ:
    '---''(_/--'  `-'\_)      http://www.io.com/~efindel/design.html