Loading ...
Sorry, an error occurred while loading the content.

[sml-dev] Re: SML reactions from W3C members (was SML < XML ?)

Expand Messages
  • Michael Champion
    ... From: Gavin Thomas Nicol To: Sent: Wednesday, December 01, 1999 6:36 PM Subject: [sml-dev] Re: SML
    Message 1 of 12 , Dec 1, 1999
    View Source
    • 0 Attachment
      ----- Original Message -----
      From: Gavin Thomas Nicol <gtn@...>
      To: <sml-dev@egroups.com>
      Sent: Wednesday, December 01, 1999 6:36 PM
      Subject: [sml-dev] Re: SML < XML ?


      > > DOM would be more interesting - but my point was you ought
      > > to be able to demonstrate the SML 'simplicity' by trimming
      > > the APIs.
      >
      > Probably the single most complicated part of the DOM is
      > that the tree is "live". That has nothing to do with
      > the use of XML.

      You should have come to this week's working group meeting, Gavin.

      6 -- count 'em -- SIX hours of discussion on how to resolve the mess that
      is created when parsed entities meet namespaces. And they probably took
      another six hours off my life because of the chocolate consumed to fight the
      depression and caffeine consumed to fight the urge to sleep. And we ended
      up deciding to essentially put up a sign saying "There Be Dragons There"
      because
      the problems are too complex to solve without mandating a lot of machinery
      that would serve little purpose but to present a pretty face on a very, very
      ugly mess.

      I had a few private chats about SML at the meeting which were pretty
      interesting. First, there appears to be zero, zilch, nada support for the
      idea of
      SML. Don't even THINK about approaching the W3C for their blessing,
      Don. It would be D.O.A.
      XML is *so* much simpler than SGML that the general opinion is that
      it is simple enough. Also, there *are* a lot of people using parsed
      entities, CDATA
      sections, comments, and processing instructions that didn't chime in on
      XML-DEV
      when we discussed all this stuff.

      Also, the interest the the XML features that were dropped from SML
      transcends just the documentation-oriented people; I talked to an e-commerce
      developer who couldn't live without parsed entities!

      On the other hand, there's a very widespread agreement that the minimalist
      philosophy behind SML is appropriate and needs to be more widespread.
      There's a lot of frustration that the various XML-related groups can't get
      the minimal specs out because they're spending their time "gold plating"
      them
      with potentially desirable but complex and controversial features.

      I personally am intrigued by the idea of SML because it *mandates* the
      minimalist
      philosophy (Occam's Razor, as Robert LaQuey emphasized in his article).
      Even
      more important is the notion of a clean hierarchy of functionality layered
      on top of
      a minimal core. I honestly don't know if re-building the specs up in
      layers will
      work better than a pragmatic set of ad-hoc recommendations about where the
      dragons are in the XML world. On the other hand, the question is too
      important
      to decide by default, so perhaps parallel efforts to promote the idea of
      Occam's
      Razor in the W3C world and to actually go out and re-architect the specs in
      the SML world is the best way of proceeding.
    • Don Park
      Gosh, I am shocked that good old handy weasel wasn t enough and the dreadful dragon had to be applied. ... I wasn t planning to. If SDOM is useful at all,
      Message 2 of 12 , Dec 1, 1999
      View Source
      • 0 Attachment
        Gosh, I am shocked that good old handy 'weasel' wasn't
        enough and the dreadful 'dragon' had to be applied.

        >I had a few private chats about SML at the meeting which
        >were pretty interesting. First, there appears to be zero,
        >zilch, nada support for the idea of SML. Don't even THINK
        >about approaching the W3C for their blessing, Don. It
        >would be D.O.A.

        I wasn't planning to. If SDOM is useful at all, it will be
        used without W3C's blessing. SAX turned out all right and
        I found it amusing that W3C is working on SAX for CSS.

        [snip]
        >Razor in the W3C world and to actually go out and re-architect
        >the specs in the SML world is the best way of proceeding.

        Frankly, I think of it as 'dolphin-approach': riding ahead of a
        large ship to surf the waves created by the ship. We can create
        better specs by dropping or adding features in the form of 'usage'
        recommendations.

        Best,

        Don Park - mailto:donpark@...
        Docuverse - http://www.docuverse.com
      • Robert La Quey
        At 08:45 PM 12/1/99 -0500, you wrote: Thanks Michael for the insightful report. ... Not surprising. The W3 can hardly be expected to state the unfortunate yet
        Message 3 of 12 , Dec 2, 1999
        View Source
        • 0 Attachment
          At 08:45 PM 12/1/99 -0500, you wrote:

          Thanks Michael for the insightful report.

