Re: [xml-dev] It is okay for things to break in the future!

["C. M. Sperberg-McQueen" <cmsmcq@blackmesatech.com>]

Roger L Costello <costello@mitre.org> writes:

> There is a difference between:
>
> -	Failing to understand the current world and then writing incorrect and/or incomplete code
> -	Not knowing the future
>
> My post was about the latter, not the former.

If you can distinguish reliably between the current world and the future
world, then you may be right that there is a difference.  But over time
I have come to agree more and more with Bruce Gibson's remark:  the
future is already here - it's just not evenly distributed.

You are certainly right that the failure of that web site to accept a
hyphen in my last name was not a failure of future-proofing but a
failure to understand the variety of forms taken by family names in the
U.S. today (and for the last couple of centuries).


That said, I continue to be confused by the bright line some people
appear to see between "things that it makes sense to put in a schema"
and "business rules"; the distinction seems to commingle issues of
technology (e.g. what language do you use to express a given
constraint?), issues of permanence (is this constraint permanent or
subject to change? if it changes, who decides?), and issues of ease of
change (if some constraints are subject to change and others are not,
you probably don't want to build the former into levels of your data
store that are difficult or tedious to change), mixed occasionally with
what I think is simple misinformation about existing technology (I have
heard, more often than I care to think, that the schema in an RDBMS is
effectively impossible to change, although one of the key innovations in
the development of RDBMS was precisely that it's easier to change the
schema than previously).


-- 
C. M. Sperberg-McQueen
Black Mesa Technologies LLC
http://blackmesatech.com