When i begin this list in the home page i talked about DWG format.
Well in the page 34 of the STEP e-book you can see in a photo that DXF
(and therefore DWG) is a complete CAD format.
We want build an open cad format similar to DWG...
In the page 13 they tell us:
"the real importance of STEP lies in the business
drivers that are creating a new environment for manufacturing
industries in which
the opportunities offered by STEP may be realized."
STEP was born to manufacturing industries.
You take a look to the companies working in openDWG. They are all CAD/
CAM/GIS/databases companies:
http://www.opendesign.com/membership/sustain.htm
Microstation V8 works with DWG,DXF and DGN like native formats.
and export to this formats:
IGES
CGM
STEP AP203/AP240
VRLM
Parasolids
ACIS SAT
Bentley Explain STEP export like this:
"""""Application Protocols (APs) are used to exchange data. Each AP
pertains to a different application area. For example, AP227 pertains
to spatial plant technology and contains several entities pertaining
to that area, such as pipe or elbows. Also, these APs use certain
common entities known as generic resources, such as geometry and
topology, to define solid models.
AP203 applies to representations of mechanical parts and assemblies.
AP214 applies to representations of data relating to automotive
design. Present day AP203 files typically contains the boundary
representation model, assembly data, and a limited amount of other
product information. AP214 files typically contain colors, layers, and
generic resources.
The boundary representation of a model is a CAD model represented by
its boundary. For example, a surface model consists of only the
surfaces use to make the model. Solid models include geometry
information, such as surfaces, curves, and points, and topology
information, such as edges, vertices, and faces. The geometry
information provides data about the form of the model, and the
topology information provides the connectivity between these geometric
elements and the extent of the geometric elements.
STEP AP203 files are divided into different classes according to the
type of model that they represent. For example, a surface model is a
Class 2 representation, and a solid model is a Class 6."""""
Regards.
http://open_cad_format.googlegroups.com/web/STEPbook.pdf
Put Ford Orion in a CAD file. This is a crazy WORLDDDDDDDDDD!!!!!!!!
This kind of things are to databases like Mysql, Oracle, postgree,...
With a database server you can share this information between all
company.
Autocad and Microstation have database conections to this kind of
crazy things.
Regards.
The point is that applications that share data across networks, people,
machines tend to define protocols first, formats second. That is also
why I was trying to delay a big debate on encodings - people are getting
less interested in the format, and you can change it without effort.
At the moment, people like to put shared data into database servers as
you describe it. To exchange this data, it is frequently encoded into
files, that is why people care so much about xml. It is the same with
CAD. Big installations have drawing databases, which are better called
building databases, and use regular file formats more or less for exchange.
The point is how to strucure the data. Than you can write it into a
file, archive it on tape, store it in databases.
It may appear strange to put CAD data into databases if yoy think of it
as lines. But do not forget the meaning of CAD, Computer Aided Design.
Design is not just lines. The metadata associated with these lines is
just as valuable.
CU Lars.
On Oct 8, 11:07 am, "Lars O. Grobe" <gr...@gmx.net> wrote:
>
> It may appear strange to put CAD data into databases if yoy think of it
> as lines. But do not forget the meaning of CAD, Computer Aided Design.
> Design is not just lines. The metadata associated with these lines is
> just as valuable.
I didn´t know this problems. I thought in CAD only like lines, colors,
layers, and therefore for me was very easy the problem hehe.
But it´s supposed that lines, colors and layer are always the same for
all??
At least we can share that.
?it´s important share database information between all different
fields??
if it´s important this then build a CAD format for all is a very
difficult task hehe.
