Charles Vilnis – BookRouter & Allusive Information Systems


Winter 2002 (Vol. III, No. 4) Table of Contents

URLS: http://www.allusys.com and http://www.bookrouter.com

When Shirley Bryant approached me with a request to do an interview for the IOBA’s The Standard newsletter, it forced me to reflect on what it is my colleagues and I do at Allusive Information Systems and how we got here to be doing it….

From personal experience, I can tell you that book dealers and computer people are very much alike – generally independent, a bit eccentric and definitely passionate about the ingredients of their respective vocations. Just as book dealers do much more than simply buy and sell books, computer folks do a whole lot more than simply sit and code, or create computer systems for clients. The world of books and the world of computers exist – both cultures of communication.

I began as a citizen of the world of books….

After a stint in New York working for a Japanese trading company where I could use my Japanese language skills acquired in college to a practical end, I moved to northern California in 1975. I was working as a finish carpenter in 1977 when an acquaintance asked me to become his partner in a small used bookshop in a small apple town where the freight trains still ran down the main street (called Main Street, by the way) on their way to the cannery. The day I came by to talk to him, he gave me a set of keys and put my name on the checking account. Suddenly I was a book person – a full partner in a business that had grossed almost $14,000 the year before!

Most instances of autobiography now turn to the enormous success that followed – fortune and the respect of colleagues accumulating through the years – the subject turning back in self-satisfaction to consider a life well led…

In my case, I would have to say that, while I hope I have gained at least some regard from my colleagues and learned a little bit about the craft of dealing in the books of East Asia, what I really found that day over 25 years ago was a passion, a love for books and for the world of books. I did become a member of the Antiquarian Booksellers’ Association in 1980 (at a time when the ABAA was full of young booksellers in their late 20’s to their early 40’s). A few years later, I moved to Boston and became involved in the life of Helen Kelly and her business, the Boston Book Co., founded in the 1970’s. It was Helen who gave me the chance to see what a large urban book business could be and who opened my eyes to the joys of domestic life – we married and had a son in the late 1980’s.

Helen had one of the first computers in the old book trade: In 1982, she bought an original IBM PC with specifications laughable by today’s standards – but it was top of the line (with a hard drive, no less) and cost her about $10,000! Not so laughable a sum then. Soon after I merged my book business into hers, I became involved in the care and feeding of that by then venerable PC and its successors – then creating a network…..I was hooked.

Helen had introduced me to my second passion – the world of computers.

Helen had equipped that original PC back in 1982 with Bookease – Marc Younger’s professional bookseller’s package, built on DOS, which he had created for his own bookseller wife Helen in the same year. By the early to mid-90’s I had become Boston Book Company’s information technology department – our computer network had grown – we had gone far beyond the first PC to the blistering speed of the 486! Marc had succeeded in creating a version of Bookease that networked well and stably, we had bought a commercial space specifically for our rare books. Our book business was prospering… we cranked out catalogues, published lists and I personally had also become our Asian book department of one and was doing a land-office business with Japan.

But off on the horizon, no bigger than a man’s hand, a dark cloud was churning. The Japanese market was in decline; that was clear. Techno-friends were starting to talk about the growth of a “network of networks” called the internet and I decided to take a look. By 1992 I had an email address and a shell account so I could start exploring the ‘net (not via the Web, though, it was too early for that –I didn’t even use Windows….). But even in my halting fashion, I found my way in cyberspace and what I found astonished me: the internet was full of librarians!

Viewed from the near side of the dot-com divide, the internet of those days seems quite quaint. Even folks with “com” domains were careful to give as little offense as possible to the dominant anti-commercial culture. The ‘net was for the exchange of information – if you wanted to share bibliographical information as a bookseller with your colleagues and with the library and academic worlds, that was fine, but booksellers should not flog their wares.

