On Thu, Aug 09, 2001 at 11:34:44AM +0200, Lars Gullik Bjønnes wrote:
[snip!]
>
> and follow an implicit DTD.
That was my point. Any text-based file format follows an implicit DTD.
> John Weiss <[EMAIL PROTECTED]> writes:
>
> | We can do both design and DTD simultaneously. When we need a new cell
> | or block or attribute, we'll end up discussing it, won't we? Well,
> | there are four ways to have that discussion:
> |
> | - In English (the imprecise Lingua Franca of the developer's list)
> | - In yacc-format
> | - In an SGML-DTD-like format.
> | - In an XSL-like format.
>
> - C++ code.
Ah, but C++ was designed for coding algorithms, not for defining DTDs.
Besides which, if someone wants to ask if the part of the LyX file
format that handles Insets should have a "width" parameter, they're
not going to do so by sending everyone the code for parsing it.
They're going to ask, "Should we add '<Inset width="..." ...>'?".
Let's see, that's a 5-word question vs. my ~16-word english
explanation.
I think we can leverage our discussions about an XML LyX file format
into a Schema with little effort. And if we can't, then we won't.
>
> I belive we pretty much agree.
>
> My right foot is to begin by cleaning up the .lyx format with out
> introducing too many new constructs.
Yes. And, we could probably move towards XML gradually, too.
--
John Weiss
"Not through coercion. Not by force. But by compassion. By
affection. And, a small fish." -His Holiness, the 14th Dalai Lama
