Results: ISO 12083 Survey

Number of Responses: 11

Questions/Responses

1. Do you use ISO 12083 or based your DTD on 12083?

3 Respondents answered "No, I do not use ISO 12083

Other Responses included:

• I did use ISO 12083 to tag 2 French books see http://www.bibl.ulaval.ca/doelec/pul/
• I don't use yet SGML for practical purposes, but I discuss it (including ISO 12083) at my lecture on text processing standards. I am also involved in the evaluation of the Polish translation of ISO 12083.
• In the last three years we have worked extensively building and deploying implementations of several 12083 derivatives including Elsevier DTD 3.0 and 4.1 and the Blackwell DTDs. In addition, we've written a DTD based on 12083.
• Yes. ISO 12083 is used as written by the Library of Congress, Network Development and MARC Standards Office for several important technical documents. Use of this standard has expanded to another unit within LC that has found its relationship to other standards (AAP in particular) to be valuable, particularly in conversion of text using other markup to an ISO standard.
• No we do not use ISO 12083
• My office uses 12083 for technical documentation which is not extremely complex but not simple either. Our documents do not have features like mathematical equations, chemical formulas or pictures in them.
• We are not currently using ISO 12083 or any derived DTDs.
• No, but one of the standards defining organizations (SDO) whose documents we are converting to SGML (via a subcontractor) references the ISO 12083 math fragment in their DTD.
• We use 12083 for all of our journals now, including the math. We have had to substitute CALS tables this year due to customer requests. We made only a very few modifications to 12083, most notable in the reference section where we added finer granularity, removed title (we never put the title in a reference, costs too much money! Almost all physics journals /never/ include the title in a ref across all publishers BTW).

2. Have you changed ISO 12083 for your own use? What changes did you make?

• Very slight change at the beginning of my instances:
  <!ENTITY % m.bib "(no?, (%bib;)*)" (deleted title)
  <!ELEMENT row - O (tstub?, (cell*|sqcell*)) (added Sqcell for Panorama)
  <!ELEMENT CITEREF - - (#PCDATA|EMPH)* (added emph to citeref)
  Of course, the document character set is to changed, if it is to be used for Polish texts. The choice is between ISO 8859-2 or ISO 10646, I personally prefer the latter.
• No comments were provided. However clearly derivatives were created.
• We have not changed the DTD for ISO 12083 for our own needs except to correct some small errors that did not allow it to be parsed properly. We did, however, make us of the "<note>" and "<noteref>" tags in a way I do not believe they were intended to be used. We needed a way to markup hypertext links within and between documents. ISO 12083 does not appear to have an equivalent to the HTML "<a>" (anchor) tag which normally contains URLs that support inter and intra-document links. We would like to see a special tag in ISO 12083 that was defined to support links like the HTML "<a>" tag.
• NA
• We have elements in author group not allowed in the spec. This is not parameterized. No legal way for us to add these elements. Need more flexibility. We deviate on this and other points where flexibility is not allowed.
• We have not made any changes as we thought there was value in staying with the established tags so we could take advantage of the DTD and any software that might be developed based on the DTD -- without having to alter same.
• In marking up one of our documents, we had occasion to wish we had a tag that was available in a propriety IBM markup tag list. The element was "<xmp>" (example). It was designed to function much as the "<lit>" (literal text) tag except that its content model allowed the use of emphasis tags for bold, italics, underlining, etc. Examples in technical documentation often have special formatting requirements that are not strictly speaking related to structures. Like typical literal text, the creator wants to control exactly how the text is layed out, except that use of boldface, italics, and other emphasis is often desirable. If WG6 is making a wish list, we'd like to see a "<xmp>" tag added.
• At present the "<toc>" tag is an EMPTY element. It cannot contain other elements or parsed character data. We'd like to have the option to include marked up text within the "<toc>" tag in front matter. Not all systems allow on-the-fly generation of table of contents information. Since we use ISO12083 to produce printed output and our Web version of documents, we'd like to actually generate and store the table of contents as text in the document. This would require a change to the standard to specify a content model for "<toc" other than EMPTY.
• Supress all elements that are not using to make it easier for editorial staff.
• We add elements where we need new tags (<URL> tag for example. Figures needed to be developed. The spec leaves this to the authoring organization. New table element had to be added.
• We do not follow the philosophy of having automatically generated text. We archive the information, we must capture section numbers for example. So we changed models to have all numbers and all punction clearly within the data. Author list is an example where punctuation is critical.
• We wanted to be ISO compatible. But we needed different levels of granularity of elements, especially in front matter. We had to add these.
• We find the strict leveling of sections/subsections is not usable in the real world. Authors don't author this way.
• We don't like the parameter entities. Any change makes a lot of work. Our DTDs are nearly compliant but look very different than those in the standard.
• We don't like generated text. We archive.The DTD is not suitable for SGML publishing. So if we cannot print what use is the standard?