But soon there was an accelerating move to make more than pure information available over the ‘net – fueled largely by the splashy, graphics-heavy and user-friendly World Wide Web. By 1993, Boston Book Company was on the Web and by early 1994 we had established a searchable database of our wares there. Interloc was still a Bulletin Board Service not open to the public, ABE and Bibliofind were just stirring, Amazon had only burned through their first few millions – spring was in the air.

As the dot.com boom evolved in the mid to late 1990’s it became clear to me and many of our colleagues that the easy answers for marketing that the Web seemed to provide for small businesses like ours were not an unmixed blessing. Bibliofind, ABE, AntiQbook in Europe, Interloc (later Alibris) and soon many other sites were jostling each other for a piece of the pie – trying to lure dealers to post their listings and to draw eyes to view them as well. And just like that we all seemed to get bulldozed by big money – folks you would never expect it of were suddenly interested in books. Capital was being poured into the process – relatively big fish were being swallowed by bigger fish in the internet world – Bibliofind was grabbed by Amazon which had discovered a sudden interest in old books and ways to market them and Interloc became Alibris absorbing Bibliocity, hoping to create a whole new market for old books by raising their visibility through the pages of the New Yorker…. It was a strange time.

There seemed to be a two-fold attack on our way of life. On the one hand, it was clear that money was trying to take over the old book world – whether that was sensible or not, whether there could be any reasonable return for the outlay of cash required to get noticed on Wall Street. One could foresee a future in which old book dealers would become glorified scouts for the various huge marketing venues. Not the sort of future most of the book people I knew relished.

The second attack (or just plain change, to be more neutral about it) was a revolution in our day-to-day activities. Traditionally there had been a “food chain” in the old book biz – at least since the demise of the truly great generalist booksellers – dealers sold to other dealers up that chain, retailing to their own customers along the way and wholesaling to specialists or those with more savvy about marketing at the high end. The romance of the vagabond bookseller hunting for gold on the less-traveled roads of America was being written every day, even just a few years ago. Yet the growing access to information (and misinformation, alas) over the internet meant lots of people knew lots more about books and markets than ever before, and the out-of-the way bookshops got much less remote. In the meantime, the multiplication of electronic venues, and the growing need for sophistication in all things digital, meant that there was less time for actually handling the books we all love…. Our customers, who had relied on us for expertise, were going to the search sites or to Amazon, to find information certainly, but more often to find stuff cheaper.

Ironically, the computer, which was supposed to make everything easier, ended up on the verge of separating us from our wares, making us work harder and maybe even making it harder to make a living.

The old book world responded with a whole raft of marketing solutions – ideas like the IOBA, high-end venues like WorldBookDealers, the ABAA/ILAB website and search engine, the joint ownership represented by TomFolio, etc. Each was an effort to reclaim the old book world from the new giants. We had all seen too clearly the utter demise of the independent new book dealers. No one wanted to follow them.

I supported all those efforts as best I could, but there had to be an answer from the computer side of things, as well. In that atmosphere, I decided to try and create some tools that would make it easier to interact with the brave new world of bibliospace, to alleviate some of the new drudge work that our friend, the computer, had created for us. As a member of Boston Book Co., I felt I was a good position to understand the joys and pains of 1990’s bookselling – we operated a used bookshop near the BU campus, and had a rare book operation in Jamaica Plain, a Boston neighborhood. At the rare shop we handled general antiquarian items, but we also pursued a fairly serious specialization in the world of Japanese books. In short, because of our internal diversity, whatever we could create to help us at the Boston Book Co. would probably help many of our colleagues, as well.

If I turned to my book business for inspiration and experience, I turned to my next-door neighbor for digital muscle. My neighbor, Mark, is a very skilled database developer with over 20 years experience in creating database tools for hospitals and educational institutions. He and I started to discuss the idea of using the internet not merely as a marketing medium but also as a foundation for helping booksellers more easily control their businesses as well. Our talks, at first theoretical and purely speculative, soon became very practical and, by early 1999, we had formed an LLC: Allusive Information Systems. Mark quit his job and came on board as a full-time in-house developer.

