tsr printworkflow1 by sandeshbhat


									Print Workflow

PDF on the Fly: Tools and Strategies for Automatic Generation of PDF Files
There is a bewildering multiplicity of approaches and workflows for producing PDF “on the fly,” and the quality of the PDF files produced often leaves much to be desired. Is dynamically-generated PDF suitable for use in prepress and print? And how can you develop new fields of business by using this technology?


he PDF file format is an important de facto standard well beyond the boundaries of the printing world. The most diverse industries have implemented PDF in IT networks and in their client communications. In fact, dynamic generation of PDF has been most embraced outside of the printing industry, mainly because it eases some of the problems encountered when printing HTML on office printers. Now, however, the various industries and markets are gradually growing together. Suddenly, the techniques for creating low-end PDFs (used primarily in the IT world) are also of interest in the printing industry. But there is a great deal of confusion in this field. The Web specialists in corporate IT departments have no idea what to make of the expressions “high-end PDF” and “ISO-standard PDF/X.” On the other hand, the prepress operators of the printing industry frequently cannot get a satisfactory level of quality from the low-end files that come from the IT world, so they have not taken the technology that produces these files seriously. Both worlds can learn a lot from each other about what is important and essential with respect to quality and production methods, while keeping the data as flexible as possible. Both worlds may be wasting important opportunities if they do not develop and implement these new technologies together.

One phrase, many meanings. Anyone trying to select the right methods for dynamic generation of PDF will find that it isn’t easy. The often-misleading phrase “PDF on the fly” can contribute to the problem: Almost every vendor of a system for flexible PDF generation via a server (on a LAN or on the Internet) uses this marketing cliché in describing her product. Still, if you look closely at current production methods, you can clearly distinguish four approaches: • Direct PDF output from a particular operating system or printer driver. Conversion from PostScript via Distiller or Normalizer. Normalizing of PDF via database publishing, followed by Distiller or Normalizer. Direct PDF creation using a library or specialized software such as Pansite’s Formatting Objects Processor.




Actually, the production methods for dynamic PDF generation aren’t so very different from the processes that have long been used to create PDFs. The distinctions are primarily in the control of the necessary tools via the Internet or LAN and in the dataResponse in real time. Marketers often want to respond to Web requests with a nicely formatted, personalized PDF document. This conceptual workflow shows how it might be done.

Volume 2, Number 10 • The Seybold Report • Analyzing Publishing Technologies • © 2002 Seybold Publications


Print Workflow
preparation methods. Perhaps it would be better to speak of “remote control of PDF creation,” especially in situations where Acrobat Distiller (or a competing product) is doing the PDF generation. As we shall see, the only method that really justifies the phrase “PDF on the fly” is generating PDF via a PDF library. Roadmap. Keep in mind that there are more solutions on the market than we can possibly describe. Part one of this article will provide an initial overview and discuss some of the code libraries that programmers can use to simplify the creation of PDF documents. In part two, slated for our next issue, we will discuss several software products and a variety of sample applications and user reports from around the industry. it would be when using Distiller or Normalizer. Instead, it is produced directly inside the computer through the use of calculation and previously-prepared elements. This means there is no need to generate and manage source data (e.g., the Postscript files that would otherwise be used for PDF creation), and no storage space on the server is required for it. In technical terms, a “library” is a collection of commands, class methods, macros and program segments that can be assembled in a development environment, making possible the programming and execution of repetitive processes. Use of a library greatly simplifies and speeds up the creation of sets of data, since the entire process happens in the working memory of the system. Such libraries are mainly of interest to programmers, system developers and integrators who want to integrate specific functionality into their own applications. Endusers are better advised to focus on packaged products from various system integrators; we’ll discuss those later in this article. Adobe PDF Library. One of the best-known programs that uses a PDF library is Adobe InDesign. Version 1.0 was theoretically able to save PDF files directly, but it didn’t do a very good job. This was fixed in version 1.5, and now it is reasonable to say that PDF is one of InDesign’s native output languages. The Adobe PDF Library supports the integration of Adobe PDF into third-party applications. It can be licensed from Adobe or Datalogics, a Chicago-based firm that was for many years an Adobe subsidiary, developing a impressive number of applications of its own based on the Adobe PDF Library. A wide variety of system integrators use the functionality of the Adobe library to add flexible options for PDF creation to their applications. Adobe itself emphasizes that it uses its own library in its applications, since “Adobe PDF” can only be produced by the Adobe PDF Library. At first, this may look like a promotional effort by the Adobe marketing department. But as you look at the many open-source libraries and their widely varying capabilities, it quickly becomes clear that there really are qualitative distinctions here. For example, many freeware and shareware libraries cannot meet the version 1.3 specification. Instead, they simply generate unspecified PDF, which is hardly usable for those in the printing industry. There are exceptions, of course, as will be evident from examples later in this article. Adobe provides the further assurance that future releases of its library will keep its clients current with the state of the technology. In other words, whenever a new PDF specification comes out, the license holder can safely assume that he will soon have the opportunity to upgrade all of his customers. With the PDF Library, Adobe is predominantly addressing the needs of large organizations, so it is making an effort to extend the Library to all relevant

