LRUD (was Re: Roadmap)

[Mike McCombe]

Interesting to see all the LRUD issues coming up again. Speleogen kept the
LRUD data in a separate file from the Survex data. This was firstly so that
Survex didn't need to be modified to parse and discard the extra data and
secondly because they're very different from the survey leg data. Apart from
station naming issues (i.e. within a *begin/*end block) the order in which
legs are specified in Survex is irrelevant. This has the advantage that
leapfrogging (or not), junctions etc. are all handled easily in Survex. On
the other hand, to use LRUD dimensions to recreate passage walls, the
sequence of stations is an essential piece of information. A station at a
junction, for example, would have different LR values depending on which
branch is followed. Whatever convention is chosen for the orientation of LR,
provision needs to be made for passage breaks and junctions.

A more modern approach than separate files would be to define survey legs
and passage details in a single XML file. How this is created (authoring
tools?) is another issue. Separate files still look to me to be the most
straightforward approach.

Speleogen adopted a convention of LR being perpendicular to the line from
the previous station in the sequence.  Usually the L and R points are one on
each wall of the passage. However, where the passage folds back on itself, L
and R are both on the same wall (i.e. the outer wall of the bend) and a new
point is needed to mark the corresponding point on the inside wall of the
bend. Rather than force surveyors to adhere to some complicated rules to
handle such situations, Speleogen was designed to make a sensible estimate
of where the inner point would be. As with junctions, it was assumed that
the drawing would be corrected later using a CAD package and the surveyor's
sketches.

The L/R conventions used in Speleogen were not regarded as better/worse than
any others. They just happened to be what the original users wanted.

Mike


----- Original Message -----
From: "Eric Madelaine" <Eric.Madelaine@sophia.inria.fr>
To: "Survex User Group" <survex@survex.com>;
<Eric.Madelaine@sophia.inria.fr>
Sent: Monday, April 22, 2002 1:12 PM
Subject: Re: LRUD (was Re: Roadmap)


>
> I've been following the LRUD thread from time to time, and I think I can
give
> my token at this point...
>
> I have been using LRUD fonctions in the french software "VTOPO" for a
while,
> and also, on a couple of occurences, with Toporobot.
>
> VTOPO provides more or less what you are speaking about: vertical
measurement,
> and a choice of either bissectrices or orthogonal side measures.
> I also use frequently a number of tricks for better rendering of the
volumes,
> that make my sources files look somewhat different from the underground
notebook,
> like duplication of stations at which the dimension change brutally (with
a zero
> length leg in between), or addition of intermediate stations within a leg
> when needed.
> This scheme works pretty well for printing a plan of the caves with
> understandable walls (I have not said realistic...).
> It is much more difficult with elevations. I guess that the Toporobot idea
of
> having non vertical UD measures at the top of a pit, for exemple, could be
a
> very desirable addition. Or (may be using the same kind of tricks than
before),
> having several UD mesures, like vertical, bissectrice AND horizontal at
the top
> of a pit, horizontal at intermediate stations in the pit, etc... Of
course,
> this supposes that the direction of measurement is made clear, if you want
to
> keep the simple scheme where LR is used for the plan, and UD for the
elevation;
> and then combine in a simple, rectangular section, volume model for the
3D.
>
> Eric.
>
>
>
>
> --
> Survex http://lists.survex.com/mailman/listinfo/survex
>