1 Sep 2000 raph   » (Master)

Seybold

I spent the last three days at Seybold SF, much of the time helping out in the Artifex (commercial licensors of Ghostscript) booth.

A few impressions first. Apple had a huge presence - a gigantic booth, lots of people doing demos on the machines, etc. I like to see competition for Wintel, but on the other hand, it's very difficult for me to get excited. Yeah, the cube has style, but in terms of performance it's just a 450MHz uniprocessor. I think the cheapo dual Celeron I bought last year for just over a thousand bucks probably outperforms it.

Linux, on the other hand, had virtually no presence. Corel had Corel Draw 9 for Linux tucked away into a corner of their booth, and there were a few companies doing server stuff that just happened to use LInux, but that was about it. This is a bit surprising to me, because Linux really seems to have a lot to offer for graphic arts, starting with generic server tasks and going from there.

There was a lot of XML there. This is hardly surprising, as XML seems to be one of the big hype waves in "digital publishing" right now.

PDF

PDF is becoming the dominant interchange format for graphic arts documents. A great many apps on the floor were showing much improved PDF import and export capabilities. This includes both Adobe's own products (notably PhotoShop 6) and competitors, including the upcoming Corel Draw 10.

PDF is making a lot of money for Adobe. It's not surprising, then, that Adobe is pulling a classic decommoditization strategy. The PDF 1.4 spec (not yet published) has a bunch of new stuff that it's going to be difficult for competitive products to implement. That, of course, includes the blending and transparency stuff that I'm implementing, but also the ability to re-wrap text. They also showed a beta of Acrobat 5 running on Palms and WinCE devices, including the Compaq iPaq.

Electronic "books"

E-books were a major theme at the show, with Microsoft massively showing off their Reader platform, including ClearType. As should be expected from Microsoft, it looks really good - clearly real typographers and UI designers were involved in this product. They will probably get a lot of users just by being so available.

I'm wildly ambivalent about ebooks. The whole concept seems to be organized around "digital rights management." The person who coined that phrase must have been remarkably insensitive to miss the Orwellian overtones. Sure, I'll sign a contract and agree to have my rights managed, and digitally at that. Welcome to the future.

In any case, most DRM is implemented around the concept of the "trusted client," or client software that is programmed to respect some access policies. Adobe Acrobat, for example, has a simple password-based scheme and rinky-dink encryption. Quoting from the PDF book:

Note: PDF cannot enforce the document access privileges specified in the encryption dictionary. It is up to the implementors of PDF viewer applications to respect the intent of the document creator by restricting access to an encrypted PDF file according to the passwords and permissions contained in the file

Such an approach, of course, is pretty contradictory to free software. If free viewers are available, it is always possible, and hopefully even easy, to comment out the if (!password_matches) {...} section of the code, and to distribute the result widely.

Nonetheless, it is important for authors to get paid for their work. To the extent that free software is unable to meet these needs, people will be drawn towards the proprietary systems that can, and I do not blame them.

As ebooks become more popular, it is inevitable that Napster-like trading will become widespread. I think this is also important, as it provides a needed safety valve to protect against those who would restrict our right to read excessively for business reasons.

In the meantime, I find myself very much liking paper books. Aside from the obvious issues of durability (books can and do last over a thousand years), portability, high resolution, high speed random access, and so on, the culture of books has evolved an imperfect but still reasonable balance between liberty, business, and incentive for authors. Libraries, used bookstores, and trading between friends are all popular, respected approaches to sharing books.

My main complaint about paper books is that authors get far too little cut (sound familiar?). However, self-publishing remains a perfectly viable option. Much self-publishing is for "vanity," ie people who pay for the printing of their books because they're simply not good enough for real publishers, but there are some amazing exceptions. Edward Tufte's books are of course beautiful examples, and then you have eccentric thinkers such as Ted Nelson self-publication of the first edition of Computer Lib in 1974.

The document file format to end all document file formats

I find document file formats to be an endlessly fascinating area of study. The most important axis for categorizing document formats is probably structure vs. presentation. Each point on this spectrum has unique advantages and disadvantages. The structure end brings you much more flexibility for editing, analyzing, and adapting (for example, reading texts aloud). The presentation end, conversely, gives a graphic designer much more control over the actual presentation, allowing (in the hands of a good designer) much higher visual quality. The tension between these goals drives much of the continuing evolution of document file formats, and suggests that designing an uber-format is not trivial. Certainly, we haven't seen any good uber-format yet.

PDF has been planted firmly in the presentation camp. PostScript was (I use past tense because it's no longer being actively developed) even more so - it should really be considered a graphics file format rather than a document format. At least PDF adds text searchability and some notion of document structure.

From a commerical point of view, there is pressure for PDF to become an uber-format. However, nailing down the exact formatting brings you to an unresolvable dilemma when displaying in a small window: scale or scroll. Both are bad choices, and both lead to a poorer user experience compared with a more structural approach, which can reflow the text.

Now that PDF is targeting small devices, the issue has finally come to a head. Thus, the PDF 1.4 spec has additions to inch down the spectrum towards structuralism, and is capable of reflowing text intended for display in small windows. At this point, I'm not sure what they did. My guess is that they just bolted on a structural document format. If you reflow, you probably give up any real control over formatting and positioning.

In any case, I'm very disappointed in the quality of the structuralist vs presentationist discourse. Both sides tend to talk about The One True Way. To me, this approach misses important truths. You need to be thinking in terms of the quality of user experience for authors, readers, and editors, in a diverse array of contexts. For a project Gutenberg e-text, pure structuralism is a good, reasonable choice. For a magazine ad, anything less than pure presentationism is probably wrong. For everything in between, well, that's what makes life interesting :)

Even so, it's possible to make better and worse compromises. HTML, for example, neither represents the true structure of documents particularly well nor offers high-quality (much less controllable) presentation. TeX has the amazing feature that it can accurately capture the structure of the document, yet render completely consistently on all platforms, allowing great artistic control over layout. The relative popularity of HTML over TeX is of course evidence that the world is unfair.

Well, that's probably enough ranting for now.

Latest blog entries     Older blog entries

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!