////////////////////////////////////////////////////////////////////////////////
////////////////////////////////////////////////////////////////////////////////
The sample put here i like a relational database with primary key in
STEP format:
ENTITY car;
make : STRING;
model : STRING;
year : INTEGER;
owner : person;
END_ENTITY;
ENTITY person;
first_name : STRING;
last_name : STRING;
END_ENTITY;
Instances of this data model that represent a 1989 Ford Orion, owned
by the
author, are encoded using the STEP file format as follows:
#1 = CAR ('Ford', 'Orion', 1989, #2);
#2 = PERSON ('Julian', 'Fowler');
////////////////////////////////////////////////////////////////////////////////
////////////////////////////////////////////////////////////////////////////////
Regards.
http://xml.coverpages.org/ni2001-03-19-c.html
XML parsers are more mature to parse text i think.
Regards.
Hi Martin!
I would like to explain a bit of my perspective. By CAD files as a
database I mean that for me, conceptual, a CAD file contains
heterogenous information. I describe as in architecture, but it will be
similar in other fields.
I have an object. Call it entity or whatever. It often has geometric
properties (type, position, scale, rotation, ...). It often has material
properties (concrete, wood, steel, rubber, water). It has context
properties (is part of 1/5 detail drawing IV, but not visible in 1/500
scale). It has price, provider. It has an author/drawer/content creator,
drawing and review history. It may have a name, part number. It may be
just a reference to another object. It may be a text, annotation,
correction.
In many cad systems these properties have never been addressed.
Workarounds were introduced, mostly derived from paper-times. Some
people use layers to distinguish between materials, scales, or to group
objects that are supposed to appear on an output plan. Many CAD systems
than invented multiple layer systems, because projecting everything to
layers led to documents with hundred of Layers. In Vectorworks e.g.
people have one layer system to distinguish between plan-world
characteristics (e.g. is on sheet "1st floor plan") and model-world
(e.g. belongs to concrete walls). In Autocad, people frequently use a
combination of layer and color to divide their drawings. But still, just
lay two storey's plans on top of each other without making the hole
plans into groups... its a pain.
So when people say that color, layer is enough, it is often because they
are (I am too) used to workarounds. But it is a painful workaround and
makes e.g. data extraction almost impossible. That is what I mean by
metadata - it appears as line/color/layer in many drawings today because
that is what we can do with our applications.
You mentioned that while step was supported in manufacturing industries,
architectural CAD supported dwg and compatibility libraries. I think
this is the result of just the problem this group describes - no open
format. The automotive or aerospace industry e.g. have structures strong
enough to enfore open standards. So they give a standard, and the
software companies have to implement it. In e.g. building, take a small
office, five people. Be sure they already use 30% cracked software
because the licenses are far beyond their budget. If they call their
software supplier, the end up in a paid support hotline. How should they
put pressure on their CAD provider to support the standard they need?
And how should they develop this standard - ever seen a mid-scale
architectural office with an it department able to even think about such
questions? So they are told that dwg was the standard, and they are
offered to buy a solution that allows them to stay in market. Should
they say no, we want a vendor-neutral solution? So here is the visible
difference: step was designed by the costumers, implemented after the
definition of the standard. dwg was designed as the native format of one
company, and the costumers have to use it.
I think the fist time that the fine-grained structure which made dwg and
other closed formats continue to exist even though they cause all kinds
of headache and force people to stay with products even if they would
like to change has come now with the nature of open source. Here
individuals and small companies can cooperate, and this discussion here
may be an example.
Your argument for xml may be a good one as there is a number of projects
to parse and validate xml. This may leverage development. Besides that,
I do not care so much if my data is encoded in xml (files), a big
database (on whatever medium it is) or an archive of thousands of files.
xml is useful because it supports us with a big toolset to parse and
transform, things we do not have to reinvent or reimplement. I wonder if
this could e.g. solve the problem of "unordered" files as Franz showed
in his step example. Maybe a highly optimized xml transformation (that
others were kind enough to develop already) would be enough to make
something that we can use in memory out of our chaotic step file?
CU Lars.
I add a great STEP document here:
http://groups.google.com/group/open_cad_format/web/STEP_Application_Handbook.pdf
I believe that STEP is complete, but imposible to implement in CAD.