          >I had a few private chats about SML at the meeting which were pretty
          >interesting. First, there appears to be zero, zilch, nada support for the
          >idea of SML. Don't even THINK about approaching the W3C for their blessing,
          >Don. It would be D.O.A.

          Not surprising. The W3 can hardly be expected to state the unfortunate
          yet obvious fact: "We are exerting great effort and producing a pile of
          pretentious complexoturds which we expect you to just shut up and eat."

          We shall have to have a better strategy for fixing this mess than
          submitting "Yet Another Standard" to them. I have an idea in mind but
          I want to sleep on it before laying it before this group. Meanwhile
          I continue to chant:

          SML is "All the XML you really need."


          Bob La Quey
        • Gavin Thomas Nicol
          ... to sleep. ... mandating a lot ... Well, we probably spent about 100 hours on the entity model in general, and probably another 200 on liveness .
          Message 4 of 12 , Dec 2, 1999
          View Source
          • 0 Attachment
            > You should have come to this week's working group meeting, Gavin.
            >
            > 6 -- count 'em -- SIX hours of discussion on how to resolve
            > the mess that is created when parsed entities meet namespaces. And they
            > probably took another six hours off my life because of the chocolate
            > consumed to fight the depression and caffeine consumed to fight the urge
            to sleep.
            > And we ended up deciding to essentially put up a sign saying "There Be
            > Dragons There" because the problems are too complex to solve without
            mandating a lot
            > of machinery that would serve little purpose but to present a pretty face
            > on a very, very ugly mess.

            Well, we probably spent about 100 hours on the entity model in general, and
            probably another 200 on "liveness". personally, I think namespaces are
            somewhat misguided anyway... I am *not* opposed to SML if it is a subset
            (quite in favor of it, because I teach people a basic subset anyway).

            I'd *really* like to see SDOM: it would be useful even in the context of
            normal XML.
          • Michael Champion
            ... From: Gavin Thomas Nicol To: Sent: Thursday, December 02, 1999 9:45 AM Subject: [sml-dev] Re: SML reactions from W3C
            Message 5 of 12 , Dec 2, 1999
            View Source
            • 0 Attachment
              ----- Original Message -----
              From: Gavin Thomas Nicol <gtn@...>
              To: <sml-dev@egroups.com>
              Sent: Thursday, December 02, 1999 9:45 AM
              Subject: [sml-dev] Re: SML reactions from W3C members (was SML < XML ?)


              >
              > I'd *really* like to see SDOM: it would be useful even in the context of
              > normal XML.
              >

              Right! That was my original interest in SML -- the DOM WG wasn't interested
              in an S-DOM without an S-XML to guide its features, and supposedly Canonical
              XML would specify such a thing. That's not what Canonical XML turned out to
              be. S-DOM is what I'm primarily interested in, and since the W3C is not
              interested in blessing such a subset, maybe you folks will be.

              BTW, I haven't read Gavin's data model for a long time, but he was
              DEFINITELY RIGHT in the DOM development -- we should have defined the
              underlying data model before attempting an API. Very many of the DOM's
              flaws are easily traced to that.

              Speaking of which, I have a vague notion that SML *is* canonical SML, i.e.,
              there should be one and only one way of representing any item in the
              underlying data model in SML syntax. Is that naive?
            • Oren Ben-Kiki
              ... i.e., ... If you mean a single possible stream of bytes, then yes :-) I expect that there will be multiple possible variants, differing in: - Encoding
              Message 6 of 12 , Dec 2, 1999
              View Source
              • 0 Attachment
                Michael Champion <Mike.Champion@...> wrote:
                > Speaking of which, I have a vague notion that SML *is* canonical SML,
                i.e.,
                > there should be one and only one way of representing any item in the
                > underlying data model in SML syntax. Is that naive?

                If you mean a single possible stream of bytes, then yes :-)

                I expect that there will be multiple possible variants, differing in:

                - Encoding (UTF-8 or UTF-16).
                - White space (for non-leaf elements).

                So far these aren't serious issues. The next one is:

                - If attributes end up being a "shorthand" for elements, only the
                attribute-less form would be "canonical". Score another point against this
                notion :-)

                Another potential problem, which I don't consider to be serious, but others
                might:

                - Namespaces (choice of prefix).

                There may be other problems...

                The goal of keeping the number of possible variants to a minimum is a worthy
                one. We could use it as a guide in considering features.

                Have fun,

                Oren Ben-Kiki
              • James Tauber
                ... interested ... Canonical ... to ... No, Canonical XML was never supposed to be a simplified XML. Canonical XML and the simplified profile of XML were
                Message 7 of 12 , Dec 2, 1999
                View Source
                • 0 Attachment
                  > Right! That was my original interest in SML -- the DOM WG wasn't
                  interested
                  > in an S-DOM without an S-XML to guide its features, and supposedly
                  Canonical
                  > XML would specify such a thing. That's not what Canonical XML turned out
                  to
                  > be.

