LINGUIST List 5.313

Mon 21 Mar 1994

Disc: Mosaic & World Wide Web

Editor for this issue: <>

Directory

1. Lou Burnard, www/html/sgml
2. Don Harlow, Mosaic, World Wide Web, and Lynx
3. Stephen P Spackman, Mosaic/WWW/html: a mostly negative review [Re: 5.301 Mosaic]

>need a SLIP/IPP account and special software.) Mosaic is quite distinct from
>the World Wide Web, which is based on a special SGML superset known as HTML.
>You usually need to "mark up" text in HTML to display it in the Web. Mosaic

HTML is *not* an SGML superset. Rather the reverse. It is a very simple
application of SGML. SGML provides a meta-language for the description
of mark-up languages. HTML is one such, and a very primitive one. But
none the less useful for that!

Pedantically,

Lou

Discussing the World Wide Web, Richard Wojcik wrote:

> If you can forego the sound and graphics, you may wish to use a browser that
> is compatible with vt100 screens: Lynx. I use Lynx from my home account
> (rickweskimo.com), and it suits my needs perfectly. Lynx is also freeware.
> Since I can't use Mosaic with my ancient PC at home, I need Lynx for after
> hours browsing of the Web. Margaret Fleck is right to recommend this
> technology. There are vase amounts of information on languages and
> linguistics accessible through the Web. Not to mention everything else in th
e
> world.
>
And Natalie Maynor wrote:

> > At the moment, only some of this list may be able to use Mosaic. For
> > example, I don't think most of the linguists at my site have good
> > enough internet access.
>
> People who aren't able to use Mosaic can use lynx, a terminal-based
> interface to WWW that won't give you the pictures and sounds but will
> give you the text.

Actually, you don't have to give up on pictures and sounds, even with
a text browser such as Lynx. While you can't look at pictures or listen
to sounds on-line with such a browser, Lynx at least has the built-in
ability to download pictures and sounds for you, so that you can inspect
them later at home -- assuming that your computer has the proper software
to do so (and for Macs and PCs, such software is generally available on-line
as shareware or freeware).

Don Harlow donhnetcom.com
Esperanto League for N.A. elnanetcom.com (800) 828-5944
ftp://ftp.netcom.com/pub/elna/elna.html Esperanto
ftp://ftp.netcom.com/pub/donh/donh.html

Message 3: Mosaic/WWW/html: a mostly negative review [Re: 5.301 Mosaic]

What seems to get lost in the enthusiasm over World Wide Web, and for
Mosaic and the other tools for accessing it, is that its usability as a
professional resource is, well, marginal. Admittedly the network in
germany is not the best, but given the choice between ftp'ing and
printing a ten page document which I can skim in a matter of minutes -
or even ordering an offprint which I can await while getting on with the
rest of my life - and twiddling my thumbs through frequent but
unpredictable delays, without feedback, of five minutes or more (I kid
you not) each time I "just click on a hyperlink" to get the next page
(paragraph, picture, footnote or error message) - I know which system I
prefer.

 (This could be ameliorated somewhat with caching, read-ahead, and
 above all, better networks. Perhaps in a few years, and with
 substantial revisions to the underlying html language, ...)

Reliability is also abysmal: it is frequently impossible to reach even
material you have seen before, because to follow a thread through a
hyperdocument, every single machine in that thread (and there can be
dozens, all over the world) has to be up, connected, and running its
server. For any given machine the odds are in the 95% range, perhaps;
for practical tasks this means you can reach a particular page about
half the time. (To make matters worse, the version of Mosaic I've been
using seemingly can't distinguish between the node being down, its
refusing me access, and the object requested having been deleted. But I
expect that will improve with time.)

 (Reliability could be improved greatly by allowing replication of
 servers so that you don't depend on particular machines being
 active. It is rumoured that they're working on this, but we haven't
 seen it yet, and one would have thought that this was one of the
 more basic requirements for such a system.)

On top of this, it suffers from the usual problem of hypertext: rather
than asking themselves what reading orders will make sense to their
audience, authors frequently, no, typically, take the easy way out and
just put in links at random, as things seem to be vaguely related.
There are no indications of what is important, of what is prerequisite
to what, or even of which two links will take you to the same place. The
result? You spend half of your time looking at things that are either
irrelevant or repetitions, and when you finally surface from your hunt
through the web you have no way of knowing how much of what you went
looking for you have found, how much of the "document" before you you
have read.

 (This could be fixed by - well, by teaching students to write and
 then waiting twenty to forty years. Or by instituting a subweb of
 edited http servers. But the protocol itself would remain remarkably
 weak in expressivity about discourse relations, so the effort would
 remain much larger than that of composing coherent linear prose in
 the first place.)

In most respects I'm perhaps rabidly pro-tech. But this technology isn't
mature. It's like the Mac when it first came out, or the NeXT: it gives
pretty demos, and for a few tasks it may be your only choice. But as a
general tool it's just not really there yet. If you plan to use World
Wide Web for something *today*, wait until after supper for best network
performance and bring a novel to your desk.

Then again, as a source of raw amusement, it's wonderful. For pure
uncomplicated access to Neat Stuff - so long as you have no preconceived
notion of what you're looking for - it's fabulous. Not only are there
pictures, maps, papers, software, standards documents, comic strips and
movies out there, but I found a bookshop doing business in the Web!

More fun than a barrel of monkeys, but as yet not clearly more useful.

stephen p spackman +49 681 302 5288(o) 5282(sec) stephenacm.org
 dfki +1.24 / stuhlsatzenhausweg 3 / D-66123 saarbruecken / germany