Over the course of those conversations, we developed a set of principles that would be the underlying theory of Allusive. Allusive was set up to fill a niche in the old book world, to develop a set of answers to some problems that I had learned about too clearly as a bookseller.

First of all, as computer people, we had to learn to listen to the customer, to the bookseller. Technical people tend to see the world in terms of their own latest technical “fad” – they are excited by the new, want to explore the “bleeding edge” of technology. We were determined not to prejudge a solution by its technical “sexiness” alone. What works, works.

On the other hand, it is true that clients and especially booksellers who aren’t computer literate have a certain Luddite strain (which I don’t entirely disagree with) that has to be dealt with. Sometimes the old way of doing things has simply gotten too complicated. I often find that booksellers use a Rube Goldbergian mess of structures, techniques and equipment that has evolved in response to practical needs as they have changed over time, which “works”, kinda, but which is so inefficient that it isn’t making economic sense for the business.

The result of those two problems is that booksellers are sometimes stuck with systems that are either alien and unusable or antiquated and overly complicated. In both cases, the systems are further compromised by being unreliable. And where do you find somebody to help you fix a problem at a moments notice? If your technical “faddist” moves to Peru, who will understand what latest gizmo he or she has equipped you with? On the other hand, if your system is a crazy quilt of operating systems and older equipment, who can pull it all back together when it falls apart?

Mark and I came up with the notion of creating a “toolbox” of software, both for the dealer’s office and for use over the internet. We figured that the basic task of the bookseller was the creation, storage, conversion and transfer of information. The average bookseller has data that they need to store and have access to – book records, sales records, wants, etc., etc. Most everybody has some sort of database that they use for helping them with that – a software replacement for the raw memory skills and the many, many index cards we all once used (and once had, in the case of memory skills). Those databases are sometimes brewed at home out of Access or a spreadsheet or Filemaker Pro. Alternatively, there have been dedicated bookseller programs around for many years: Bookease, Homebase, BookTrakker, BookHound, etc., etc. Some were built from scratch, some on a base of Access or Filemaker – all useful as far as they went. It was the integration of those existing tools with the internet, and the growing marketing opportunity the internet represented, that we wanted to aim at.

It was at that point in 1999 that an interesting practical circumstance arose that put a point on our need to come up with tools for integration. I had been friends with Marc Younger, the developer or Bookease, for many years. He had handed over the maintenance of Bookease to other people soon after developing some simple tools for exporting internet-ready files out of it in the mid-1990s. The handover was not a success and became a classic case of the two failings we were seeing: too much tech on one hand, too little service on the other. The first group that took over Bookease for DOS meant well but never seemed to be around when there were problems or questions from the Bookease booksellers. In despair, Marc made an arrangement with another developer on the west coast – this fellow decided to rebuild Bookease from scratch as a Windows program on an Access foundation. Bookease had been brilliantly conceived by Marc and was a very stable and expandable piece of software built on a Foxpro foundation. Access, on the other hand, is not a foundation for a mission-critical business environment – at least not as it was used in the new “Bookease Pro” for Windows. The results were instability, loss of data, bookseller recriminations and the new developer’s disappearance from the scene. Bookease Pro was in shambles and Bookease Plus for DOS, while humming merrily along, was running into a limit we all saw in the late 1990’s – the Y2K dilemma. Bookease for DOS needed an update desperately and time was running out.

It was at that point that Marc Younger came to us, and asked if we at Allusive might take over the maintenance of Bookease for DOS. Though it seemed a bit far afield from our primary mission of building all new tools, I saw it as an opportunity to establish a relationship with a whole group of longtime computer users who still used the old Bookease program, and an acid test for our ability to integrate the old and the new with software “bridges.”

