From: Tim Bray \ Newsgroups: comp.text.sgml Subject: Yuri Rubinsky Date: Fri, 26 Jan 1996 23:40:49 -0800 Organization: Open Text Corporation Message-ID: <3109D701.4030@opentext.com> I don't understand how a person can found a software company, tough it out through the bad days, lead it through the vale of Internet Fairy Dust, turn a silly idea into a PC Mag Editors' Choice, simultaneously be a pusher, mover, and shaker in the WWW and SGML worlds, get a novel published, wrestle with ISO, and never, so far as I know, cut a throat, stab a back, or crush an ego. That it is possible to be successful and honourable at the same time is something we should take seriously to heart; in these days it seems too little respected, or even expected. That alone would be a monument; maybe the only important one. But there are lots of others; a substantial part of the community, the technology, and the business. Still, though, I think it would be nice if, the next time we need to name a protocol or an encoding or a language or something, we think of Yuri. He and I walked about 40 blocks home across San Fran during the last Seybold, late at night, talking excitedly about some stuff we want to do on the Web, that might yet turn into another monument. You know those hills, neither of us were exactly slim nor fit, and we laughed at our own heavy breathing as we slogged up, over and down. Tough to think that we won't be doing that again. Tim Bray From: Eryq \ Newsgroups: comp.text.sgml Subject: NEWBIE Q: is this DTD wrong? Date: 27 Jan 1996 04:48:12 GMT Organization: Hughes STX Corporation Message-ID: <4ecaqc$5gf@Jupiter.mcs.net> Folks - I just tried my hand at writing my first pass at a DTD, which I've pumped though nsgmls 1.0.1 for Linux. The DTD will (ultimately) be for slide presentations... my problem is, that in the sample document (below), nsgmls complains that the \ within a \<slide> is in the wrong place if there is any whitespace between the \<slide> and the \<title> tags. I thought this was odd, since it doesn't complain about whitespace between \<head> and \<title>, for instance. BTW, I've tried coming up with a unique tag, like \<stitle>, for the slide title. Same problem. Here's the sample document: \<!DOCTYPE talk SYSTEM "/home/eryq/src/SGML/talk.dtd"> \<talk> \<head> \<title>This is \<em>my\</em> talk\ \ \ \Introduction\ \ \What is life?\ Philosophical junk... \ \What can I do with it?\ Practical junk... \ And here's the DTD: \ \ \ \ \ \ \ \ \ \ \ \ \ \ TIA for any help you can given me, -- ____ __ / __/__________/_/ Eryq (eryq@rhine.gsfc.nasa.gov) / __/ _/ / / , / Hughes STX Corporation, NASA/Goddard Space Flight Cntr. /___/_/ \\ /\\ /___ /_/ /_____/ http://selsvr.stx.com/~eryq/ From: jorn@MCS.COM (Jorn Barger) Newsgroups: comp.text.sgml,alt.hypertext Subject: Re: Hyperlinks (was: SGML and the US DoD) Date: 26 Jan 1996 22:42:29 -0600 Organization: The Responsible Party Message-ID: <4ecafl$cv6@Mars.mcs.com> References: \ <9601252333.AA08633@fly.hiwaay.net> <4e9e8k$8bq@Mars.mcs.com> <4ebbvp$mfa@crl2.crl.com> In article <4ebbvp$mfa@crl2.crl.com>, Joe English \ wrote: >>- Any site with a traditional table of contents is a waste. ToCs should >> offer concise *summaries*, not just headings. [...] >Useful advice... but I think it's an overgeneralization. There >are many cases where a traditional TOC is exactly the right >navigational aid. If the document is something that's likely >to be revisited many times (like a reference manual, for >example), a concise TOC is the best way to find exactly what >you're looking for. Even for first-time readers, summaries may be >unnecessary and sometimes annoying if the section headings are >descriptive enough. I can accept that this *might* be true... but can you give an example? My gut reaction now, *whenever* I see a column of links each <40 chars wide, is: this person is wasting my time. Forty chars just aren't enough to detail a link. Keeping the summaries concise enough so that the whole ToC fits on a single screen *is* nice