Balisage logo

SGML 30th Birthday Party — Toasts

All of the photographs on these pages were provided by the photographer for this use, and unless otherwise specified remain the property of the photographer.

At Balisage 2016, the 30th Birthday of SGML was celebrated with toasts, stories, cake, and many friendly reunions. The invitation to the party is available here: SGML 30th Birthday Party Invitation

SGML Birthday Cake
SGML Birthday Cake

Toasts from the Virtual Birthday Party

Eliot Kimber
SGML-involvement: I was an early user of SGML within IBM and then later worked directly with Charles on HyTime and revisions to the SGML standard in the early 90s. I was an SGML consultant from 1994 until SGML was supplanted by XML for all practical purposes.

SGML-memory: I first started using SGML within IBM as a punk kid who didn't even realize it was a standard. I created docs using the GML convention of a line of dashes as a visual separator for comments (".*-----------" in GML) and wondered why my SGML documents were randomly failing. After some experimentation I realized that if the number of dashes was a multiple of 4 then everything worked.

Discovering this I proceeded to post on the internal SGML forum, asking more or less literally "What idiot came up with this comment syntax?" only to have Charles Goldfarb answer that he was the idiot and point out that there was in fact a specification for SGML and that I ought to read it. Which I did, getting the SGML standard from the local IBM library.

That was the beginning of my relationship with Charles, which endures to this day (and I'm happy to report that I saw Charles just a few weeks ago and he is in good health and as feisty as every).

Sean McGrath
SGML-involvement: I have been working with SGML/XML/HTML since about 1990. Wrote 3 books on markup for Prentice Hall in the CFG Series. I was on the W3C invited experts group for XML.

SGML-memory: Dr David Abrahamson of TCD told us comp. sci. undergrads at Trinity College Dublin, about SGML back in 1986 when it was still a draft international standard. The notion of a dynamic parser for parsers, that would then proceed to parse documents against the parser just parsed, pretty much blew my mind!

I posted a letter (yes, a physical letter) to an address in Switzerland to get a copy of ArcSGML on a floppy disk by return post. I found comp.text.sgml on on usenet. Then I found James Clarks sgmls. Then I met Charles Goldfarb. Then I saw ESIS - and it all started to make sense.

Then HyTime blew my mind. Later DSSSL did it too :-)

Although I use XML and XHTML now much than I use SGML these days, I can honestly say that most of the key ideas I have build a career around in content management have come from the SGML eco-system.

Happy birthday SGML. You changed the world and amply rewarded the effort I expended in figuring you out.

Richard Light
SGML-involvement: I attempted to introduce the concept of markup to the cultural heritage community, most notably with CIMI and the National Museum of American Art. I was Treasurer of ISUG for a number of years.

SGML-memory: in the 1990s I worked with Steve Dietz at the National Museum of American Art on the development of a CD-ROM showcasing the collections. The source material was lovingly marked up in TEI SGML, using the state-of-the-art Author/Editor software. Of course, once the CD-ROM developers got their hands on the SGML data they proceeded to hack it using a home-grown 'parser', losing contextual information and introducing new errors.

Interestingly, much the same thing happened to me more recently with a U.K.-based museum project, where the data was presented as XML. So 'machine-processible' still means something different to web developers, compared to SGML/XML types. Maybe we'd be safer with binary SGML, which they can't hack. (I don't mean that.)

Beth Micksch
SGML-involvement: SGML was my life for a long time! Started in SGML at Datalogics on SGML projects and systems, SGML instructor for GCA; chairman for OASIS and IDEAlliance years ago.

SGML-memory: 30 years already! SGML or die, remember that saying? Hello to everyone! I miss all the SGML and XML conferences and the opportunities to catch up with everyone.

Bill Kasdorf
SGML-involvement: I've been involved with SGML and its descendant technologies pretty much my whole career. See my SGML Memory below.

