From: Garry Petrie (gp_at_europa.com)
Date: Thu Feb 01 2001 - 17:31:20 CET
Received: (from mdom_at_localhost) by karto.ethz.ch (8.9.3/8.9.3/SuSE Linux 8.9.3-0.1) id RAA21163 for cavexml-outgoing; Thu, 1 Feb 2001 17:40:44 +0100 Received: from io.europa.com (IDENT:root_at_io.europa.com [216.65.131.2]) by karto.ethz.ch (8.9.3/8.9.3/SuSE Linux 8.9.3-0.1) with ESMTP id RAA21159 for <cavexml_at_cartography.ch>; Thu, 1 Feb 2001 17:40:43 +0100 Received: from europa.com (ip85.gte15.rb1.bel.nwlink.com [209.20.244.85]) by io.europa.com (8.11.1/8.11.1) with ESMTP id f11Gf0605524 for <cavexml_at_cartography.ch>; Thu, 1 Feb 2001 08:41:00 -0800 (PST) Message-ID: <3A798F58.9020907@europa.com> Date: Thu, 01 Feb 2001 08:31:20 -0800 From: Garry Petrie <gp_at_europa.com> User-Agent: Mozilla/5.0 (Windows; U; Win98; en-US; m18) Gecko/20001108 Netscape6/6.0 X-Accept-Language: en,pdf To: cavexml_at_cartography.ch Subject: Re: Groupings References: <200102011429.PAA20204_at_karto.ethz.ch> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-cavexml_at_karto.baug.ethz.ch Precedence: bulk Reply-To: cavexml_at_cartography.ch
Richard Knapp wrote:
>> A "survey" to me, only represents data that was collected at one particular instance and
>> as such, by construction, share common attributes.
>
> Reasonable point. Let's take a survey then, say the "A". It was started on 1-Jan-2001. The
> next trip with a new crew continued the survey (2-Jan-2001). A couple months later (after
> beating out some smaller leads), the survey was continued with the original crew
> (9-Sep-2001). This is one survey -- the A -- but done in three parts. How would this
> accurately be recorded?
You just said the data was recorded on different dates! When I take notes in the cave, the
first page in my note book records date, people, locations, etc. When I get home, I keep
that page with the measurements bundled together. It is kind of like going to the doctor's
office, he just doesn't start writing on the same page where he left off the last time he
saw you.
In XML the survey identifier needs to be decided as to whether it is unique or not.
Certainly, the cave idenfifier is unique, but I believe good laboratory procedure requires
data recorded for an "experiment" be organized such that it can be repeated or verified later.
Roger Schuster wrote:
>> A "survey" to me, only represents data that was collected at one
>> particular instance and as such, by construction, share common attributes.
>
> This seems to be a practice very common in the US but in Europe most cavers
> structure the data in a way in which each survey represents a certain
> passage. The structure of the data represents the topology of the cave and
> this is the way how the survey editor of software like Cave Render or
> Toporobot works.
Here is an example of the application driving the data recording. In effect, the data set
will never be larger than the capabilities of the software, which leads to multiple data
sets representing the same object, each tailored to the application.
Paul & Eleanor wrote:
>> That is the whole point of a relational database, to group data by
>> attributes.
>
> Except we aren't defining a relational database.
A relational database is an application, not data. For example, you can freely move data
between Microsoft Excel and Access.
Garry Petrie
This archive was generated by hypermail 2b30 : Thu Mar 01 2001 - 18:00:00 CET