Distiller is the standard
The first choice for PDF generation is still Adobe’s Acrobat Distiller (www.adobe.com), or perhaps a functional clone such as Global Graphics’ Jaws PDF Creator (www.globalgraphics.com). PDF generation over a network using Acrobat Distiller Server or Jaws PDF Server has lately become a standard step in the workflow in many middle-sized and large enterprises. Printers may be familiar with the Adobe Normalizer used in various workflow systems, including Creo/Heidelberg Prinergy and Agfa Apogee. In addition, a number of smaller vendors have also taken the opportunity to bring “fullfeatured” Distiller clones to market. The iDoc Distiller from IMS (www.imsllc.com) is one example of this. But the more perfect and secure the distilling of the PDF files and the subsequent preflighting processes are, the more time they require. The only alternatives are the fast PDF libraries, which are becoming increasingly important. Also, “remote control” distilling is finding increasing application in the automatic preparation of print-ready files. The question that remains is whether the image quality produced by these methods is sufficient.

Native PDF. A traditional graphics application (e.g., XPress) generates PostScript, which must then be distilled to PDF in a separate step. InDesign 2.0 uses an internal library that generates PDF in a single step.

PDF libraries: displacing Distiller?
When you use a PDF library, the file produced is not based on a source file (for example, a Postscript file) as


August 19, 2002 • The Seybold Report • Analyzing Publishing Technologies

Print Workflow
“enterprise platforms.” Currently, in addition to the Mac OS and Windows, there is support for AIX, Linux, SGI Irix, HP-UX, Sun Solaris and Digital Unix. For development, Adobe recommends Microsoft Visual C++ (v5.0), gcc v2.8.1 for Unix-family systems, Visual Age for C++ for OS/400, or SAS/C (v6.50), depending on what the development platform is. These choices seem odd to some experts: VC++ and gcc 2.8.1 are not particularly modern environments, and SAS/C is generally not the first choice when developing for mainframes. The range of functions available in the Adobe PDF Library is impressive, matching Acrobat Distiller’s capabilities. It handles optimization, linearization and compression of PDF files and can embed fonts (Type 1, TrueType and Type 3). Double-byte fonts for Chinese, Japanese and Korean are likewise supported, as are the standard security and document-encrypting functions. Existing PDF documents can also be modified using the library. But the library’s biggest advantage is the ability to generate perfect PDF files of great complexity. Where there is light, there is also shadow. Programming the Adobe PDF Library is quite difficult. Over the years, many powerful functions have been added, with the result that the library as a whole has gotten bloated. It is very powerful, but also quite slow. An example of this is the “Save as” function in Adobe Acrobat; saving just a few pages can take quite a long time. This is caused by the complicated optimizations that the Adobe Library performs on PDF files. Performance is a special problem when the library is running on a Web server. Another problem is that dealing with the internals of the library is hardly a “piece of cake.” It is best to view the library as a monolithic block, suitable for use only by C++ programmers (those who use other languages are out of luck). Adobe is also aware of problems with one aspect or another of the PDF Library. The company confirms that the library is not yet “thread-safe” (which is quite a major limitation for developers working in the server environment) and that “memory leaks” still exist (gradually degrading performance until eventually a restart is required). The API requires a certain amount of patience; the developer must read through thousands of pages of documentation in order to assemble a suitable product. This not only eats up valuable development time, it also requires the developer to learn more about PDF than he probably wants to know. The Adobe PDF Library is, as noted above, very powerful. But you really have to know how to tease the desired functionality out of it. Adobe’s easier partner. Perhaps for this very reason, Datalogics (which is still partly owned by Adobe) offers its own API (called the Datalogics Interface, or DLI) as a companion to the Adobe PDF Library. This product is intended to speed up the creation of PDF