In late 1999 we set up a Y2K upgrade system, contacted those many Bookease DOS users who had been languishing a bit since the mid-90’s, created a service contract system and Y2K converted every dealer who wanted to sign up before the January 1 witching hour. It was a great experience, an early success, and it taught us a lot about where we should be going. It was also the genesis of the online data conversion system that would become BookRouter.

As I mentioned, Bookease had a rudimentary capability to export book records in UIEE format, which had originally been developed by Tom Sawyer of the then Interloc as a lingua franca for the transfer of book data. That was great as far as it went; one could log on to the various indexing sites and upload the export files, or one could use email to send an attachment to a few of the sites, or send by FTP, or…. There were many ways to keep up – too many it seemed and I–and many of my Bookease customers–was spending one heck of a lot of time creating formats and logging on and sending by email, and it didn’t seem to make much sense.

It was in that environment that my database-savvy neighbor and I put aside the development of yet another online search site, and decided to create a universal online data converter which could take pretty much whatever text-based data that was thrown at it, convert it to an appropriate format for the various indexing sites and send it on. We also thought it would be good if we could do things to the book records on the fly…. Perhaps the dealer would only want to send books over $100 to one of their sites – or perhaps they wanted to have an instant “sale” by lowering prices, or they wanted to add an html link to their homesite to each and every record. We (or should I say Mark) created a “swiss army knife” data converter online at http://www.bookrouter.com and our first “bridge tool,” the prototype for the whole concept, was born. BookRouter went public in mid-2000. It has gone through many changes and improvements since, but the original simple idea remains the same: Upload once to us and we will take care of the rest – including all the toing and froing and talking to the sites when a mistake occurs. We wanted BookRouter to become the ultimate time-saver. At first customers were skeptical – it was really quite easy to upload files after all, and it didn’t seem to take much time – skeptical until they realized just how much time they really had been spending now that they weren’t spending it anymore.

The Bookease experience also showed that there really was a need for a multi-tiered approach to the in-house bookseller software dilemma. Bookease DOS had been built on a state-of -the art foundation that was very robust, but the platform was creaky now and it was clear that a lot of functions (direct export of data to emails, for one, “assisted cataloguing” like Homebase’s ISBN lookup, for another) that were so necessary for modern commerce could not be added via DOS. But the Bookease Pro disaster showed that simple conversion of Bookease functionality into a Windows-based environment was not the answer by itself. A “one-size-fits-all” mentality wouldn’t work where some folks were dealing with thousands and others potentially millions of book records. New thinking was needed. We had to look around for new systems, new approaches.

We had been using a Linux data server at Boston Company for many years. Linux was free, or close to it, and it was incredibly stable and bulletproof, as well. It networked well with Windows and made it possible to set up data server boxes separate from the rest of the Windows network and thus protect against the data corruption that came from inevitable Windows crashes. In addition, hardware had gotten so cheap that there wasn’t too much economic pain involved in setting up those Linux servers, as well. Any “entry-level” machine off the shelf was more than capable of doing the job. Over the course of time, I have set up a number of Bookease users with whole new systems – integrating their old hardware with a new box that hummed along – delivering data. It was almost comical – I would stop by a client’s shop to deal with some problem or another or just to check in – and the server would be running perfectly. I would reboot it just since I was there anyway. There are data servers at some of my clients’ places that are only rebooted once or twice a year.

So we had grown busier – besides the toolbox business, as exemplified by BookRouter, we were also doing systems consulting, taking care of all the Bookease DOS users, setting up hardware/software networks, consulting on equipment acquisitions (we have found that most folks overspend on hardware because they are insecure about what they really need), etc. Help was necessary. Especially since early in 2001, Mark, having built BookRouter and the online search engine, got an offer from MIT he could not refuse.