In this document there all STEP AP especification, Many companies work
only with little parts of STEP. No full implementation nowadays.
I think that we would need a surveying and architectural AP
especification...
A CAD is a crazy world hehe.
>
>Your argument for xml may be a good one as there is a number of projects
>to parse and validate xml. This may leverage development. Besides that,
>I do not care so much if my data is encoded in xml (files), a big
>database (on whatever medium it is) or an archive of thousands of files.
>xml is useful because it supports us with a big toolset to parse and
>transform, things we do not have to reinvent or reimplement. I wonder if
>this could e.g. solve the problem of "unordered" files as Franz showed
>in his step example. Maybe a highly optimized xml transformation (that
>others were kind enough to develop already) would be enough to make
>something that we can use in memory out of our chaotic step file?
>
>
>
What Franz sample do you say?
Regards.
> I believe that is very difficult for CAD companies build complex CADs. I
> would prefer other CAD format based in trees, not in layers-colors like
> you say but i think it´s a difficult task to program. For example i love
> ADOBE illustrator format, because it´s based in trees, not in simple
> layers like Autocad.
It is a feature I really love in formZ too. You can (sub)group layers
there, creating a hierarchical layer structure. But still it is layers,
nothing else but the digital transformation of transparent sketching
paper. I would prefer to define attributes as I need. And in the
view-setting, give selected attributes a "switches-visibility"-status
e.g. So I do not need a layer at all, if I can set "if value of
scale-attribute != 100 then visibility = FALSE;" in the view dialog,
prefereably not in a text line but a nice GUI dialog. Software can
filter by any attribute, why using layers?
> I believe that STEP is complete, but imposible to implement in CAD.
> In this document there all STEP AP especification, Many companies work
> only with little parts of STEP. No full implementation nowadays.
> I think that we would need a surveying and architectural AP
> especification...
It is enough to support only the part of step that is needed. I cannot
tell if a specific AP is needed, I did not study the existing ones
enough. But while the application would only support what it can
actually use, the format would be capable to store more. So that is what
makes it possible to use an application-independent format. Every
application has a - at least slightly - different feature set. If the
format supports a superset of the features, no information is lost. If
it supports only a subset of all the features of all applications I
start to loose.
> A CAD is a crazy world hehe.
:-)
> What Franz sample do you say?
He gave a sample showing how difficult it is too parse a step file. As
far as I understood it is possible that you read in an entity in step
that references values that are defined later in the stream. This gives
a lot of headache if you do not want to read in all the document and
"compile" it in memory than... So preprocessing this may take away
workload from the parser. But here Franz can tell us much more...
CU Lars.
On Oct 8, 2:25 pm, "Lars O. Grobe" <gr...@gmx.net> wrote:
> It is a feature I really love in formZ too. You can (sub)group layers
> there, creating a hierarchical layer structure.
Exactly the same in Adobe illustrator.
Microstation solve partially the problem in this way:
You only have layer but you can find sublayers by pattern matching
house1_window
house1_wall
house1_stairs
house2_wall
house2_window
house2_stairs
here you have 2 sublayers (house1 and house2), but no more than 1
sublayer posibility. I suppose that Bentley do this things to have
100% compatibility with DWG.
> prefereably not in a text line but a nice GUI dialog. Software can
> filter by any attribute, why using layers?
>
I think layers is a natural way to work in the world:
A base layer is a house
A room is inside a house (sublayer)
a table is inside the room (sublayer)
and so... (sublayer)
filter by any attribute is possible using trees. Illustrator let us
select by all kind of atributtes.
I believe that filter by any attribute and a layer/sublayer system is
totally compatible.
Maya, blender, k-3d and Shake work so, but they show us the layers
system like nodes widgets.
Regards.
http://stevenchan.us/files/tut_facingratio_surflum/17.png
Maya is the software to build all kind the 3D games and films(srek,
toy story,...) and was made in motif/openmotif.
This kind of widget i haven´t seen in GTK or Qt nowadays.
