[LandXML] Re: Any datum stated in LandXML standard ?
- From: Raymond Wilson <Raymond.Wilson@xxxxxxxxxxxxx>
- To: "'nathan.crews@xxxxxxxxxxxx'" <nathan.crews@xxxxxxxxxxxx>,landxml@xxxxxxxxxxxxx
- Date: Tue, 5 Nov 2002 09:12:19 +1300
Disclaimer: I am not knocking the format, the working group or anyone else
involved - I am just sticking my oar in, so to speak :-)
A file does not need to define explicitly the units of measure used within
it. The file format specification can explicitly determine this - the
readers/writers of the format then adhere to the specification. The DXF file
format is an example of a file format that has no implicitly or explicitly
defined units (and it is a pain indeed!).
The example you cite of the intrument vendors stating that radians are not
supported in the device and that dd.mm.ss would be preferred is one which
IMHO, should be resisted (in this example, the request is not one of support
for different units, but a different formatted representation of an existing
unit). Taking a pragmatic view, the effort for an intrument vendor to
perform a simple operation on an angle in radians to convert it to their own
internal format pales next to the effort to support LandXML. Notwithstanding
this, as the current LandXML standard already permits radians among other
angle units, the LandXML reader on the intrument must support them anyway.
I hope I am preaching to the choir here when I say LandXML should be viewed
as a format that represents, stores and permits interchange of 'land'
information. This purpose is, again IMHO, better served by a format that
expresses this information simply. Adding a layer of uniting, representation
and other semantic indirection does not, I think, serve this purpose. I
appreciate that allowing a LandXML file to have user defined units does
allow a human to relate the values in the LandXML file directly back to
their own locale when viewing the file directly - was this a prime
motivation to including them?
We should be careful when considering LandXML schema additions that
represent either 'semantic sugar' or extensions (such as the intrument
vendor request) that do not enhance the information carried within it. In
the dd.mm.ss example, adding it to the schema obligates all LandXML
readers/writers to support it, simply to allow an instrument to read the
value directly, rather than convert it from radians/degrees etc.
In closing, I am reminded of a recent NASA Mars probe disaster, where a
software component from the US and another from Europe worked in different
units (feet and meters respectively). It was not a happy marriage!
Cheers,
Raymond.
-----Original Message-----
From: nathan.crews@xxxxxxxxxxxx [mailto:nathan.crews@xxxxxxxxxxxx]
Sent: Tuesday, November 05, 2002 1:02 AM
To: Raymond Wilson; landxml@xxxxxxxxxxxxx
Subject: RE: [LandXML] Re: Any datum stated in LandXML standard ?
Without detailed units of measure information, the data in any file is
useless. One specific issue was the difference between angular units and
directional units. The angular units were assumed to be radians prior to
last year. At one point, I also though SI units simplified the schema, but
in practice more complex definitions are required.
In a discussion last year, several instrument vendors indicated that radians
were not supported natively in the current hardware and that dd.mm.ss format
was supported. If the field instruments were to ever support LandXML,
additional units of measure needed to be added to the schema.
The changes were broadly discussed and accepted by the working group, just
as any change occurs to the schema.
Nathan Crews
www.landxml.org
-----Original Message-----
From: Raymond Wilson [mailto:Raymond.Wilson@xxxxxxxxxxxxx]
Sent: Sunday, November 03, 2002 3:48 PM
To: 'landxml@xxxxxxxxxxxxx'
Subject: [LandXML] Re: Any datum stated in LandXML standard ?
This is a related issue I thought I'd bring up (sic) for some clarification.
I was not a party to the discussions that resulted in the LandXML v1.0
standard and am curious as to the motivations for allowing the LandXML
format to have defined linear/angular units.
In my experience file formats that have this ability tend to be more
difficult to work with than if the format mandated (for instance), SI units.
This would mean linear units were always meters and angular units may be in
radians expressed clockwise from the north azimuth. The effect of this is to
reduce the complexity of the reader/writer by removing (among other things)
the need to perform an arbitrary unit translation rather than a single
defined one.
Anyway, I would be interested if anyone here could enlighten me to the
reasons LandXML adopted user defined units instead of SI units.
Cheers,
Raymond.
-----Original Message-----
From: nathan.crews@xxxxxxxxxxxx [mailto:nathan.crews@xxxxxxxxxxxx]
Sent: Saturday, November 02, 2002 12:20 AM
To: bleblanc@xxxxxxxxxxxxxx
Cc: landxml@xxxxxxxxxxxxx
Subject: [LandXML] Re: Any datum stated in LandXML standard ?
Yes, this is the right place to discuss any public issues regarding LandXML.
I will make sure the LandXML-1.0Docs.zip file is fixed.
The Units element details all units of measure contained in the data
structure, which includes linear units and angular units. The
CoordinateSystem element describes the coordinate system and datum used.
Nathan Crews
www.landxml.org
-----Original Message-----
From: bleblanc@xxxxxxxxxxxxxx [mailto:bleblanc@xxxxxxxxxxxxxx]
Sent: Thursday, October 31, 2002 1:35 PM
To: landxml@xxxxxxxxxxxxx
Subject: [LandXML] Any datum stated in LandXML standard ?
this format, some or all of this message may not be legible.
-- Attached file included as plaintext by Ecartis --
Hi,
First, I don't know if it's the right place to ask my question. If not, I
apologize and would like to know where to do so. I tried to find what I'm
looking for in the LandXML-1.0Docs.zip document that is on your web site.
Unfortunately, the file is corrupted, Winzip displays the message Bad CRC
ca3d7493 I downloaded it three times to ensure that corruption didn't
occurred during data transmission. Anyway, he it is...
Here at Ministry of Transports of Quebec (Canada), we use Bentley's Inroads
civil design software. Through Inroads and all software involved in our
road project workflow, we apply a combined scale factor (obtained by the
multiplication of grid scale factor and mean sea level reduction factor) on
each input lengths to express field data on a projection plan (something
equivalent to the State Plane systems used in several US states). The goal
is that all reports produced by Inroads show grid level related coordinates
and field level related dimensions (stations, intervals, radii, lengths,
etc.).
Recently, for reporting purposes, we decided to export some data (surfaces
and alignments) in LandXML format. We also plan to use the same standard to
upload road designs in our electronic fieldbooks, for field deployment
purposes.
We'd like to know if somewhere in LandXML standard, there is any statement
about the reference datum to use for any dimensions contained in LandXML
data files. It's quite obvious that coordinates should be expressed on the
projection plan, but what about all linear dimensions like curves radii and
alignments stations? For us, these should be ground (field) referenced. Is
this topic addressed by the LandXML standard?
Hope my question is clear enough. If not, feel free to contact me for
additional information.
Best regards,
www
(O o)
---------------------.oOO--(_)--OOo.----------------------------------------
-
Benoit Leblanc Tel: (514) 873-2263
Ministere des Transports du Quebec Fax: (514) 873-8203
Direction des Technologies de l'information
Montreal (Quebec) Canada bleblanc@xxxxxxxxxxxxxx
----------------------------------------------------------------------------
-
Other related posts:
