Chad McNeely

Have you added wall peaks, or fit walls to roof? I've had several instances where the walls became corrupted, such that moving a vertex up caused the actual wall corner to move the corresponding distance downward, below the z=0 baseline. The result was "figure 8" shaped walls, and the only remedy was to delete the wall and redraw it. The walls frequently became corrupted again, so the exercise became routine...

Pat, I had tried your workaround- as soon as the updated new PIO is pasted into the symbol space, its wraps disappear. I also tried creating a new symbol, saving it as the same name as the old, to "write over" and replace the old, but that just corrupted the old ones... I know the PIO needs to reach out and feel the wall to get the wraps correct, but it doesn't seem necessary for them to go away when things unrelated to the wrap are changed. Mike, Indeed. I rarely use wraps at openings, or even walls with components to wrap, so making PIO's into symbols is generally an efficient way to spread doors and windows around, edit them, and manage extra bits of geometry I might need to add. Except when it doesn't work. Bug report on the way...

I want to wrap the finish components of a wall into the jamb. I inserted the door PIO, adjusted the settings as needed, it looks fine. Made a symbol of the PIO for use elsewhere, still looks fine. Spread the symbol all over the project, then needed to make a change. Ooops. Entering the symbol to make any change including color, size, or class eliminates the wall wraps. The only way to edit the symbol instances seems to be to create a new symbol and manually replace each instance of the old one.

Vectorworks>Preferences>3d>3d Conversion resolution: Set to high or very high. Removing the hidden line facet lines is controlled by the setting in your later post. Ths short answer is to use them as you see fit... I don't overlap them because it just bugs me to do so, and walls will end up reporting the additional height in the OIP. With the ability to texture floor edges in '09, it makes it even easier to run floors to the building edge. For interior walls setting on stepped floors (I think that might have been the case in your example?), you can use the "Fit walls to roof" command, and fit the bottoms of your walls to the floor geometry below.

Class visibility issue? Scale issue, meaning the scale of the symbol won't fit the polygon, like say on a sheet layer (1:1) viewport? The symbol contains 3d stuff? (although I think the command gives an "invalid object" warning for those, and doesn't go through the script)

Easy, Pat! While his rant is a little off base, there's a nugget of truth to it... With 2008, when you make a symbol of a PIO and have just one trim selected, say Exterior, and you flip the door because you want outswing instead of inswing, for instance, the trim flips with the symbol/PIO (as I would expect), but the texture remains defined as though the door wasn't flipped (if you're lucky) or gets lost entirely, frequently. Several releases ago, I could just manually reassign the classes to the other trim to get the look needed, but then PIO's were given some additional "intelligence" to make them relate to the wall's exterior or interior face. Pity the chump (me) that always made my window and door PIO's into symbols before insertion! Simply, there's been a longstanding disconnect between the interior/exterior textures set in the View tab of the Settings window, and the geometric size and location indicated in the OIP. I haven't played with it in a new working file in '09 yet, but am hopeful based upon a simple test file...

Welcome to the world of hybrid objects. Once you get used to it and understand it, you'll appreciate it. Many VW objects are "already" hybrid- walls, roofs, floors, window and door PIO's, etc. When you build stuff yourself, you need to make it hybrid if you want it to show other than wireframe. Generally the easiest way to do this is to create the 3d stuff, and then select it all and convert copy to lines. Edit the lines to show the appearance you want, including composing/drawing over with filled polys. Now select the 2d stuff, the 3d stuff, and create symbol. In a 3d view, you'll see the 3d stuff, and in a 2d view, you'll see the 2d stuff. That's hybrid.

I bug-submitted a similar problem back in the middle of the VW2008 cycle. I recreated that file in VW2009 -it was essentially five 3-point 3d polys with ascending and descending elevations, nearly the simplest possible geometry for a DTM, which VW2008 mangled. 2009 built an identically mangled DTM.

Ever hopeful, I tried some Site Model geometry in both VWA 2008/3 and VWA2009/1. The geometry comes from a current project, and consists of several low vertex count 3d polys, and a couple of 3d loci at the corners. The existing site is graded already with some steep cuts and a pad area, so no modifiers have been used yet- I just need a reasonable approximation the dirt that's there. In 2008, the DTM displays many of the 2d contours incorrectly despite adding lots of additional 3d polys in the offending areas- the DTM irrationally connects points across intervening geometry in the classic "spilling dirt" pattern we've come to enjoy lo these many years. But at least the 2d contours were as consistently erratic as the 3d contours... In 2009, the Site Model created from the same geometry has vastly improved 2d contours, almost perfect even. But the 3d stuff is just hilarious. "Spilled dirt" is too kind- this is a landslide...

Personally, I would prefer if your posts were just helpful, followed the forum rules: http://techboard.nemetschek.net/ubbthreads/ubbthreads.php?ubb=faq#rules ...and you limited your adverts to your signature. But flagging them as adverts ahead of time is a close second...

Another workaround for somebody else that had the problem was to add a wall peak along the slope (in line, so it doesn't change the shape). Try it in a few locations along the rake- I think it needed to be near the offending texture wings.

Is your class visibilty setting set to something other than "show/snap/modify"? Sometimes, the stuff on the bottom gets selected "through" the stuff on top if the stuff on top is not in the active class, when one of the more limiting class visibily options is used. Or if the stuff on top is a group or symbol that combines visible and invisible classes in the container/contents. Otherwise, sounds like something goofy with VW. Maybe try adding a new layer for the floor if all else fails.

Could you flag your posts as ads when you "use the forums to post or transmit advertisements or commercial solicitations"? I like the helpful posts you provide, but sheesh, enough of the ad spamming!