files several-fold. The DLI reduces the redundancy of functions in the Adobe Library, and it avoids entirely the PDF Edit Layer (Adobe’s cumbersome solution to the problem of adding text and other elements to a file). The limit of 2,000 pages per calculation is also lifted. While the Adobe PDF Library works with only one graphic type (the so-called “bitstream format”) and thus requires the programmer to code various image conversions, the DLI version accepts several formats. In version 4.0, these include: TIFF, GIF, JPEG and several vector formats. DLI runs on the same platforms as the Adobe PDF Library. Licensing the DLI and PDF Library is not a trivial undertaking. The programmer pays an initial licensing fee plus an annual royalty that depends on the number of platforms on which his application runs, as well as the number of copies. This can make the use of the library an expensive proposition. (Despite requests, we were not able to get specific pricing from Adobe or Datalogics.) Datalogics provides support for system integrators, including consulting and continuing development of DLI. As you might expect, the Adobe PDF Library is integrated into Datalogics’ own products. For example, Datalogics Pager (a high-speed pagination package) and Datalogics Formatter (a tool for variable-data printing) both make use of the Adobe functions. Formatter is controlled via a GUI or a command-line interface on the Mac and Windows platforms, and it understands most current database formats, including XML, flat files and data provided via a ODBC interface. Datalogics has worked successfully with Creo, Nexpress and HP Indigo in this area, and the Adobe PDF Library is used in these implementations. Among the largest of the company’s 800 clients are Ford, IBM, Datev (Germany’s largest accounting firm), Merck, Boeing and the U.S. government. Competitors. Other vendors, such as the British company Global Graphics, offer libraries of their own, designed primarily for integration into an OEM product. Global Graphics offers its Jaws PDF Library with an accompanying developer’s kit that lets licensees add PDF-generation functions to their own products. The software, written in C++, converts EPS or PostScript files into PDF. It runs on Windows, Linux and Solaris platforms. Jaws PDF Library has practically the same

Datalogics API. The DLI presents a simplified programming interface to the Adobe Library, avoiding some of Adobe’s limitations and speeding up the process of PDF generation.

Volume 2, Number 10 • The Seybold Report • Analyzing Publishing Technologies


Print Workflow

Graphics to go. ReportLab’s online demo shows its ability to combine database content with lastminute changes (supplied via the HTML form seen at left) and produce a formatted PDF presentation (right).

range of functions as the Adobe PDF Library, at least where conversion of Postscript to PDF is concerned. As of mid-2002, the company is still limited to the PDF 1.3 specification. Features for Web-based control are not currently planned; the company prefers to let its OEMs add this capability.