for return visitors. But they'd probably also like #-links to the subtopics in each 'chapter'... And, in fact, they probably won't be using all the chapters, so they'd do better to bookmark just the ones they really use! >>- The main purpose of links isn't to enable exploration, it's to allow >> you to decide what you can safely *skip*. >Also good advice, but again I think an overgeneralization. >*One* purpose of hyperlinks is to allow readers to decide what they can >skip. I've discovered too much useful information on the Web >by wandering around to agree that exploration is less important >than pruning. I accept this, but I think the other lesson is more urgent to emphasize now. > Plus, I've waded through *way* too much junk to >agree that the pruning is particularly effective. We can learn to get better and better at it... surely? [other observations-- entirely agreeable-- trimmed] j -==--- . hypertext theory : artificial intelligence : finnegans wake . _+m"m+_"+_ lynx http://www.mcs.net/~jorn/ ! Jp Jp qh qh ftp://ftp.mcs.net/mcsnet.users/jorn/ O O O O news:alt.music.category-freak Yb Yb dY dY ...do you ever feel your mind has started to erode? "Y_ "Y5m2Y" " no. From: Gary Merrill Newsgroups: comp.text.sgml Subject: Re: Authoring Tool Comparison Methodology Date: 27 Jan 1996 01:10:37 GMT Organization: Interpath -- Providing Internet access to North Carolina Message-ID: <4ebu2d$k51@redstone.interpath.net> References: <4dh58f$nbr@blackice.winternet.com> > gray@winternet.com (William Gray) writes: > This is long, but I wanted to provide enough background. Short version: > We are required to develop a business case that recommends that: > * We all move to Interleaf > * We all move to Interleaf SGML > * We all move to Adept (likely to cause mutiny among the Interleaf diehards) > * We adopt Adept for the native side of the shop, leave Interleaf in > place for the SGML-stupid side of the shop, and run both tools > simultaneously. > As a matter of curiosity, why is the list of possibilities so restricted? Why, for example, is FrameBuilder so conspicuouslyu absent? In addition, I should think that one of the primary criteria should be the *productivity* of the writers (which in general turns out to mean "ease of use") for a given tool. These things are *very* difficult to evaluate. In fact, I think the only successful way of getting anything close to a true comparative evaluation is to take a sample document that is not currently in any of the competing representations and set a team to producing that document (all the way to the production version) with each of the competing tools. Only then will you really have any idea of the "gotchas" that are lurking beyond hype and appearance. It is only when you really have to *use* such a tool to create a *real* document that all good and bad points are exposed. This also exposes problems with the "maturity" of the product and with the support that is provided. Well worth the time and effort. ---------- Gary Merrill From: David Megginson \ Newsgroups: comp.text.sgml,alt.hypertext Subject: Re: Hyperlinks (was: SGML and the US DoD) Date: 27 Jan 1996 06:17:01 -0500 Organization: University of Ottawa Sender: david@baeda.english.uottawa.ca Message-ID: \ References: \ <9601252333.AA08633@fly.hiwaay.net> <4e9e8k$8bq@Mars.mcs.com> <4ebbvp$mfa@crl2.crl.com> <4ecafl$cv6@Mars.mcs.com> Reply-To: dmeggins@aix1.uottawa.ca In-reply-to: jorn@MCS.COM's message of 26 Jan 1996 22:42:29 -0600 In article <4ecafl$cv6@Mars.mcs.com> jorn@MCS.COM (Jorn Barger) writes: > [...] There >are many cases where a traditional TOC is exactly the right >navigational aid [...] I can accept that this *might* be true... but can you give an example? ====================8<====================8<==================== \ \ \ \Regional Listings\ \ \ \ \

Dining Guide: Canadian Restaurant Listings by Region\

\

Table of Contents\

\

Restaurant listings are available for the following regions:\

\ \
