See my comment at

http://www.futureofthebook.org/academicpress/2006/04/learning_from_the_related_proj.html#comment-5139

To share the work we are doing through the Public Knowledge Project, I would focus on how we are incorporating (and test) reading tools to be used in publishing research articles using Open Journal Systems. These tools enable a vareity of readers to build a context for their reading, by connecting them to related "open access" resources from within the field of study, as well as from outside it, through media, government, and the Web. See our demo journal page.

Another thing to consider when looking at publishing something electronically is managing the workflow, and here is a diagram of the online flow as we have developed it for Open Journals Systems.

No doubt this will earn multiple nominations, but I'll predictably mention Ayers' and Rubin's _Valley of the Shadow_ project. A remarkably good realization of scholarly hypertext.

On the subject of other hypertext classics, David Kolb's _Socrates in the Labyrinth_ and George Landow's _Victorian Web_, which he's maintained faithfully for more than a decade.

As for Web-based publishing projects that have stuck with me for years: (1) The venerable PlumbDesign Visual Thesaurus -- still nothing like it; and (2) Scott McCloud's "The Right Number," which is the best introduction I know to what John Cayley calls "writing on complex surfaces."

I'll start by mentioning the Gutenberg-e project. More later.

Some of these push the idea of publishing a bit, perhaps too far:
http://www.michaelberube.com
http://www.theonion.com
http://jot.communication.utexas.edu/flow/
http://papress.com/thinkingwithtype/
http://www.visualcomplexity.com/vc/
flikr

And, of course, I'm partial to http://www.vectorsjournal.org

With apologies--until today I had assumed I need not mention the William Blake Archive among these "good example[s] of something published electronically," but I just noticed that, though Related Projects links to IATH Projects, the IATH home page these days is actually organized around a small subset of IATH projects that doesn't include stalwarts like the Rossetti Archive, the Whitman Archive, and the Blake Archive.

So forgive me for saying this myself--but the Blake Archive (www.blakearchive.org) is a revealing instance of a particular kind of web-based scholarly publication with a notable track record and a profile that is different in several respects from the others that have been mentioned.

The Blake Archive's stated goal is to provide *only* research-grade materials (produced by uniform, well-documented standards) capable of supporting scholarly research. Not that we shun normal people looking for Blake pix,it's just that they aren't in our target audience.

Maybe the distinguishing features of the Archive are its highly articulated scholarly editorial standards for texts & images and its solutions to what I think of as the picture problem, some manifestations of which will be well known to anyone who ever tries to publish someone else's pictures electronically. We have (for example) a cooperative arrangement with 25+ international libraries/museums/galleriesthat allows us to publish, without special licensing fees, both lower resolution [100dpi jpeg] and higher resolution [300dpi jpeg] versions of *every* image, individually color corrected, in the Archive; for one large/significant category of images we offer an experimental image search function alongside text search; we organize access to Blake's output around individual objects (rather than works), etc.

Naturally I'm not claiming that by "solutions" to picture-problem issues I mean universally applicable solutions, permanent solutions, or competely satisfactory solutions. We have only tentative, fragile solutions. But I think the Archive usefully exemplifies, in a way that nothing else known to me does, some persistent and interesting issues that have to be faced in some important kinds of scholarly publishing.

This is too long already--but just one caveat: if you happen to take a look at the site and find something irregular happening, it may be because we're at the long tail end of a tedious and frustrating transition from SGML to XML that still causes new problems to crop up hourly. We'll all be well soon, I'm sure.

Just a few more:

Wikis haven't come up yet so it couldn't hurt to add Wikipedia to the list.

A few others I read regularly:

Nation (non-subscriber content)

Salon (day pass)

Furtherfield

Culture Machine ("Culture Machine is currently taking the form of an international electronic journal. Acting as additions or supplements to the e-journal are the Culture Machine Reviews section and the Culture Machine InterZone.")

Blogs: Billmon; Juan Cole; DefenseTech; We Make Money Not Art

Romantic Circles