SGML-memory: I first heard about _GML_ (aka Goldfarb/Mosher/Lorie) when Charles Goldfarb spoke at an early annual meeting of SSP, the Society for Scholarly Publishing. (I started attending when I was six.) I was immediately hooked. In those days, back when faxing was cool, it wasn't as easy as it is today to tune in to and stay on top of such things. I gradually found out about GenCode, and DCF, and the AAP Electronic Manuscript Project, and the MAJOUR DTD (thanks to Craig Van Dyke, then at Springer), and ISO 12083, and on and on. So how am/was I involved with SGML? It is in my professional DNA; in all the XML I deal with in my work today (and EPUB, and HTML5, and JSON, which I know may send some shudders up the spines of Balisagers, as they sometimes do to mine) I never forget where all this came from. This has shaped my now very long career fundamentally and utterly permanently. When people ask me when I'm going to retire I say "when I die." Not a joke. This stuff is too much fun. Why would I quit doing it?--Bill Kasdorf

Liam Quin
SGML-involvement: First used Author/Editor in 1987, then worked for SoftQuad for 7 years, then helped to kill it (or give it new life?) with XML.

SGML-memory: I got to visit potential SoftQuad customers with Yuri Rubinksy, and to hear his arguments for using SGML. We stayed up late into the night talking about possibilities. Yuri had insane travel schedules and had had the experience of rushing down the walkway to find the 'plane door closed, and knocking on it to get it open for him. Because he knew that if you knock on doors hard enough they open.

We worked together on HoT MeTaL, the HTML editor built out of Author/Editor, and on the SGML World Tour CD. He and I visited the NCSA Mosaic team to persuade them to include "th" elements to support accessible tables for Braille.

In a hallway at a 1984 HTML conference in Chicago Yuri and I were there with Stu Wiebel of OCLC (Joseph Hardin of NCSA may've been there too, it's been too long) and out of that came Dublin Core. The feeling that the Web needed help from library and metadata professionals, from information specialists, but didn't know it needed that help, lives on. We still have to knock on doors. Information design is still hard, as is thinking about all people and cultures and their needs, but Yuri and others at SoftQuad helped by knocking on those doors. Let's all keep knocking.

Bob Dreyfuss
SGML-involvement: Starting in 1997, worked with ASTM International converting its then 60,000 pages of standards into SGML and creating an SGML-first workflow.

SGML-memory: To all those who survived SGML conversions and never gave up when the going got tough and now reap the many benefits that it provided.--Bob Dreyfuss, ASTM International

Francis Cave
SGML-involvement: I first learnt about SGML at an event hosted by the International STM Group in Amsterdam in late 1982, where Charles Goldfarb, Bill Tunnicliffe and others were drumming up support for SGML to become an International Standard. With Joan Smith and others I was involved in organizing the first international Markup Conference in Oxford in 1984. I - along with Martin Bryan, John Trevitt and others - represented the UK at several ISO Working Group meetings and was a founding member of the International SGML Users' Group. I still have one publisher customer using an SGML-based system that I helped them to construct almost 20 years ago, and they have no desire to switch to anything else.

SGML-memory: ISO rules meant that the brand of English used in drafting the international version of the SGML standard mostly had to follow British English spelling conventions. So, my personal thanks go to Charles Goldfarb for putting up with us "limeys", who constantly "nit-picked" about English grammar and spelling.

Graham Every
SGML-involvement: In 1992, I came home from the first day of my SGML training course and said: "If this thing takes off I'll be really surprised. It's too complicated and too convoluted." Over the next 24 years I've built an extremely enjoyable and rewarding career in SGML and then XML, so I guess I have to admit I was wrong!

SGML-memory: I owe it all to Pam Gennusa, who first hired me as a junior programmer at DPSL. She taught me everything I know about SGML and -- equally as importantly -- how to teach SGML & XML to others. So my toast is to Pam to thank her for the incredible opportunities I've had as a result of her decision to hire me.

G. Ken Holman
SGML-involvement: Member of JTC 1/SC 18 and JTC 1/SC 34 since 1989, also serving for five years as committee Secretariat Manager.

SGML-memory: A fond memory of the SGML committee work comes from one of my very first international meetings, perhaps my first, where I revealed in an off-hand comment at the table my ignorance at an important distinction in the specification. From both sides of me, Lynn Price and Charles Goldfarb, *in unison*, corrected me with identical phrasing. Almost as if it had been practiced. It underscored for me that one cannot simplify concepts for end users to the point where the simplified labels end up being ambiguous with other important concepts. Others must have done this as well based on the reaction. If I didn't have the concepts right, how could I expect my customers to get them right? That experience and memory sticks with me to this day, many times getting me labeled as a pedant in my work and teaching, but being precise and pedantic will lead to fewer problems than being imprecise and ambiguous.