It´s more common this:
http://www.oz9aec.dk/software/gpredict/loc-tree-sign.png
This is a gtktreestore widget similar to sublayers system in
illustrator.
Ad:
> far as I understood it is possible that you read in an entity in step
> that references values that are defined later in the stream. This gives
Yes, thats correct.
You must import all at once and order it, if you want to put it into an
interpreter.
> a lot of headache if you do not want to read in all the document and
> "compile" it in memory than... So preprocessing this may take away
it took 2 months for 2 people to write the step-translator; this is not so
much.
BUT you must have native support for all types of objects you want to import
(analytical described surfaces and curves) or change it at least into a
supported type of object (eg if you cannot support a Parabola you have to
translate it into a B-Spline).
We have decided to create gCAD3D as an interpreter-system, which is completly
different from most CAD-systems.
We made the step-importer (without Express-Tool) to show that a
interpreter-system can work also with acceptable speed.
Greetings from Vienna,
(from AUSTRIA, not Germany !)
Franz
> I think layers is a natural way to work in the world:
> A base layer is a house
> A room is inside a house (sublayer)
> a table is inside the room (sublayer)
> and so... (sublayer)
I don't think layers are naturally, lines can only on one layer at a
time, lines on different layers can not share properties.
If the base layer is a house and the room is on a sublayer some lines
of the house has to be duplicated in the room layer.
I like to work layer independent, a line or object has a set of
properties, it is a wall or a window etc.
These properties don't include color linetype etc, it is just a name.
All these lines with there properties compiles a model, a building,
car or map.
Different people like to look at a different way to a model, you have
to define a presentation for a particular view.
A presentation consists of the properties normally found on a layer,
color, weight, style, a presentation also contains filters.
An architect, a electrician or a constructional engineer all want an
other presentation of the model.
For example the constr. eng. doesn't want to see the electric wiring
(filter), the electrician wants to see the wiring in thick lines
(presentation).
But the all want to see the holes through the construction.
If you want to achieve these things with layers many many layers has
to be used, all the different people working on the project must use
the same layers.
The more layer there are the more difficult it gets to get all lines
on the right layers, it is far more easy to bind a property to a line.
There is also a big difference between working in layers and trees.
If you want to work with trees i think this is only possible when
working with objects (root = car, branches = wheels, engine, etc.)
With trees it is more difficult/impossible to share properties between
lines, and this is important in gis, a line can not be part of a house
and also be part the grass. I don't want to use two lines to get this
done.
So is it possible to define a format suitable for all purposes?
If you look at AutoCAD it serves all but it is not very good in it (it
is good 2D production drawings).
Greetings,
Gertwin
> You must import all at once and order it, if you want to put it into an
> interpreter.
Do you think that using some kind of compile step would help here? I
mean a tool validating and reordering the file, so that this memory
consumer does not happen in the main app? I was wondering if the
xml-tools available would help to do so. I have to look how they solve
this problem in xml (how do they reference values in xml that are not
defined yet?).
> BUT you must have native support for all types of objects you want to import
> (analytical described surfaces and curves) or change it at least into a
> supported type of object (eg if you cannot support a Parabola you have to
> translate it into a B-Spline).
Yes, that looks like a huge task. But if one thinks not about writing
from scratch but reusing code, it gets less awful. I think that there
are already engines out there that are able to handle at least the very
most of step geometry objects (brlcad, cascade, ...). So if one binds
one of these libraries, at least the second option (converting to
something already supported) should be possible.
> We have decided to create gCAD3D as an interpreter-system, which is completly
> different from most CAD-systems.
Is this documented? I imagine something like what happens in brlcad when
using scripts to call the generator/modifier tools on the model database?
> We made the step-importer (without Express-Tool) to show that a
> interpreter-system can work also with acceptable speed.
I see.
> Greetings from Vienna,
Ah, greetings back there ;-)
CU Lars.
I just changed the subject line.