                  No, Canonical XML was never supposed to be a simplified XML. Canonical XML
                  and the simplified profile of XML were separate work items.

                  > Speaking of which, I have a vague notion that SML *is* canonical SML,
                  i.e.,
                  > there should be one and only one way of representing any item in the
                  > underlying data model in SML syntax. Is that naive?

                  Would it be any trouble to allow for whitespace to be repeatable?

                  James
                • Gavin Thomas Nicol
                  ... Canonical ... since the ... I would be *very* interested in pursuing this.
                  Message 8 of 12 , Dec 2, 1999
                  View Source
                  • 0 Attachment
                    > Right! That was my original interest in SML -- the DOM WG
                    > wasn't interested in an S-DOM without an S-XML to guide its features, and
                    > supposedly Canonical XML would specify such a thing. That's not what
                    Canonical
                    > XML turned out to be. S-DOM is what I'm primarily interested in, and
                    since the
                    > W3C is not interested in blessing such a subset, maybe you folks will be.

                    I would be *very* interested in pursuing this.
                  • Joe Lapp
                    ... Let s do it, but not yet. Let s make sense of the object model first. I keep getting pulled into meetings, so I haven t had time to give it much thought
                    Message 9 of 12 , Dec 2, 1999
                    View Source
                    • 0 Attachment
                      At 05:32 PM 12/2/99 -0500, Gavin Thomas Nicol wrote:
                      >>[...] S-DOM is what I'm primarily interested in, and
                      >since the
                      >> W3C is not interested in blessing such a subset, maybe you folks will be.
                      >
                      >I would be *very* interested in pursuing this.

                      Let's do it, but not yet. Let's make sense of the object model first. I
                      keep getting pulled into meetings, so I haven't had time to give it much
                      thought yet.
                      --
                      Joe Lapp (Looking for some good people to
                      Senior Engineer help create XML technologies that
                      http://www.webMethods.com connect businesses to businesses
                      jlapp@... over the web.)
                    • Michael Champion
                      ... From: James Tauber To: Sent: Thursday, December 02, 1999 11:34 AM Subject: [sml-dev] Re: SML reactions from W3C
                      Message 10 of 12 , Dec 2, 1999
                      View Source
                      • 0 Attachment
                        ----- Original Message -----
                        From: James Tauber <jtauber@...>
                        To: <sml-dev@egroups.com>
                        Sent: Thursday, December 02, 1999 11:34 AM
                        Subject: [sml-dev] Re: SML reactions from W3C members (was SML < XML ?)


                        > No, Canonical XML was never supposed to be a simplified XML. Canonical XML
                        > and the simplified profile of XML were separate work items.


                        Thanks for clarifying this. I should have made it clear that it was our
                        MISUNDERSTANDING of what Canonical XML was always about that was the root of
                        this confusion, not that the intent of Canonical XML changed.

                        Is there any W3C activity to define a "simplified profile of XML" underway?
                      • Joe Lapp
                        ... I exchanged a few emails with Tim Bray about this several weeks ago. The XML Syntax group attempted to initiate an XML Profile effort, but failed for
                        Message 11 of 12 , Dec 2, 1999
                        View Source
                        • 0 Attachment
                          At 12:33 AM 12/3/1999 -0500, Michael Champion wrote:
                          >Is there any W3C activity to define a "simplified profile of XML" underway?

                          I exchanged a few emails with Tim Bray about this several weeks ago. The XML Syntax group attempted to initiate an XML Profile effort, but "failed" for not having agreed upon the requirements. The effort was to define XML subsets for low resource computing and other specialized computing environments. The effort is now defunct.

                          I see such a diverse set of requirements in this group that SML may be in jeopardy. I suspect we'll find out based on whether we can agree on a data model.

                          I suggest that somebody else use Gavin's syntax to get the model conversation going. I want to see what this group is thinking before I chime in.

                          --
                          Joe Lapp (Looking for some good people to help design
                          Senior Engineer and build the Internet's business-to-business
                          webMethods, Inc. XML infrastructure. We are 100% Java.)
                          jlapp@... http://www.webMethods.com
                        • James Tauber
                          ... underway? We made a start on it but found that the activity was ill-defined. The parts of XML that can be thrown out versus those that must stay in wildly
                          Message 12 of 12 , Dec 2, 1999
                          View Source
                          • 0 Attachment
                            > Is there any W3C activity to define a "simplified profile of XML"
                            underway?

                            We made a start on it but found that the activity was ill-defined. The parts
                            of XML that can be thrown out versus those that must stay in wildly differs
                            depending on application.

                            James
                          Your message has been successfully submitted and would be delivered to recipients shortly.