We brought in a part-time developer, Steve Clay, to take care of further work with BookRouter and the search engine on a consult basis. In addition, we hired an office manager, first Cathy, then BJ – who sends out bills to our customers and makes sure at least some of us get paid. We added Scott Pezza to take care of customer service. He is the one who fields questions, sets up new clients, writes conversion scripts for the occasional exotic data format beyond the reach of even BookRouter’s built-in tools – in short, he is the glue that holds the day-to-day operation of BookRouter together. And he is a gifted developer as well. Over the last many months, he has been hard at work creating a new tool for the toolbox: He has written an order management system which integrates and organizes all the email orders from the various online listing services – puts them into a coherent form, generates reports, creates packing lists and invoices, keeps track of clients, etc., etc. All this can be done either locally over an intranet or over the internet when the user is away from home. We haven’t come up with a name for this analogue to BookRouter, a Swiss army knife for orders, the results of our labors, but we are thinking about one – for now it is just “OMW” – Order Management Web.

We are refining the online search mechanism, which now exists in-house as a database of our BookRouter users. Soon we will be offering that database “backend” as a feature to our BookRouter clients: they will be able to search their own inventories from their own websites without potential customers being diverted into the ordering systems of the indexing services. The search results, logo, colors and all, will appear to come straight from the dealer’s own search engine.

We have outsourced development work on the successor to Bookease DOS to a group of programmers in Europe and further east and are working closely with them to dot the “i”‘s and cross the “t”‘s. The program is called Biblioware and it is a total departure from the standard in-house bookseller’s database program. It has a front-end with all the bells and whistles derived from the best of the bookseller programs, a user interface that can potentially plug into any modern database you please – whether it be our home grown Java data backend that will allow the Java version of Biblioware to be totally self-contained and totally compatible with any operating system, or a heavy-duty Linux/Unix/NT database for heavy lifting client/server operation, for people with thoughts of putting a million book records online.

Carrying the “toolbox” idea even further, we are also working with a fellow who has created a wonderful assisted cataloguing system that is at least as accurate as Homebase’s ISBN lookup feature, but which will also allow the lookup of older, pre-ISBN material, as well. Its functionality will be integrated into Biblioware, the successor to Bookease DOS, but it will also be “modular” so that it can integrate with any bookseller’s program out there. One needn’t have to convert to a whole new system like our Biblioware or stick to an essentially “light-duty” database like Homebase in order to derive the benefits of assisted cataloguing technology.

Obviously, even with outsourcing some development, we are going to need more hands to make all this work. Scott needs an assistant, I need an assistant, we need an extra full-time developer on staff – the wish list is long. We “come in” in the morning (sometimes we log in from far away, actually), clear up any problems re Bookease or BookRouter first thing, then start to work on toolbox projects or in setting up connections with new customers or new indexing sites. We have, in the past, had to create entire new programs to allow bulk uploads to Amazon or Half – there is never a dull moment at Allusive Information.

Finally, we are always looking for new ideas from our book-selling colleagues. We are committed to making this “computer thing” work for all of us, not just for the ones with the money or technical savvy to bulldoze their way to a solution. So far we have created BookRouter, which we are very proud of as a tool for reducing the day-to-day drudgery of the online bookseller. We have set up a stable maintenance contract environment for all the Bookease DOS users still out there. And we are hard at work on a whole group of new capabilities or tools – centered around the new bookseller’s program called Biblioware, but consisting of a group of universally adaptable modules: the order management system, the assisted cataloguing system, the online search engine. All of these will follow the BookRouter model of being universal and adaptable to as many legacy systems as possible. It all sounds very complicated, but, hopefully, the complication is under the surface where the user never sees it. Take one, or take them all, the tools are created to make things just a little bit easier and more simple for the bookseller client.

With that, I think I have to say goodbye to you all. This essay ended up being a lot longer and a bit more technical than I thought it would be when I started. I am always ready to field any questions that anyone might have about what we are about, but for now, from Scott, BJ, Steve, our European friends and myself, I would like to wish all of you in the world of books heartfelt best wishes from the world of computers.

The Standard: The Journal of the Independent Online Booksellers Association

Check out the Independent Online Booksellers Association Website