>> I think layers is a natural way to work in the world:
I agree if you think about "natural way to work" as sitting on a table
witch sketching paper. Which is rather productive and the reason that
using layers can work out.
>> A base layer is a house
>> A room is inside a house (sublayer)
>> a table is inside the room (sublayer)
>> and so... (sublayer)
Well, this is true in most drawings. If you enter a house you will not
perceive it as such. Layers are on top of each other, the furniture is
in the room, the room is surrounded by walls, etc. Things are not all 2d
and lying stacked as on my desktop ;-)
> So is it possible to define a format suitable for all purposes?
> If you look at AutoCAD it serves all but it is not very good in it (it
> is good 2D production drawings).
Well, the point is that if you use attributes that the user defines, it
is his decision how to use them. E.g. for a small 1:200 sketch that I
want to complete on a week-end, I may choose just to define an attribute
"layer" and use the system just as in Autocad. In a project with some
10.000s m2, that is supposed to be applied, I would like to work out a
more sophisticaded attribute system. Maybe even one allowing namespaces.
Meaning that office A can define a:materialName="wood_cherry", office b
has b:materialName"wood124" for it, occice c uses c:LAYER_MAT="Cherry"
and still one can map them and there are no conflicts between names. So
keep the user free to use attributes as layers or whatever he/she needs...
CU Lars.
Yes, you are correct, if only a identifier of the layer/attribute is
stored in the element and in a separate table the attributes or layers
are maintained it is the most flexible.
Should a table in which the layers or attributes are stored be part of
our file format, has gcad3d or step such structures?
Also i want to see our file format able to store data along with the
elements, just as xdata in AutoCAD or user data in MicroStation.
I don't care if this data (also for attribute/layer) information is in
a separate file if we have some kind of "include" system that
references external files (or database links).
Greetings,
Gertwin
In xml-world, you can store more or less random text in every entity. E.g.:
<cad:geomObject>
<cad:circle center="0,0,0" radius="1"/>
<larsAnnotations:description>
This is a circle.
</larsAnnotations:description>
<larsDrawingStucture:Option visible="TRUE"/>
</cad:geomObject>
So even if the application only knows the cad:-namespace, one could have
a way to make entities from unknown namespaces visible and editable. And
one could still enable the user to define such entities as a filter
criteria e.g. to toggle visibility, line color etc. BUT I do not know
how to do this in step - I do not know about their xml encoding, would a
step/xml validate if mixed namespaces?
> Should a table in which the layers or attributes are stored be part
> of our file format, has gcad3d or step such structures?
My idea would be to use namespaces to divide ONE document. It would
allow easier handling in non-file encodings, e.g. when stored in databases.
> Also i want to see our file format able to store data along with the
> elements, just as xdata in AutoCAD or user data in MicroStation.
This generic approach would allow whatever you want.
CU Lars.
Layer is the layerNr; 16 bit (65000 Layers max).
UserData is a pointer to a internal datafield, binary or ascii; size variable.
(*) = not yet implemented.
One elementType should be Applicationdata; eg to store csv-data; if there are
functions to correlate excel-fields with internal variables you can use
excel-sheets directly.
Not yet implemented, tested.
What do yot think about FreeCAD:
http://juergen-riegel.net/FreeCAD/Docu/index.php?title=Main_Page
(OpenCascade (Step, Iges, some Mesh-formats) + Qt + Pyton scripting;
native format is Opencascade(Ascii)+xml) ?
Franz
I think that opencascade (FREECAD), BRL-CAD, Archimedes and gcad3d have a
good future.
good demos of opencascade.
http://www.opencascade.org/showroom/demos/
The problem of opencascade we have 250 Megas of source code hehe. People
need several years to study all posiblities and many people working there.
Little by little...
They are like trolltech and Qt:
A company give CAD services and offers also all source code to the community.
COMPANY services:
http://www.opencascade.com
COMMUNITY source code:
http://www.opencascade.org
Great companies like BWM and EADS use this technology.
http://www.opencascade.com/customers/
Regards.