Open source: no-cost PDF creation?
The idea of PDF technology has charmed a wide variety of developers into deciding to do what Adobe did and develop an alternative PDF library. The most diverse libraries for the most diverse applications have been popping up on the Web, some freeware and some shareware. Probably the most successful are ReportLab (www.reportlab.com), Etymon PJ (www.etymon.com), and Thomas Merz’ PDFlib (www.pdflib.com). The three have different areas of specialization: The Etymon library is primarily Java-based, while ReportLab creates PDF using an XML-based reportgenerating language and is written in Python. In addition to its open-source product, ReportLab offers for purchase a line of PDF tools for system integrators. Etymon, developed in 1998 by Nassib Nassar, has become the standard library for generating PDF in the Java environment. Following the Free Software Foundation’s GPL rules, any developer can use these libraries for free in their own free software. They can also license the libraries for use in commercial products. The Java class library works well in pure Internet workflows, creating mail-order paperwork, simple files for printing, or similar undemanding documents. But if the library is tested with large page volumes, it quickly becomes evident that the software becomes very slow and uses up extremely large amounts of memory. The Perl library, PDF on the Fly, (www.ep.cs.nott.ac.uk/pdf-pl/) from the Electronic

Publishing Research Group in Nottingham, U.K., is an open-source project that anyone can freely use, providing the use is not commercial. It generates simple PDF files that are fine for use on Web sites, but hardly suitable for professional use. (But if you look closely, you will see that even this small library has support for DeviceCMYK and DeviceGray.) Unfortunately, the library is only at alpha-release level and is no longer supported. So, in practice, it will no longer play a role. Open-source project goes commercial. PDFlib (from PDFlib GmbH in Munich, www.pdflib.com) has achieved almost a cult following. The man behind it is the well-known international PDF expert Thomas Merz. Merz, also known for his book Postscript and PDF Bible, started developing his own PDF library in 1997. He is pleased with the way it has been adopted, primarily in the IT world, where it has been used for years to add PDF generation to various applications. Like the original Adobe PDF Library, PDFlib is a set of developer’s tools with which PDF files can be generated on a Web server system (IIS or Apache, for example) in a vendor-independent way. The library, which is available as open source or as a commercial product, supports PDF specifications 1.2 through 1.4, depending on the compatibility mode selected. It provides comprehensive compression options for text, vector graphics, image data and file attachments, and it will create an arbitrarily large number of pages in any format. You can tell by looking at the list of functions that this is one of the most professional libraries on the market. Not only are all conceivable font functions supported, but the software will even query IBM eServer systems for the corresponding “system encodings,” as well as for the character metrics for proper formatting of the text. Most of the important international symbol sets are supported, including the Euro


August 19, 2002 • The Seybold Report • Analyzing Publishing Technologies

Print Workflow
sign (the handling of which can still cause problems in Europe). Starting early on, Merz was careful to include the features that were important for prepress. The library provides comprehensive vector-graphics functions, including support for pattern fills, the use of templates for repeated use of text and vector elements. Image reuse (for example, a logo that appears in several places in a brochure) is optimized. The use of transparency (masking) in images is also possible. The hypertext functions are highly valued by PDFlib users. Via PDFlib, you can create bookmarks, links, page transitions such as fades, and much more. In addition to the actual PDF library, PDFlib also offers a PDF-Import (PDI) library. PDI, in combination with PDFlib, lets the programmer import existing PDF pages and add them to new documents being generated on a just-in-time basis. In this way, page backgrounds or parts of pages can be imported and superimposed. The positioning of ads, the removal of selected elements, the scaling of pages, and the copying of an element onto a series of pages—these are all supported. PDFlib offers an easy-to-use API that lets any developer with a bit of experience in graphics and text formatting integrate PDF support. The vendor supports it with a comprehensive Web site; an English reference manual is available now and a German version will be published shortly. Internally, PDFlib supports an impressive variety of scripting languages, which reflects the investment that has been made in the development process. PDFlib supports ActiveX, Visual Basic, ASP, Delphi, Cold Fusion, ANSI C and C++, Cobol, Java (including servlets and JSP), PHP, Perl, Python, RPG, and Tcl. Also impressive is the list of the platforms on which PDFlib can run: Windows, Mac OS 9 and 10, Linux, Sun Solaris, HP-UX and several others, right down to IBM eServers of the i-series and z-series. An initial beta version for Microsoft’s .NET is available for download on the Web site. The open-source version of the library is distributed along with SuSE-Linux by Aladdin Free Public License, and more than 10,000 users get it each month through this channel alone. PDFlib is also widely known and used in the PHP developer network. (Information on PHP integration is available at www.php.net/manual/de/ref.pdf.php.) In addition, Merz and his team have become an important factor in the for-profit “PDF-on-the-fly” domain, having sold more than 2,500 commercial licenses. Licensees include firms such as Deutsche Bank, Wells Fargo, UBS, Credit Suisse, Vodaphone, Telecom Italia, SER, iXOS, Brio, Allianz, the United States Postal Service, EFI and Quantas. In the future, PDFlib plans to pay even more attention to prepress. This may be due in part to its cooperation with the Berlin software developer Callas, a market leader in workflow automation and preflight. Support of spot colors will be improved, color management with ICC profiles will be added, and PDF/X1a and PDF/X-3 will be supported. The company has recently signed licenses with Pantone and HKS. Further changes are under consideration: an XML interface, better options for personalization, and support for RealBasic for the Macintosh. Although PDFlib is open-source (though requiring a license for commercial use), PDI is offered only as closed-source software. However, the terms are not onerous. The various licensing models, which start at around $500, range from a simple runtime license to server licenses and source-code licenses. Market success seems to prove the correctness of this approach. Why is it always the small companies that seem to do things better than the big ones?

