From: devinkouts_at_earthlink.net
Date: Wed Feb 07 2001 - 16:04:12 CET
Received: (from mdom_at_localhost) by karto.ethz.ch (8.9.3/8.9.3/SuSE Linux 8.9.3-0.1) id QAA20565 for cavexml-outgoing; Wed, 7 Feb 2001 16:01:18 +0100 Received: from [209.70.170.131] (brick.cist.saic.com [209.70.170.131]) by karto.ethz.ch (8.9.3/8.9.3/SuSE Linux 8.9.3-0.1) with SMTP id QAA20561 for <cavexml_at_cartography.ch>; Wed, 7 Feb 2001 16:01:16 +0100 From: devinkouts_at_earthlink.net Received: from cist.saic.com by [209.70.170.131] via smtpd (for karto.ethz.ch [129.132.127.159]) with SMTP; 7 Feb 2001 15:01:55 UT Received: from earthlink.net (unverified [10.43.39.246]) by exmail.cist.saic.com (EMWAC SMTPRS 0.83) with SMTP id <B0000717156_at_exmail.cist.saic.com>; Wed, 07 Feb 2001 10:02:58 -0500 Message-ID: <3A8163EB.E5F160D7@earthlink.net> Date: Wed, 07 Feb 2001 10:04:12 -0500 X-Mailer: Mozilla 4.6 [en] (WinNT; U) X-Accept-Language: en To: cavexml_at_cartography.ch Subject: Re:Groupings Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-cavexml_at_karto.baug.ethz.ch Precedence: bulk Reply-To: cavexml_at_cartography.ch
By natural I mean replicating the fashion in which cave survey is
commonly conducted, and recorded. Five examples of what seems to be
natural are given below:
HA36 HB1 28 30 265
>From David MacKenzie's Walls, read as: From Station, To Station,
Inclination, Distance, Azimuth.
A1 A2 21.75 63.50 -28.00 2.60 2.60 2.60 2.60
>From Larry Fish's Compass, read as From Station, To Station, Distance,
Azimuth, Inclination, Left, Up, Down, Right
B6 B7 104.00 37.50 -28.00 10.0 10.0 30.0 3.0
>From Garry Petrie's WinKarst, read as From Station, To Station,
Distance, Azimuth, Inclination, Left, Right, Up, Down
2 4 7.89 254 -11
>From Olly Bett's Survex, read as From Station, To Station, Distance,
Azimuth, Inclination
Shot=r1 r2 9.80 276.00 -12.00 NAN NAN 1.00 1.00 0.00 2.00 ()
>From Taco Van Ieperen's OnStation, read as From Station, To Station,
Azimuth, Inclination, Distance, Left, Up, Down, Right
I completely agree that the XML format should not try to replicate or be
unduly influenced by one of the existing formats or another. But I would
consider all of these to be natural representations of the way we
accomplish and record a "survey shot". They all follow a basic model,
that looks like this:
>From Station, To Station, Distance, Azimuth, Inclination, (and
optionally) Left, Right, Up, Down
But these shots can be difficult to deal with if simply written into a
file with no organization. They most be arranged in some fashion that
makes it easier for them to be interpreted, associated, related,
indexed, whatever. All of the above formats use some method to organize
shots into "surveys" (which can be segments of larger surveys), and
delimit each survey so that it can be distinguished from other surveys
in the file. For example:
David MacKenzie's Walls uses # and ; to exclude everything that is not a
survey shot. The delimiter of surveys could be construed to be the file
itself.
Larry Fish's Compass uses the Cave Name and a special character to
delimit a survey. Everything appearing between these two delimiters can
be viewed as part of a single survey (which can be incorporated into a
larger survey later).
Garry Petrie's WinKarst uses #SURVEY NAME: and #END tags to delimit a
collection of related shots as a survey.
Olly Betts' uses *begin and *end to delimit a series of shots into a
survey in Survex.
Taco Van Ieperen uses Begin=Survey and End=Survey to delimit a series of
shots that comprise a survey in OnStation (CDI format).
All of these constructs follow a model that organizes "shots" and their
attendant data into surveys. Surveys are expanded over time by one of
two methods; either (1) opening an existing survey and inserting new
shots between the delimiters of that survey, or (2) by creating a new
delimited survey with new shot data, and linking that new collection to
previously existing surveys. These approaches support the way in which
cavers conduct surveys, one survey trip at a time. And each trip might
add stations to an existing collection of survey data, or add a new
survey that can be standalone, or imbedded in or linked to previously
existing survey data. That's how it gets done day in and day out. And it
seems like a pretty easy thing to model in XML.