3. If you don't use ISO 12083, why not?

• NA
• There are many flaws in this standard, one of them consisting in the unprecise references to American documents. I mean first of all the reference to ICADD. I located the ICADD mailing list and tried to subscribe to it, but it appeared to be locked by the list owner. I found with AltaVista an interesting URL "SGML: ICADD in ISO 12083" http://ftp.sunet.se/pub/etext/ota/TEI/software/docs/SGMLinfo/ICADDiso.html but the URL appeared to be obsolete - there is no such a document at the server. Another example is the reference to MIL-M-28001 EREVIEW I was unable to trace down any information on this military standard.
• NA
• It is way too complicated, yet it is not flexible enough to represent the things I need to have in the journal I publish on the Internet
• We use AMS-LaTeX
• Until recently, we did not have the infrastructure in place to support SGML data. Now that we do, it is possible that we will purchase information products from vendors which will use some form of ISO 12083. We would consider using it for some locally-developed projects, particularly if the mathematics model were more fully developed, but there are no definite plans at this point.
• The DTDs we use were developed before I joined IHS, so I can't say for sure. I suspect a major reason for it was the lack of SGML editor support for it then (5-8 years ago).
• The SDO whose documents we are converting has deprecated use of the ISO 12083 mathematical markup (FORMULA and DFORMULA), because of a lack of (built-in) support for it in ArborText's SGML editor then (1 year ago).

4. Other Comments:

• I am very interested in the XML version of ISO 12083! In the meantime I would appreciate very much your help in obtaining detailed information on ICADD and EREVIEW.
• I'd like to say that in my use of ISO 12083, I have found it to be very useful and usable, even the way it is (i.e., given its current repertoire of tags and attributes). Despite its foundations on the early AAP and ANSI/NISO standards, I believe its documentation structure remains valid, even in today's world of hypertext documents and the Internet. The small number of things that are missing are far outweighed by the large number of features that are present and work well. Despite the somewhat negative tone of the introduction to the survey (the suggestion that the standard is "flawed" being on the top of my list), I think ISO 12083 is a valuable standard and deserves wider praise and implementation.
• We do not yet have experience or need in this area, but believe the standard could benefit from a header element with a content model like that in TEI (Text Encoding and Interchange) DTD. The TEI header elements have been discussed widely for their usefulness in harvesting metadata from machine-readable text. We believe we may need to convert between TEI and ISO 12083 for some documents. The lack of granular metadata elements in or before the front matter in ISO 12083 may discourage its use. We suspect it has kept some TEI supporters from even considering ISO 12083 (although it has not discouraged us up to now).
• I am under the impression that there were a number of big players (Kluwer?) who had a lot invested in ISO 12083. We are interested in considering ISO 12083 for encoding technical documentation and possibly bibliographies. We've already invested heavily in coding technical documentation. We are hoping other ISO 12083 are not ready to "throw in the towel" on the standard.
• On a very positive note, we've really like the ISO 12083 tag set. It has not failed us yet (in providing a variety of useful elements) and our technical documents are quite complex. Unlike TEI, ISO 12083 is more prescriptive in terms of document structure (sometimes requiring the ordering of elements different than our existing paper documents), but ISO 12083 has proven rich enough to allow us to encode all the structures we've encountered. And I think we haven't cheated! The occasional rigidity has made for very consistent documentation that converts easily to HTML. We like to be able to share that conversion success with other 12083 users too!
• Up to now our greatest concern has been the lack of support for the standard. Even the DTD is not available in machine-readable from a Web-accessible site. This would seem to be *essential* in today's Internet environment. There also seems to be a need to a users group and a healthy maintenance agency where questions about implementation of the standard can be addressed. IS there any hope that better support is in the offing? We'd like to share the news of our successes and have a place to turn when we get stuck. We hope WG6 can help us and other users. (We hope we're not the only ones now!)
• I'm not sure where the 12083 authority web site is? Does one exist?
• I am writing as a user of ISO 12083 to respond to the questions in the survey that you circulated. I was surprised at the negative tone of the introduction to the survey. I need to know if there is real ISO commitment to maintenance and propagation of the standard so I can make decisions about continuing to use it. Would I would be in a more stabile and better supported environment if I used a proprietary markup? That is what is important to me as a user, since the initial markup of continually updated and used documents is a major investment. I hope that the survey results will lead the working group to needed changes and a sound program for support and propagation.
• What's the status of XML compliant 12083? This is of enormous and compelling interest to me. We would like our on-line platform to begin delivering experimental/alpha versions of XML (including MathML) by the end of the year. We would like to see a few users actually embed material in their own web-pages from our journals using the XLink/Xpointer model from w3c. We figure that stuff will be in the browsers (particulary msie and mozilla/netscape) before we're ready anyway...
• I'd like to see the XML version allow for MathML in place of 12083 math. It's a simple thing to do with the DTD, politically I have no idea and standards-wise yikes! But people will just do it anyway as soon as MathML is working in a browser (actually it works in Amaya from w3c but it didn't quite hold up under the physics assault I put it to :)
• If you have 12083 XML in some organic state I would definately like to help complete it, I will have to do it all myself if no one has been working on it :) If no one has worked on it I'm willing to start in on it and maintain a web page and email group to work on it. This is critical to AIP /now/ for products coming up next year or at least by end of 1999 – F. Scott Johnson, AIP

