Petri

Ooh, CADintosh is so expensive! A whopping USD 32 - can you believe it! Clever people save heaps of money by buying CorelDraw: it costs only USD 400...

This is curious: When I create a Color Palette in VW, it is editable. When I create one externally, it is not. Curiouser and curiouser: When I copy data from the externally-created into an internally-created one, it sometimes is editable, sometimes not. It seems that if I only paste the actual colour descriptions, not the header & "closer", it is. (The header used in generating the non-editable files was copied from a custom palette.) All the files are in user's Application Support, with R/W permissions. Where in the manuals or help is this documented?

Well, obviously this was not supposed to be easy. What might if (!cmyk['C']) cmyk['C'] = 0; if (!cmyk['M']) cmyk['M'] = 0; if (!cmyk['Y']) cmyk['Y'] = 0; if (!cmyk['K']) cmyk['K'] = 0; mean?

Well, the user-friendly Plug-in Editor does that... Sort of... 1. The Plug-in Editor not only provides the information, but also enables one to change it. 2. The Report Generator gives you column headers and formulae for each object. The latter should be quite adequate when investigating. In order to know what a tool/object can do, you do need to place an instance in the document. A menu item that would list the details of every tool might be interesting, but - at the end of the day - somewhat useless.

I create a colour with CMYK values 0% 6% 26% 40% I close the dialog. I open it again and VW has decided that the colour was not good and has changed the values to 37% 33% 48% 2% What's the point in this new system? Looks suspiciosly like the old one - substitution of colours. EDIT Here's the pudding, as requested

Good question. Some 10 years ago the Benelux-distributor Design Express wrote an SQL-extension to MiniCAD, but I gather it was not considered as useful at Diehl Graphsoft. At present, there is no mechanism for accessing external databases.

I'm not saying that the difference would be big, but the "flatter" the shapes, the bigger it gets. May well be below measurement accuracy in most cases, but the maths is different and the squashed circle is not an ellipse.

There is a clumsy, partial workaround: Create a site model. Define its settings. Delete it. Save as template. New models will use the settings. (Not thoroughly tested.) It appears that the dreaded "There already is a site model in the file" no longer haunts us, so it's possible to do it like this. This still does not make it possible to transfer settings to existing models, so something should be done to the mechanism.

You are correct, too. This is a somewhat problematic issue with no ideal solution - especially as it is too easy to flip a door. In full "product modeling" the question becomes particularly urgent - the data record is not, AFAIK, a part of IFC export.

This is exactly as it should be. Sorry, I meant to say "as it is". There is a database, but the Plant Database does not actually access the database. It just creates a bunch of text files they call "the database". The entire system is a sorry excuse for a database. You can open the "actual" database, too, and edit it. Unless you tell VW to regenerate the text files, the plant tool has no idea of your editing efforts. (Mind you, it is very easy to end up with any number of versions of the said text files, but without any idea which version the tool actually uses. My last count was 8 data sets in different locations.) A total dud.

The computer system of the construction firm does not eyeball drawings. It just places orders for doors as specified; a door meaning a full assembly, ready to be installed. No perspective needed.

And after all this effort, you can't distinguish left-handed doors from right-handed ones? Fire-rated doors from others? (Some of which may be of steel construction, with steel jambs.) Living is easy in the fair mathematically-challenged country of NC!

Could be. Primarily your problem is that you don't read the answers to your questions.

About spirals: http://en.wikipedia.org/wiki/Image:Confusion_of_Tongues.png

Love the idea... How about transparent leather in a Chesterfield? Indeed. And now with 24-bit colour everywhere, one should be able get quite reasonable results. I'm anything but a rendering expert, but I'd think that with a set of textures using Object Color and - 2 or 3 levels of reflectivity - 2 or 3 levels of transparency - 2 or 3 types of bump in various combinations, one could get reasonable results of practically anything. Just an untested theory, though.

This capability does exist: when you say "Report", you get the alternative names as column headers while the real names are in the formulae.

One of the better 3D features of VW is that you can draw a 2D shape in any 3D view and then extrude or sweep it. If this possibility would be taken away only because some architects want 2D shapes (uselessly at z=0) to be shown in 3D, we'd take a huge leap backwards. Since only the plan view is 2D, you are indeed advocating a forced switch to plan mode. If someone is able to create a massing model of a building, surely he or she can either draw a few 3D polygons or convert 2D shapes to those.

Downside? Well, let's just say that a certain Mr. Euclid has shown that there is a downside. Let's assume you have two 3D lines created in different orientations. How would you join them? I don't think there is an NNA engineer working on proving Euclid wrong.

2D is always 2D. Once more: if you want everything in 3D, use only 3D-objects. Draw them on a plane. Do not use any 2D objects. What's the big problem here? It might be handy to have clairvoyant 2D objects that would behave like 3D objects in the way we would like them to, but that is not going to happen. Oh yes, about "all other programs": ArchiCAD behaves exactly like VectorWorks in this respect, except that it can't even convert 2D lines etc into 3D.

If an innocent user changes the name of a layer or a class, a query may be ruined. When an average user sees a message saying something like "the criteria are invalid and you have to change them in VectorScript" he or she will simply panic. No normal human being can deal with an error in something like =DATABASE(((L='SURFACE MATERIALS') & ((T=ARC) | (T=POLY) | (T=POLYLINE) | (T=RECT)) & (C<>'MATERIAL-Asphalt') & (C<>'MATERIAL-Concrete') & (C<>'MATERIAL-Stone'))) I'm not sure how this could be resolved. In FileMaker Pro, all names of tables, fields, layouts, scripts etc. a user (or even a developer) sees are "aliases"; internally they are, I gather, persistent numbers. VW classes are numbered, but the numbers are not persistent - and probably could not be: classes can be added by copying and pasting objects or by any other simple action. Maybe we could have aliases (alternative layer & class names) that would be used in the normal user interface, including queries & reporting, in the same fashion as we have alternative parameter names in PIOs? PIO reports, by the way, should be able to use those alternative names. Localisation of objects is rather pointless if users need to figure out what the "real" name of a parameter is. In this respect, alternative names for data record fields would also be welcome.

Compliance with SILLY_CONVENTIONS may be rational.

In simple terms: 2D objects do not exist in any 3D coordinate system. I'm pretty sure that this concept is beyond your comprehension: it must be difficult to realise that 2 is not 3. Now, I do appreciate that with your whopping four years of experience and little, if any, background knowledge, logical & mathematical concepts are difficult. This, as such, would not be stupidity. Your unique stupidity arises from your unwillingness to learn even the basics of spatial geometry. Just stick to drafting.

MESSAGE(GETTYPE(GETOBJECT('2/A6.3'))); The word "digital" in "digitalmechanics" seems to carry the literal value of zero for comprehension. Zero fingers?

Geez... Hey, read a book on spatial geometry. Enrol on a course of Extremely Elementary Geometry & Utterly Basic Coordinate Systems at your local, friendly Mechanics Institute. Hire a maths teacher. ANYTHING! (It would not help, though. How about a book about Simple Arithmetics? One finger, two fingers, three fingers - six!)

This explains what I saw the other day!