More alternatives: commercial libraries and applications
Although open-source projects can be helpful to the programmer, anyone developing a commercial product faces two obstacles: license restrictions and updates. Many (though not all) open-source licenses allow free use of the code only for non-commercial purposes, and many projects are manned by volunteers who work at their own pace, perhaps in their spare time, and who may lose interest when they’ve met their personal goals. It’s nice to get code for free, but sometimes it’s better to pay for code that does what you need. Here are some companies that offer the latter. PDF Tools. Glance AG Software Engineering in Switzerland (or, more precisely, its 2002 spin-off PDF Tools AG, www.pdf-tools.ch) offers a successful library. The company has been gaining experience in PDF publishing and on-the-fly PDF creation since 1995. More than 500 customers and partners now use its products; among the best-known OEM partners in the U.S. are IBM, Boeing and ActivePDF. In addition to a full-featured library, with a range of functions that bears comparison with any competitor, the company also offers its own tools for server-side PDF processing and manipulation. Three tools in particular—the PDF Preparation Tool, the PDF Image Import Library, and

PDFlib for programmers. PDFlib, a professionally oriented PDF-generation library, and its companion PDFimport library offer powerful capabilities. They are available under open-source or commercial licenses.

Volume 2, Number 10 • The Seybold Report • Analyzing Publishing Technologies


Print Workflow
online PDF-generation process is controlled by a Java program that can be integrated into the Web environment via an API. Upon receiving a Web request, the Java application launches the PDF-generation process and makes the new PDF available to the client’s browser. Templates can be used in generating PDF, and they can be populated with information from a database (including images, graphics and personal data) and mixed with XML information. The company’s server suite, consisting of the Prep Tool, e-doc assembler, the Java API and the actual application server, performs the assembly of the data and the PDF generation. The system appears quite flexible. The PDF templates, for example, can be external files pointed to by URL references. All dynamic data is brought into the system via the Java API, XML, or a database interface (JDBC) prior to PDF generation. It’s a solid solution. But the complex licensing approach and the ability to order functions individually leave the potential customer somewhat confused. The company plans expanded support for XSL/FO and PDF 1.4 in the next release. This is high-quality Swiss work that a lot of international partners have adopted. ReportMill. ReportMill, from Cupertino, CA (www.reportmill.com) has also taken an interesting approach. The eponymous software package ReportMill is a developer’s tool for generating dynamic PDF or Flash. The tool lets developers quickly format any type of content into the form of a report and make it available as a PDF or Flash file. ReportMill can combine several data sources in its reports, and it generates individual report documents for systems such as Apple’s Enterprise Objects Framework and WebObjects. Using the Apple Framework (which is available for Windows NT, Solaris, HP/UX and Mac OS X Server) ReportMill can create any static report (or, with Flash, an animated report) without the need for any other plug-ins. In addition to Apple itself, which uses ReportMill in its Apple Online Store, the technology is used by Bell Atlantic, MCI WorldCom and Merrill Lynch. Developers pay about $500 for a workstation developer license or about $5,000 for a server developer license. Websupergoo. For developers in the Microsoft environment who are using ASP, Visual Basic or .NET, as well as for applications based on Active-X, Websupergoo (www.websupergoo.com) offers its ABCpdf product. This tool is particularly notable for the fact that it has been kept “lean.” With it, simple PDF files can be created very quickly. It can also be integrated quite quickly into the customer’s own software, in the various programming and scripting environments mentioned above. (According to the vendor, ABCpdf is not a “true” library. It completely avoids the problems of compiling and linking an external library by strict implementation of the so-called “com8 August 19, 2002 • The Seybold Report • Analyzing Publishing Technologies