\ \ ====================8<====================8<==================== (Providing an explanation of each region would simply annoy Canadian readers, at least, and would keep the whole list from fitting on a single screen.) My gut reaction now, *whenever* I see a column of links each <40 chars wide, is: this person is wasting my time. Forty chars just aren't enough to detail a link. That depends on the nature of the readers. I teach writing, and the first thing I tell my students is that there is no such thing as correct grammar or a correct writing style -- instead, you use the method which is most appropriate to your audience (in an academic essay I would have written "A writer should use the method which is most appropriate to...", but the net is more informal). There is no set of rules formulated which can take away from the author (or document designer) the obligation to think about the document and its audience. Keeping the summaries concise enough so that the whole ToC fits on a single screen *is* nice for return visitors. But they'd probably also like #-links to the subtopics in each 'chapter'... And, in fact, they probably won't be using all the chapters, so they'd do better to bookmark just the ones they really use! This is an oversimplification. A first-time user might have a great deal of expertise in the field and might be able to find information quickly based only on succint chapter titles; on the other hand, an experienced user might need to consult the table of contents frequently to find information in different sections. Of course, by now most of this little to do with hypertext theory and almost nothing to do with SGML -- it's more along the lines of "Tips and Tricks for Web Document Designers" -- so I beg the indulgence of the group. David -- David Megginson Department of English, University of Ottawa, dmeggins@aix1.uottawa.ca Ottawa, Ontario, CANADA K1N 6N5 ak117@freenet.carleton.ca Phone: (613) 562-5800 ext.1203 WWW: http://www.uottawa.ca/~dmeggins FAX: (613) 562-5990
From: mtbryan@sgml.u-net.com (M Bryan) Newsgroups: comp.text.sgml Subject: The Rubinsky Prize Date: Sat, 27 Jan 1996 14:19:00 GMT Organization: The SGML Centre Message-ID: <4edc9p$n6q@nuntius.u-net.net> Reply-To: mtbryan@sgml.u-net.com The news of Yuri Rubinsky's unexpected death has greatly saddened those of us who had the privilege to know him. My memory of him will always be of his tireless energy in promoting the role of SGML as a tool for the benefit of mankind, especially its less fortunate members. He saw the advantages that SGML had to offer for the disabled long before the rest of us, and clearly promoted the idea of preparing information in a form that would be reusable in the widest possible number of environments. This work has already borne many fruits and will continue to be fruitful for decades to come. Perhaps the best tribute we could pay to Yuri's memory would be to set up an annual prize for the group that has done most during the past year to develop SGML-based services for the underpriviledged. I would suggest that the SGML '96 would be the ideal time and place to present such a Rubinsky Award, which The SGML Centre would be proud to sponsor. Martin Bryan E-mail: mtbryan@sgml.u-net.com The SGML Centre Churchdown, UK Phone: +44 1452 714029 --- Martin Bryan, The SGML Centre, 29 Oldbury Orchard Churchdown, GLos GL3 2PU, U.K. Phone/Fax: +44 1452 714029 From: jenglish@crl.com (Joe English) Newsgroups: comp.text.sgml Subject: Re: NEWBIE Q: is this DTD wrong? Date: 27 Jan 1996 11:08:41 -0800 Organization: Tagheads Message-ID: <4edt7p$s9s@crl12.crl.com> References: <4ecaqc$5gf@Jupiter.mcs.net> Eryq \ wrote: > >I just tried my hand at writing my first pass at a DTD [...] >nsgmls complains that the \ within a \<slide> is in the wrong place >if there is any whitespace between the \<slide> and the \<title> tags. >I thought this was odd, since it doesn't complain about whitespace >between \<head> and \<title>, for instance. > [...] > \<!ENTITY % textitem "#PCDATA | %font | %phrase "> > \<!ENTITY % text "(%textitem)+"> > [...] > \<!ELEMENT slide - O (title, %text)> You've run into the "mixed content" problem. First an explanation: An element is said to have "mixed content" if it contains #PCDATA anywhere in its content model, and "element content" otherwise. The parser uses this distinction to decide how to treat whitespace: In element content, whitespace is ignored, and in mixed content it's treated as data. In your DTD, the SLIDE element has mixed content, so in: \<slide> \<title>...