Respondents:

• Guy Teasdale
  Bibliothèque de l'Université Laval, Sainte-Foy, Québec G1K 7P4
  TÉL: (418) 656-2131 - 2090 FAX: (418) 656-7897
  http://www.bibl.ulaval.ca/info/pagepers/teasdale/welcome.htm
• dr hab. Janusz S. Bien, Instytut Informatyki UW
  Janusz S. Bien, DrSc, Institute of Informatics, Warsaw Uniwersity
• Irina Golfman
  President
  Inera Incorporated
  815 Washington St. #3
  Newton, MA 02160
  617-969-3053 (phone)
  617-969-4911 (fax)
  irina@inera.com
• Randall Keigan Barry
  Senior MARC Standards Specialist
  U.S. Library of Congress
  Network Development and MARC Standards Office
  101 Independence Avenue, S.E
  Washington, DC 20540-4102 U.S.A.
  TEL: +1-202-707-5118
  FAX: +1-202-707-0115
  NET: rbar@loc.gov
• Sally H. McCallum,
  Chief, Network Development and
  MARC Standards Office, Library of Congress
  Washington, DC 20540 USA
  smcc@loc.gov (Fax: 1-202-707 0115) (Voice: 1-202-707 5119)
• Peter C. Gorman
  University of Wisconsin
  General Library System
  Automation Services
  pcgorman@facstaff.wisc.edu
  (608) 265-5291
• Kevin Rodgers
  kevinr@ihs.com
  Lead Software Engineer
  Information Handling Services
  Electronic Systems Development
  15 Inverness Way East, M/S A201
  Englewood CO 80112-5776 USA
  1+ (303) 397-2807[voice]/-2244[fax]
• F. Scott Johnson
  AIP
  scottj@aip.org>

Summary

Use of Standard

• It is difficult to tell the actual use of the standard from the survey responses. Those that use ISO 12083 seem to agree that the standard is quite useful, although they have suggestions for change.
• In addition, XML may have cause those that did not use the standard before to consider its use more seriously. Again changes must be made to address this audience.
• My sense is that LaTeX use is strong in the mathematics-based community. But putting "smart" data on the Web may motivate those users to make a shift to XML for delivery. Can ISO 12083 be compatible with both LaTeX and MathML?

Changes/Updates to the Standard

• More flexibility
• Simpler to use, simpler to get started, easier to "comply"
• Ability to add metadata
• XML links for Web presentation of data
• MATHML support
• XML version of DTD in standard
• Table support compatible with model provided by tools

Other

• Web site access to standard
• Help for users, news group or Web site where Q&A can be interchanged
• Tool support

(Return to TOC)