On Oct 10, 7:20 pm, Franz Reiter <franz.rei...@cadcam.co.at> wrote:
>
> Attributes (for each element) should be:
> ObjName(*), Linetyp/Color|Texture, Layer, Refsys, UserData(*),
> Dim, Pick, Hilite, Hide, Group,
> symbolic, visible, hidden.
>
> Layer is the layerNr; 16 bit (65000 Layers max).
> UserData is a pointer to a internal datafield, binary or ascii; size variable.
> (*) = not yet implemented.
>
> One elementType should be Applicationdata; eg to store csv-data; if there are
> functions to correlate excel-fields with internal variables you can use
> excel-sheets directly.
> Not yet implemented, tested.
>
Franz do you have this advices save in your documents?. I say this
because if not i save your advices for the future...
The next week we are in a magazine in spain called todolinux.
In spain people in the public administrations are forcing to give open
standards formats and the source code in all the works. In Spain if
you know we have the gvSIG software. This software is in the osgeo
cover like you can see:
http://www.gvsig.gva.es/index.php?id=gvsig&L=2
http://www.osgeo.org/
In this project there a lot of public money and they want build a free
CAD also,...
What is it happens with the standards opened up in your countries???.
In Spain we have many distributions public GNU/linux as linex,
lliurex, molinux,...
http://lliurex.net/home/ (Valencia)
http://www.linex.org/joomlaex/ (Extremadura)
http://www.guadalinex.org/ (Andalucia)
http://www.educa.madrid.org/web/madrid_linux/ (Madrid)
http://trisquel.uvigo.es/ (Vigo University)
http://pulsar.unizar.es/gluz/ (Zaragoza University)
http://www.linuxglobal.org/ (Cantabría)
http://tirwal.terueldigital.es/ (Aragon)
http://www.molinux.info/ (Castilla-La Mancha)
...
In engineering projects we have office documents and CAD documents. By
now i only see MS office (.doc,.xls,..) and DWG. I am totally sure
that from European Union they will need an open CAD format like open
office,...
Regards.
Hi Martin,
Maybe it is an idea to put the open_cad_format document chapters on
several discussion pages in this way it is more easy to discuss this
format.
I mean a page for the "introduction", for "Bases 2D-3D", for "Texts",
"Points", etc.
I think we should take this document as a start of our CAD file
format.
I think if we can provide a good/solid CAD engine library which can
and will be used by several drawing/CAD applications we have
succeeded.
It is impossible to define a standard format, it can become a sort of
standard if there is a large user base.
Greetings,
Gertwin
Franz do you have this advices save in your documents?. I say this because if not i save your advices for the future...Hi Martin, Maybe it is an idea to put the open_cad_format document chapters on several discussion pages in this way it is more easy to discuss this format. I mean a page for the "introduction", for "Bases 2D-3D", for "Texts", "Points", etc.
i think that i can put out the next gcad-version in 2 or 3 days; this version
will have clothoids and a gis-mesh (imported from Landmark.xml).
This could be a good base for discussions around gis-format.
With the next Version i had to modify the general fileformat;
(the file is now separated by SECTION's; new is "SECTION PTAB"
and "SECTION MESH"; subModels now are in "SECTION MODEL <modelName>"
was "MODEL <modelName>" before.
Since this is the basic format i have tried to explain it in the new
http://groups.google.com/group/open_cad_format/web/oca-page
What should be improved ?
Maybe you should also provide test-examples at opengeomap or open_cad_format;
- a point file (max 50000 points)
- a corresponding file with the breaklines (dxf ?)
- the mesh for the points and breaklines (Landmark.xml ?)
- the image from googleEarth (or photo from airplane) for that area
It should be possible to depict that in the new format ..
Franz
this is in the programming docs and in the inlude-files for the programfunctions. The programming-doc is not translated yet from german ..
i think that i can put out the next gcad-version in 2 or 3 days; this version will have clothoids and a gis-mesh (imported from Landmark.xml). This could be a good base for discussions around gis-format.