Glance’s e-doc assembler. The serverside code combines templates, database records and Java modules to manufacture PDF for delivery through the Web.

the PDF Print Preparation DLL—are ideal enhancements to the library and give the client the opportunity to obtain exactly the modules he needs. There are, however, a few functions already available from the competition that must still be added to this library. These include document locking and linearization. A command-line tool for precise control of individual functions would also be a good addition for PDF Tools AG to make. The company offers a variety of APIs for direct integration using C or C++ on the most common platforms, including Windows, Linux, Sun Solaris and HP-UX. If you want to use the Prep Tool on a Unix system, you must first install the corresponding runtime library. PDF Tools AG provides a complete toolset for normal PDF generation on the server. If you want to create PDF on the fly, you would use the suite of programs called “e-doc assembler.” This supports PDF 1.3. (The 1.4 version is in development.) The entire

It’s a report. ReportMill lets programmers combine almost any kind of data into a PDF page.

Print Workflow
ponent threading models.”) Unfortunately, the library is not suitable for prepress applications, since it generates only PDF 1.1 files—that is, files that meet only the 1994 specification. ClibPDF. FastIO Systems (www.fastio.com) offers ClibPDF, an ANSI C library that is capable of generating PDF files via a user-written C program. No Adobe products are required (other than Acrobat Reader to browse through the resulting file). The API is roughly comparable with Thomas Merz’s PDFlib. ClibPDF is recommended for dynamic generation of simple documents and individual PDF Web sites. It can combine data from several sources into a single report, as ReportMill does. But generation of several PDF files simultaneously is not completely without problems. Support for Chinese, Korean and Japanese fonts was recently added, and the library (unlike Adobe’s) has been made thread-safe. You can also get a companion PHP interface at www.php.net/manual/en/ref.cpdf.php, and there is an interface for RUBY, an object-oriented scripting language, at www2s.biglobe.ne.jp/~Nori/ruby/. The package is free for private use and can be licensed for commercial use. But we doubt you should; the ClibPDF developers have apparently decided to let the project die. No updates have been published since 1999. Only among PHP loyalists is the library still being used. PDFever. Perl Script Studio offers a simple PDF generator under the name “PDFever.” (You can see it at www.perlstudio.de/pdfever/pfrontpage.html. Despite the URL, the developer is based in Canada.) This tool is not comparable with the more powerful libraries, but it has its uses within a restricted domain of text-toPDF conversion. It is particularly attractive when you consider its price: $195. The working method of this Perl program is quite simple. The program takes text that has been created or selected by the user and fills predefined boxes with it. Then, following an analysis by the program, the material is saved as a PDF file. It will be clear to any sophisticated user that highend PDF files for prepress cannot be expected from such a program. But it does have a number of interesting functions. Text effects can either be rendered by PDFever or treated as “real” graphic objects; hyperlinks and URLs are automatically recognized and linked; and grayscale and CMYK colors (in addition to DeviceRGB) are supported. The import and export of XML and the use of templates have recently been implemented. Several simple examples can be found on the vendor’s Web site. Self-effacing. Big Faceless Organization (www.big.faceless.org), in the U.S., calls its Java library for PDF generation “The Big Faceless PDF Library.” The Java class library in its Extended Edition not only supports PDF generation but also provides a PDF browser for