\ the space characters after the \ start-tag are parsed as a #PCDATA content token -- which is not allowed at that point in the model group. One way to fix it is to define a container element for the slide content: \ \ Another approach is to define one or more "block-level" container elements, and require that all character data appear inside those throughout the DTD: \ \ \ \ \ \ \ If you ever plan to add block-level constructs like bulleted lists, I'd definitely recommend the second approach, since it will save you a lot of hassle in the long run. (HTML 3 is still plagued with mixed-content problems...) As a general rule, if an element allows #PCDATA anywhere in its content model, it should allow it *everywhere* to avoid mixed content parsing problems. It's best to use a single repeatable OR group for any element which allows #PCDATA. --Joe English jenglish@crl.com P.S. For a first pass attempt your DTD looks pretty good. From: "Alva T. Stone" \ Newsgroups: comp.text.sgml Subject: Re: HTML is not a subset of SGML? Date: 27 Jan 1996 20:32:48 GMT Organization: FSU Law Library, Tallahassee, FL 32306-1043 USA Message-ID: <4ee25g$h0@news.fsu.edu> References: <4e94i5$lip@news.fsu.edu> To: atstone@law.fsu.edu Mea culpa, mea culpa, mea maxima culpa! My apologies to Mr. Eric van Herwijnen. The statement that HTML is technically not a subset of SGML was *not* in his second ed. of _Practical SGML_, after all. I had spent so much time reading and trying to understand his manual about 4 weeks ago, that I'd forgotten having also read, about the same time, some very basic sources of information about the W3 and HTML. What I was remembering actually had been written in version 6.1 of Kevin Hughes' _Entering the World-Wide Web: A guide to Cyberspace_ (Palo Alto, CA: Enterprise Integration Technologies, c1994). He says, "The standard language the Web uses for creating and recognizing hypermedia documents is the \ Hypertext Markup Language (HTML) \. It is loosely related to, but technically not a subset of, the Standard Generalized Markup Language (SGML), a method of representing document formatting languages." [bracketed information mine] I should also point out that the opinions stated in the paragraph directly above this one are solely those of Mr. Hughes and do not necessarily represent the views of Enterprise Integration Technologies. :-) Lastly, I want to express in public my thanks to Ray Dassen, C.M. Sperberg-McQueen, and David Durand for taking the time to clarify for me (a non-programmer) the distinctions between HTML and SGML. You guys are great! --Alva From: "Pat Benjamin" \ Newsgroups: alt.jobs,biz.jobs.offered,comp.jobs,comp.jobs.offered,comp.text.sgml,misc.jobs Subject: SGML Systems Integrator Date: Sat, 27 Jan 96 14:58:52 -0400 Organization: The Prescott Group Message-ID: <3031858618.10.p01292@psilink.com> Responsibilities: - Lead Engineer - Software/systems integration - Information processing products Experience: - BSCS or Equivalent - Integration, programming and scripting skills - Familiar with Arbortext, Avalanche, SoftQuad, FrameMaker, Omnimark, other structured editor/desktop publishing programs Personality Traits: - Self starter, high initiative - Prefers corporate environment - Team player - Excellent communication skills Location: - East Coast - Fee paid by hiring company Salary: Commensurate with skills Contact: Pat Benjamin / Joy Norswortyh (609) 354-7651 Voice (908) 819-7600 Voice or Fax resume to: (609) 354-8274 FAX (908) 985-0704 FAX All correspondence held in strictest confidence!