Re: [xml-dev] provocations and realities (was Re: Fwd: [xml-dev]Not usin

["Simon St.Laurent" <simonstl@simonstl.com>]

On 4/9/13 1:52 AM, Liam R E Quin wrote:
> But you are still taking a strong rhetorical position :-)

Making the shocking normal generally requires shocking people first.

It's not mere rhetoric - it's that the reality is disconcerting, to put 
it mildly.

>> At least I'm getting a gentler reaction than Walter Perry used to get,
>> though it may just be that I'm offering a milder line.
>
> I always felt his arguments were not about schemas per se but about
> specific vocabularies,

Yes.  I hope that Walter will eventually call me on that heresy, and 
then we'll have proper splintering.

I agree with Walter's broader position on standard vocabularies. 
Schemas earn an extra level of contempt because they are the carrier of 
the problem, the concrete form we markup folks have chosen for that 
particular set of toxic values.

Why target the carrier?  That comes from the Christopher Alexander work. 
  Structuring architecture around fixed drawings - a temptation at every 
level - imposes costs at every level.

It's not that he doesn't draw, but rather that he doesn't make 
schema-like promises in those drawings.  Unless they are of completed 
work (and maybe not even then), they are open to change through 
conversation.

> and remember well that Michael Sperberg-McQueen,
> after Walter had said the markup language doesn't matter because the
> data itself is sufficient, refuted him in German. Or something
> sufficiently German-sounding to fool me.

I'm learning German from my five-year-old, slowly.

I - how can I say this? - rarely find Michael's refutations persuasive, 
even when they are in a language I pretend to understand.

> He said in private conversation afterwards the the primary documents
> with which he was dealing contained only a few items - I think price,
> quantity, stock ticker symbol, buyer, seller, date/time. In almost all
> cases the lexical forms would be sufficiently distinctive that there
> would be no confusion, especially with ticker symbol/price redundancy.
>
> But there are other use cases for which that's not the case, of course.

Yes.  I can't speak for Walter, but in past conversation we haven't 
found intrinsic reasons why the approach can't work with more complex 
documents.

(A parallel in Alexander might be the Large-Scale Production chapter in 
The Production of Houses, or the entire project described in The Battle 
for the Life and Beauty of the Earth.)

> We also agree that Christopher Alexander is worth reading. I should very
> much like to get to this book but it won't be until after the
> Restoration of the House.

That is deliciously ironic.  I suspect you've already seen A Pattern 
Language, but if you haven't, a glance through might be good before the 
Restoration.  If not, the Site Repair pattern is always useful...

Thanks,
-- 
Simon St.Laurent
http://simonstl.com/