importing data and for (somewhat limited) editing of PDFs. Both versions, basic and extended, offer comprehensive support of Unicode fonts, file locking, embedding of TrueType and Type 1 fonts, bar codes, hyperlinks and URLs. Prepress functions, such as extensive font handling (taking into account ligatures and kerning), process color and spot colors, have been accounted for. The software is completely Java-based. It requires JDK 1.2 and runs on all systems on which the JDK can run (Windows, most Unixes and OS X). Developers can obtain the software for free, but sites that deploy it must buy licenses. Single-CPU licenses start at about $275 for the basic version, while a 10CPU license for the Extended Version costs $4,500. Big Faceless also offers a report generator that is based on the PDF library and offers functionality similar to ReportMill. Llionsoft. Llionsoft (www.llion.net) takes a different approach. With its LLPDFlib 1.2, it provides a simple Pascal library that can create PDF files without a DLL or other supplementary software. The approach follows the example of Delphi’s Tprinter, a printer driver that’s well known in the IT and developer world. Pricing for this library is between $70 and $500. AdLib and Amyuni. The Canadian vendor AdLib eDocument Solutions (www.adlibsys.com) offers a variety of PDF-creation tools. The one we are interested in, AdLib eXpress Server (priced in the U.S. from $1,270 to $1,750), can produce PDF files from more than 300

Variations. These pie charts were generated automatically and packed into a PDF page by The Big Faceless PDF Library.

More To Come
In our next issue, we will continue this topic with a look at some actual installations and conversations with several users. TSR

Volume 2, Number 10 • The Seybold Report • Analyzing Publishing Technologies

Print Workflow
from Amyuni Technologies (www.amyuni.com), the most appropriate application area is the print-on-demand sector. This is how some large vendors (e.g., Xerox) are using AdLib’s technology in their products. If you buy the AdLib PDF Relay add-on (about $1,250) you can access the conversion functionality across the enterprise via a browser. This gives you the ability to send the finished file to the appropriate recipient by e-mail, without requiring a “hot folder” function. But complete Web functionality, particularly the all-important security mechanisms, is not available yet. Prerequisites for the AdLib tools are a Windows NT or 2000 server and the IIS 4. It’s a good idea if the client has a Windows machine too: Our tests with a Macintosh running OS 9.2 weren’t very satisfactory. The technology from Amyuni that AdLib uses should not really be called a library. Rather, it is a PDF converter, somewhat like PDF Writer, which does not make use of PostScript data, but converts directly from Mac QuickDraw or Windows GDI. This makes the PDF files that are created (mostly according to the old 1.2 specification) completely unusable for prepress, even though AdLib has added font embedding and has corrected some bugs. This is clearly a product for the “corporate publishing” market—the output of simple documents without exacting typographic requirements. Amyuni itself also offers a client/server solution TSR and a PDF viewer for Windows. About the Author
Bernd Zipper is a business and technology consultant specializing in PDF, new media and cross-media publishing projects. He is also the publisher of the German news site www.pdfnews.de. He can be reached at zipper@zipcon.de.

PDF Relay. This add-on lets AdLib’s application return completed PDF jobs to the user by e-mail. Without it, AdLib simply deposits its file into a directory, from which the user can eventually retrieve it.

different file formats using no intermediate PostScript file. The user can choose between “native” and “integrated” methods. If native mode is chosen, the file’s source application (for example, MS Word in the case of .doc files) must be installed on the server. If the integrated method is chosen, the files are converted using AdLib’s own proprietary techniques. Unfortunately, the company does not provide a Web interface. But developers can create JavaScripts or Visual Basic scripts that will make the application’s functionality available over the Web. According to AdLib, which employs PDF conversion technologies


August 19, 2002 • The Seybold Report • Analyzing Publishing Technologies

To top