Toast: I would like to toast the standards leadership and tutelage of Jim Mason to steer the SGML committee and its members so well the way he did, amid tensions from all sides within and without, to bring it to where he did before handing over the reins and prods. SOOOOOWEEEEEEEE!!!

Bob DuCharme
SGML-involvement: I did DTD design and Omnimark work from 1994 - 2000 and wrote the book "SGML CD" about free SGML software for Prentice Hall as part of Charles Goldfarb's series. I attended and spoke at the big main conference each year going back to SGML 95.

SGML-memory: I'd like to toast Robin Cover. For many of us who played the role of SGML "experts" at our respective employers, the ability to quickly and easily look up just about any SGML thing that ever happened on Robin's "Cover Pages" played an important role in our maintaining the appearance of experts.

Michael Miller
SGML-involvement: Been working with SGML products since the beginning of time or all 30 years.

SGML-memory: I would like to toast all the pioneers and characters of SGML who made this an interesting experience over these past 30 years. At the top of that list would be Yuri Rubinsky.

Marla Banks
SGML-involvement: I've worked with SGML documents at the Library of Congress. Started with an experiment with Congressional Research Service, continued as technical support for the American Memory site and for Encoded Archival Description (EAD) finding aids.

SGML-memory: Here's a toast to SGML tools like SoftQuad's Panorama, the SGML viewer; Microstar's Near & Far Designer, for DTDs; WordPerfect for its attempt at SGML authoring; Framemaker+SGML. But here's a very special toast to OmniMark, the programming language for SGML (and XML)!

Martin Bryan
SGML-involvement: Member of BSI committee asked to review and approve the standard in 1987. Wrote the first book devoted to teaching the basics of SGML.

SGML-memory: SGML introduced me to people all round the world, as it today introduces people all around the world to each other. We started off working on something to help produce documentation, little realising that we were to help create a world where you checked things first on a computer rather than in a book. But in doing this we have made it easier than ever for people to record their thoughts for posterity, often in the form of books (even if many of these are electronic!) Where would the internet be without SGML's sibling XML?

Today I see many of the standards that I worked with over two decades on everything my children use. Having reviewed MPEG2 and MPEG3 I now find my camera and phone used MPEG4. My wife's camera sends photos direct to the kids by email, again thanks to the standards we worked so hard to create. We can be proud of our pioneering work. Well done everyone.

Robert Wheeler
SGML-involvement: STM publishing work at AIP, then migrating to XML.

SGML-memory: Possible Toasts:
- SGML, I hardly knew you, but now stand on your shoulders. Cheers!
- SGML, you knew it --nothing ever ends-- Cheers!

Jean Paoli
SGML-involvement: I was the technical director of the team building the Grif SGML WYSIWYG Editor, in a startup of INRIA. I started working on SGML around 1987, participated in so many SGML conferences and met, learned and worked with so many wonderful people in the SGML Community. I joined Microsoft in 1996 and was part of the original W3C effort that created XML.

SGML-memory: I remember when using text to represent data was a novelty, when sending text data structures to enable interoperability between heterogeneous systems was very rare, when creating documents with semantic information was considered difficult or impossible: today, those concepts, originally championed by SGML, are at the base of the openness of the Internet and core to its future development, including Big Data, Machine Learning and the problems of structured and unstructured data and documents.

The Internet is standing on the shoulders of a lot of giants, and many of those giants are here today, the giants who created SGML and are celebrating its 30th anniversary.

Josh Lubell
SGML-involvement: I developed an SGML application to publish the first version of ISO 10303-210, a standard for representation and exchange of digital printed circuit board design data.

SGML-memory: SGML text processing was the focus of my early years at NIST, which led to a long career where I have found all sorts of interesting ways to use markup technologies to solve systems integration problems in engineering, manufacturing, and cybersecurity. Happy 30th birthday!