Classes do not have an effect on stacking order. Layer stacking order can be adjusted with stacking order overrides in viewports. Within individual design layers you can change the stacking order of objects using the "send" commands, which will then be maintained for that layer's occurances throughout the drawing.

type "-.625" in the x (for a horizontal line), y (for a vertical line), or length (in polar mode, for a diagonal line) field of the OIP, after the amount already shown there? In your example, the x, y, or L field would read: 10'0"-.625 There's no way to use the cursor cues to achieve this that I know of, if that's what you're after (other than typing the new completed length, or the arithmatic that gets there). Other ways to achieve this are: -->selecting the 2d reshape tool, lassooing the end point, and using the move command. -->drawing a line perpendicular to the line in question, double-clicking your key for the offset tool to bring up the options, entering .625 or 5/8, offsetting the line, and using it to trim the "main" line, and then deleting the trimmed line and the offset line(s). -->draw a circle or arc starting with the center at the lines endpoint, in the OIP enter .625 for the radius, and use the circle or arc to trim the line. Delete the trimmed end and circle or arc when finished. ...and on and on.

If you only need to see the elevation of the staircase, make a separate viewport for it, set the view to the elevation you want, crop as needed, and copy/create your 2d notes that relate to the staircase into that viewport's annotations.

Certainly not intuitive at first, definitely. Open the two tabs at the left of the screen, and the Texture Maps and Hatch Patterns windows from the buttons along the top. One of the Left side tabs has a button to select "Vectorworks" as the file format. Also note what the target path of any saved files will be. That done, once you click on one of the textures in the Texture Maps window, the Texture Finder (where you previously selected the VW file option) will display the stone family and color. Sort through them to find the ones you want, and highlight the ones you want with a dot in the Hatch Patterns window. Once you have the ones you want (or all of them!) selected, in the Hatch Patterns window click the "save to disk" icon. Go find the files where the Target Path said they would be. The .mcd files created will have hatch patterns of the selected stone style created, images of the stone style in each of the colors available, as well as textures (in the Resource Browser) of the stone style in the available colors. In VW, use File>Open to bring the .mcd files into your version of VW, and thereby make the resources available to your current project. HTH.

I have no idea what the area reported in OIP for the Site Model refers to, I've never dug around to see. While the data might be suspicious already, the area of the Site Model isn't necessarily related to the cut or fill volume of the area you modify. You can ty a simple test, say a small pad or site modifying 3d poly, with a tight fence to get a nearly cubic excavation, and see it the cut volume nearly matches the cube volume. I still can't believe after all these DTM releases that this wouldn't at least occasionally work...

Sounds like it would be possible with two DTM surfaces, say existing for the bottom of the fill and proposed for the bottom of paving/top of fill. The guy in charge of Landmark says that 2009's cut and fill calculations are now reliable. I think it would be a good idea to manually check at least a subset of the volumes, just to satisfy yourself that VW wasn't being silly. If the volumes don't compute correctly, the two-DTM approach might still be useful. Once satisfied that each DTM surface is reasonably representing the surfaces in your survey, you can manually cut a bunch of sections through them, trace the area between the DTMs, and multiply the area of each traced poly by half the distance to its neighbors, or use Simpson's Rule, to get an interpolated volume. With some classing of the area polys used, a worksheet might be a little helpful.

The point was, in the post I quoted, you seemed to be implying that the volume reported was incorrect. If you need the wall's footprint area, you can use the wall's volume ( a cell with "=VOLUME") divided by the wall's average height (a cell with "=WALLAVERAGEHEIGHT"). I used an additional column that said something like "=D3/B3". The rest is just factors applied to those values as you need. This seems a lot simpler than you all are making it out to be, but I remain a spreadsheet ignoramus despite occasional efforts to beat additional knowledge into my head.

Looks right to me...

Measuring a roof as a rise/run instead of as an angle is the key. The valley will be a line whose x:y dimensions are proportional to the roof faces' rise:rise. So in your example, two same-pitched roof faces each have the same rise, and a line with equal x and y happens to be a 45 degree angle (between 90 degree bearing lines). If a 6:12 roof runs into a 12:12 roof, it takes the 6:12 roof half as far to match the height of the 12:12 slope, so the valley will have x:y dimensions that are 2:1 or 1:2 depending on which way things meet. Another way to skin the cat is to draw (just 2d) lines parallel to the bearing line, such that they are, say, one foot above the bearing height. So at 3:12, this line is 4 feet in, at 6:12 it is 2 feet in, etc. When you connect a line from the intersection of the bearing lines to the intersection of these lines of equal height, you have an accurate hip or valley angle. This works for any pitch, and at any intersection angle.

For planar surfaces, top view, convert copy to lines, hidden line render. Where the eave and ridge lines stop are the end points of the valley, generally. But you should (IMO) have already drawn these lines in 2d before starting to build roof faces. You then have a guide for creating the roof face polys, a reality check on the VW output, and a way of keeping your head in the game, so to speek. I believe I can actually draw a roof faster this way, too, not counting the simple roofs where "Create roof" actually works. The goal is actively and accurately placing the valleys and such, versus throwing up geometry and letting VW mash it together for you, and then passively using the results to try again, and again.

By "flare" do you mean a mitered return? If so then... As you have learned, VW won't let you use extrude along path to do this in one step. You can make a path that comes out from the wall, across the wall and back to the wall; or you can have a path that has some spring to it when viewed from the front. But not both together. Probably the reason is that the joint between curved profiles and straight returns is fairly complex. The miter surface is a curve that takes some figgerin to get right. Your example looks like a gentle curve though, and should be doable by drawing a section plane, inclining it to match the slope at the bottom of the curve, and then doing something similar to the short pieces used for the returns. Example attached.

Why? No... But adjust the end points of the bearing line and they will reliably snap.