Re: The politics of finding errors Was:XML

From: Lemming <thiswillbounce_at_bumblbee.demon.co.uk>
Date: Thu, 21 Oct 2004 22:45:34 +0100
Message-ID: <f3bgn0h7np6sm0cbkjhk7487qnm163t285_at_4ax.com>


On Thu, 21 Oct 2004 16:12:08 GMT, Bernard Peek <bap_at_shrdlu.com> wrote:

>In message <GsydnV_UCto7h-7cRVn-uQ_at_comcast.com>, Laconic2
><laconic2_at_comcast.net> writes
>>I'll leave the question of HOW XML settled arguments "at a stroke" to the
>>person who made that claim.
>>
>>My only point was that, provided it truly did that, it added a large amount
>>of value.
>
>That was my original claim. The industry had several "standard" data
>interchange formats, but there was no single authoritative description
>of the formats. I could have tried to nail down an existing standard,
>but each of the companies would have argued for their own particular
>variant. And in any case a lot of the data files were hand-edited and
>didn't comply with whatever standards were supposedly being used.
>
>A single XML format had a number of effects. Although it didn't solve
>data quality problems those had been quite rare, the main problem we
>faced were formatting. Each company wasted approximately half a skilled
>person just reformatting data files. One of my best developers left
>because he was tired of spending his time reformatting files rather than
>what he was supposedly paid for.
>
>Another benefit to my company was that it established itself as
>custodian of the schemas, consolidating its position as the centre of
>the industry's IT research.

That is a convincing argument. I was focussed on the data processing aspect of the problem, and what happens when things go wrong with the data. Your anecdote is more concerned with how much easier it is to agree on a format if you begin with a pre-existing - and perhaps more importantly, widely accepted - method. Even if that method is less than optimal for the purpose, if you can get agreement on it's use then that's a benefit in those situations where politics holds sway over technical considerations - a situation which is more common than most of us would prefer it to be.

Lemming

-- 
Curiosity *may* have killed Schrodinger's cat.
Received on Thu Oct 21 2004 - 23:45:34 CEST

Original text of this message