Re: [xml-dev] RE: The XML abstraction leaks

[David Carlisle <d.p.carlisle@gmail.com>]

On 11 August 2016 at 21:50, Costello, Roger L. <costello@mitre.org> wrote:
> Thank you David and Michael. I would like to follow up please.
>
>
>
> I really am fascinated by the topic of abstractions and leaky abstractions.
> I hope that lots of people will pitch in with their thoughts and opinions on
> this topic.
>
>
>
> 1. What is “the XML abstraction”? I said that it is this:
>
>
>
> An XML document contains data and the data
>                is surrounded (delimited) by markers. More
>                concretely, an XML document contains pairs
>                of start-tags and end-tags, sandwiched between
>                them is character data and possibly other
>                start-tag, end-tag pairs.
>
>
>
> Do you agree that that is the XML abstraction?

No not at all: XML's tag syntax is, as you say above, a _concrete_
syntax grammar, it is the opposite of an abstraction.

>  If not, would you provide
> your definition of “the XML abstraction” please?

XML is a concrete syntax not an abstraction at all, abstract data
models somewhat related to XML would be XDM or the Xpath model or the
DOM, none of which are defined by the XMl rec.

>
>
>
> 2. What is a “leaky abstraction”? From reading  Joel Spolsky’s article as
> well as Peter Seibel’s description [1] I have arrived at this definition:

>
>
>
>                The XML abstraction leaks when users must
>                understand to some extent how software that
>                implements the XML abstraction—the XML
>                processor—works internally.
>
>
>
> Do you agree with that?

No, it would need to be an abstraction, and leak in some way to be a
leaky abstraction.
Also saying that something leaks just because users need to read the
instructions on how to use it is a strange use of the term.

> If not, would you provide your definition of what it
> means for the XML abstraction to leak please?
>
>
>
> 3. Would you identify a leak in the XML abstraction please?
>
>
>
> /Roger
>
>
>
> [1] Peter Seibel has a wonderful discussion of leaky abstractions in his
> book, Practical Common Lisp. He gives an awesome example of a user-defined
> function (abstraction) that leaks.
>
>

David