Re: [xml-dev] Four fine text-based data formats ... liberate yourselffro

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

On 3/25/13 6:37 PM, Liam R E Quin wrote:
> On Mon, 2013-03-25 at 09:10 -0400, Simon St.Laurent wrote:
>> I have severe problems with using schemas to lurch toward the maximum
>> possible straitjacket that can fit some part of a problem.
>
> You could replace "schemas" in there with "leather straps" or "UML" or
> almost any sort of constraining technology.

No, I don't think you could.  UML is not integrated as deeply with the 
surrounding technologies as schemas are with the XML tool set.  XML has 
demonstrated itself an expert at creating constraints, appropriate or not.

> XML Schemas are not a substitute for thinking and understanding. Neither
> (as others have said) do they necessarily enforce a waterfall model.
> Continuous refinement and agile spirals are equally possible.

They are possible, but not "equally" possible.  I thought you'd spent 
some time around standards organizations?

> A lot of people came to SGML and XML from "big engineering", civil
> engineering (leather straps, I think, are not very civil) and were used
> to the idea that you do a formal specification before you try building
> anything.

It's not just big engineering - it's even those drunken sots in 
publishing.  And medicine.  And shipping.  And finance.  And, well, name 
your field.

> A lot of money gets wasted that way, but on the other hand agile
> development for aircraft or space rocket design also has problems. The
> cost of experimentation is too high, so you use simulations, and tests,
> ideally a long way away from anywhere else, e.g. in Sim-mule Asian
> Cities.
>
> You could argue that publishers enforce a waterfall too - first the book
> proposal, table of contents, sample chapter, then the agreed-upon
> outline and schedule, and then a sequence of words, magically followed
> by publication of a physical object. In fact, of course, the proofing
> and editing cycle can also be a spiral, and the final book doesn't need
> to be very linear (it's hard to make an effective narrative within a
> non-linear rhetorical framework but see e.g. "Patchwork Girl").

Many corners of the publishing world are actually starting to recognize 
the waterfall portion of that cycle as a disease, and striving to make 
the cycles more public.  We have to - for all the complaints about lack 
of curation and quality, self-publishing approaches pay none of the 
costs of that waterfall.

(Writing itself rarely follows a waterfall - more like a river that 
floods and dries, starting on one course and ending up in another.)

> So when you rail against XML and say how much more comfortable you are
> with JSON, it sounds to me as if you're really railing against those
> shoe-wearing bureaucratic civil engineers who wanted
> first-this-then-that.

No, it's not specifically about engineers.  It might be fairer to say 
that it's a complaint against lawyers - tipping my hat to Charles 
Goldfarb - binding that paranoid culture of required and adjudicated 
prior agreement with various layers of computing technologies.

It's also quite definitely about a strangely industrial model of 
development, in which the specifications are built and the machines set 
to work.  (Hence the Christopher Alexander quote earlier.)

> I agree that's a problem, although if several different organizations
> are involved it's often necessary politically if not technically, and
> the schema there can help (as David and others suggest) as a sort of
> documentation.

Yes, the disease often starts that way.  It may stay mild, or it may not.

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