From: jorn@mcs.com (The HyperTerrorist)
Newsgroups: alt.hypertext,alt.culture.www,comp.infosystems.www.authoring.misc
Subject: WWW-REVIEW:  Kiss My Clown Pants, Patrick Lynch!
Date: 03 Aug 1995 20:17:24 MET
Organization: The Responsible Party
Summary: a scathing critique of the Yale Style Manual

Their hypertext theory is stuck in an old-fashioned, elitist mindset that just doesn't fit with the democratic utilitarianism of the WWWeb! The point of hypertext design on the WWWeb is to make the most information available as clearly and easily as possible. Jargon is a pure waste, and formal structures are more often a hindrance than a help, and the conventions of academic publishing just don't work in WWWeb-space...

Possibly the most frequently cited online guide is Patrick Lynch's Yale Style Manual at: <URL: http://info.med.yale.edu/caim/StyleManual_Top.HTML>

Being from Yale, it's not surprising that Lynch assumes a pose of omniscience, and writes in a style designed to impress rather than to communicate directly. This is the approach *required* by academic publication, along with heavyhanded citations to other 'research', and clumsy segregation of appendices at the end, and unhelpfully abstract tables of contents, and a boring 'introduction' dealing with side issues...

So the overall effect is that you feel you're back in school, and the amount of work you have to do before getting to any useful insights is outrageous!

Gareth Rees offers a much better overall model at: <URL: http://www.cl.cam.ac.uk/users/gdr11/style-guide.html> beginning with a 'table of contents' that concisely summarizes *all* the central points of his presentation. If the summary is too terse, all you have to do is click for more detail! If it's *not* too terse, you can confidently skip that whole section, if you're in a hurry (as who ever isn't?!).

Lynch then begins to dictate a set of often-very-dubious rules for ideal WWWeb structure, that seem too often be rigidity for rigidity's sake:

"Many critics and designers of graphic user interfaces have noted the disorienting effect of scrolling on computers screens..." he claims at <URL: http://info.med.yale.edu/caim/M_II_3.HTML >, and proceeds to assert that three screenfuls of information is the maximum allowable. Unfortunately, this is pure fiction! Short pages are annoying, and linking between them is much more disorienting than scrolling...

At <URL: http://info.med.yale.edu/caim/M_II_5.HTML > Lynch diagrams his ideal page layout, with bloated footers and too-scant headers, contrasting this at <URL: http://info.med.yale.edu/caim/M_II_1.HTML > with any freer design style, which he scorns as "clown pants"!

He defines an ideal use for 'home' buttons that completely fails to orient one to the broader structure of his site. Going 'home' from the contents-page suddenly offers two further unexpected 'homes'-- the Med School, and Yale. These would make better sense at the top of the contents page, but he's fixed a rule that forbids this.

Having broken some pages into two-- an A and a B-- he fails to define his 'prev' and 'next' buttons consistently-- sometimes they skip the B page and sometimes not. (A missed opportunity, for his prev and next buttons, would have been to include a hint at the content of those target pages.)

Each page has a bland graphic at the top that proves to be a home button, though it has no text, and is set inconsistently as to which home it will take you to. None of the graphic buttons offer ALT text for lynx.

The various appendices are implemented in a "stairmaster" design, with no prev or next, so you have to keep returning to the ToC (table of contents) if you want to visit them all. (The last one, at: <URL: http://info.med.yale.edu/caim/GUI_Glossary.HTML > looks to me like completely irrelevant Macintosh material!?)

Footers regularly take up almost a full screen, including a multiline copyright notice, and an explicit statement of the URL for the convenience of those printing the text to a file. (Not a terrible idea, but obtrusive here.) The design of the footer makes it unclear if you've reached the bottom of the file, and in fact on many pages he sneaks in text-style footnotes *under* the footer!

And finally (for now) the inline submenu at: <URL: http://info.med.yale.edu/caim/M_II_5.HTML > was apparently never proofed for misconnected links...

But this site is really generally much better than the average academic hypertext style guide! It makes several very good points, and includes many very nicely drawn GIFs...

This critique may be the first in a series... so if your style guide makes similar errors-- get cracking, shape up, or watch out!!! ;^/

j                                 hypertext theory as if the wwweb matters
jorn@mcs.com                      http://www.mcs.net/~jorn/html/hyper.html