With the next Version i had to modify the general fileformat; (the file is now separated by SECTION's; new is "SECTION PTAB" and "SECTION MESH"; subModels now are in "SECTION MODEL <modelName>" was "MODEL <modelName>" before. Since this is the basic format i have tried to explain it in the new http://groups.google.com/group/open_cad_format/web/oca-page What should be improved ?
Maybe you should also provide test-examples at opengeomap or open_cad_format; - a point file (max 50000 points)
- a corresponding file with the breaklines (dxf ?) - the mesh for the points and breaklines (Landmark.xml ?)
- the image from googleEarth (or photo from airplane) for that area It should be possible to depict that in the new format .. Franz
> >this is in the programming docs and in the inlude-files for the
> >programfunctions. The programming-doc is not translated yet from german ..
>
> German is a difficult language for us,hehe :-[
For me German is not so difficult to understand, so post this document
if you want.
> Yes in your format we can respect don´t repeat primitives like GIS.
> (nodes->arcs->faces). I believe that people uses Microstation and
> Autocad many times before work in GIS software like Arcview or Geomedia.
> People from Geomedia for example can join entitites from a DXF file. In
> your format we don´t have this problem because you can use nodes, acs
> and faces like in GIS.
I think CAD must be our primary goal.
I think GIS is shifting from files to databases, files have a limited
area, there are always problems with f.i. roads which spans over
several drawings, in databases are not limited to a paper size.
If we have a CAD module in which we can define primitives (arcs,
lines, plines etc.) and we have a separate module which exports the
oca file into a database it is a better solution than try to build a
full GIS application on top of our oca format.
If the geometry is in a Oracle, MySQL or PostGIS database the geometry
can be viewed by many GIS applications.
Also geometry can be bound to administrative records like addresses,
maintenance schemas etc.
Greetings,
Gertwin
> >- a corresponding file with the breaklines (dxf ?)
Breaklines: what is the source before it comes to autocad ?
> >- the mesh for the points and breaklines (Landmark.xml ?)
how do you create the mesh ?
i think AutocadCivil can export the civil-mesh as Landmark.xml;
it can be exported also as traingle/quads (dxf .).
Franz
On Thursday 18 October 2007 18:27, Martín RV (OPENGeoMap) wrote: adMaybe you should also provide test-examples at opengeomap orwhat kind of point files??what is used most ? can the format vary ? Please describe.
- a corresponding file with the breaklines (dxf ?)Breaklines: what is the source before it comes to autocad ?
- the mesh for the points and breaklines (Landmark.xml ?)how do you create the mesh ? i think AutocadCivil can export the civil-mesh as Landmark.xml; it can be exported also as traingle/quads (dxf .). Franz
> breaklines are lines join to triangulation points.
> >Breaklines: what is the source before it comes to autocad ?
if the surveyor wants to have some points connected with lines;
eg if there is an edge in the landscape, you want to have exactly that edge in
the mesh. Are the points marked in the points file or is there an
extra-file ?
>for example in MDT-TC-IP you have several primitives and can export all
>to ascci files (not landXML).
so the software MDT-TC-IP reads the point-files, the breaklines (source ?
format ?) and does the triangulation ?
Protopo is a ARX-program and puts the mesh directly into Autocad ?
Franz
clear now; are there such files in the net ?
breaklines are lines join to triangulation points.Breaklines: what is the source before it comes to autocad ?if the surveyor wants to have some points connected with lines; eg if there is an edge in the landscape, you want to have exactly that edge in the mesh. Are the points marked in the points file or is there an extra-file ?
for example in MDT-TC-IP you have several primitives and can export all to ascci files (not landXML).so the software MDT-TC-IP reads the point-files, the breaklines (source ? format ?) and does the triangulation ?
Protopo is a ARX-program and puts the mesh directly into Autocad ?
Regards.