For example:
I think we can all agree that a survey <shot> or <leg> if you prefer has
a certain set of data that pertains to it. Those elements are outside
the scope of this discussion but I just want to recognize their
existence for arguments sake. The focus here is on <shots> and
<surveys>.
I've just shown you five examples above wherein <shots> are organized
into <surveys>. A natural construct, so lets illustrate it in XML like
this:
<survey>
<shot>first shot</shot>
<shot>second shot</shot>
<shot>etc...</shot>
</survey>
This is just one what in which it can be done with XML but the point
here is that a "survey contains shots".
Now to carry this one step further, a lot of this discussion in this
thread seems to be about how people define, create, extend or relate
surveys to each other. Some take the approach that surveys are
individual collections of data. Illustrated like this:
<survey name="first survey">
<shot>first shot</shot>
<shot>second shot</shot>
<shot>etc...</shot>
</survey>
<survey name="second survey">
<shot>first shot</shot>
<shot>second shot</shot>
<shot>etc...</shot>
</survey>
Others take the approach that surveys are one monolithic collection of
shots that you merely append new collections of shots to. Illustrated
like this:
<survey name="the only survey">
<shot>first shot on first survey trip</shot>
<shot>second shot on first survey trip</shot>
<shot>etc...</shot>
<shot>first shot on second survey trip</shot>
<shot>second shot on second survey trip</shot>
<shot>etc...</shot>
</survey>
There are a few more ways people like to break down their surveys, but
their all basically just variations on the two themes above.
So what's the problem here? Nothing, there is no problem with using
either of these methods. They both represent natural or commonly
employed methods of describing the shot elements that comprise a survey.
The issue lies within the practices of our community, we have
individuals who prefer one approach to the other. How do we reconcile
this? By making sure the XML standard can accommodate both approaches.
So to state the need: A survey must be capable of delimiting a
collection of shots. Furthermore , that survey must itself be expandable
by any of three methods, (1) the inclusion of new shots into the
existing survey collection, or (2) the inclusion of a new survey into
the existing survey collection (creating a parent/child relationship),
or (3) the association (e.g. linking) of one survey to another (creating
a sibling relationship).
This approach creates the constructs necessary to create and expand
surveys by either adding new shots directly, or by adding new
subordinate surveys, or by adding sibling surveys. It further allows
surveys to be organized under a "super" survey construct, which would
facilitate the organization of surveys into morphologic representations
or "cave system" representations.
Examples of these constructs, expressed in XML could appear similar to
the following:
(1) the inclusion of new shots into the existing survey collection
<survey name="the only survey">
<shot>first shot on first survey trip</shot>
<shot>second shot on first survey trip</shot>
<shot>etc...</shot>
<shot>first shot on second survey trip</shot>
<shot>second shot on second survey trip</shot>
<shot>etc...</shot>
</survey>
(2) the inclusion of a new survey into the existing survey collection
(creating a parent/child relationship)
<survey name="the first survey">
<shot>first shot on first survey trip</shot>
<shot>second shot on first survey trip</shot>
<shot>etc...</shot>
<survey name="the second survey">
<shot>first shot on second survey trip</shot>
<shot>second shot on second survey trip</shot>
<shot>etc...</shot>
</survey>
</survey>
(3) the association (e.g. linking) of one survey to another (creating a
sibling relationship).
<survey name="the first survey">
<shot>first shot on first survey trip</shot>
<shot>second shot on first survey trip</shot>
<shot>etc...</shot>
<survey name="the second survey">
<link>Xpath link to the second survey</link>
</survey>
</survey>
<survey name="the second survey">
<shot>first shot on second survey trip</shot>
<shot>second shot on second survey trip</shot>
<shot>etc...</shot>
</survey>
I could throw in a few more examples to show how you might organize
surveys into cave systems, or morphologic organizations, but I really do
have to get back to work. I think all the basic constructs are available
here to play with, Parent/Child, Sibling and straight sequential. Please
take some time, kick it around and respond with your thoughts.
Thanks,
DSK
-- Devin Kouts Caver Systems Engineer www.psc-cavers.org
This archive was generated by hypermail 2b30 : Thu Mar 01 2001 - 18:00:00 CET