P Retondo

Hi Katie, I saved this file hurriedly at the end of the day. I think I shut down VW and clicked on the "Save file" dialog box, which saves it directly to the office server via Ethernet. I probably then closed down Outlook and Windows Explorer, then shut down. I was in too much of a hurry to remember any detail. VW 10.1 WinXP/Server = Win2K

Thanks for the input. Jim, have you ever developed any evidence about what caused the corruption? I've never had a file that was quite so large, and have never used NURBS tools before, so those are two of my main suspects. Does anyone know the meaning of the error code? I would like to request that NNA programmers develop a bulletproof file save procedure, that would do error-checking to ensure that the file is saved correctly before the save file operation is considered complete. [ 04-11-2003, 03:28 PM: Message edited by: P Retondo ]

BAZ, zoom line thickness mode is intended to allow the user to see graphically the differences between line thicknesses. I use it only to check my line weights. Turning that mode off allows one to work with greatest accuracy, as all lines are displayed with minimal thickness in all magnifications. Trying to work in close with line thicknesses zoomed is like trying to look at a drawing with a microscope - as you say, if your magnification is high enough all you see is a screen of black. If it worked as you suggest, we would not have WYSIWYG capability. [ 04-10-2003, 09:21 PM: Message edited by: P Retondo ]

I just attempted to open a major project file, and the file is damaged (error message: "-39 End of file reached during Read Attempt to read file FSREAD"). The recover file is missing an entire floor of walls, none of the saved sheets work, and there are 2 layers named Layer-1, the second of which contains a strange ungrouped window or door object. Even if I can recover this project from yesterday's backup, I am faced with losing at least a day's work. More importantly, I have no idea what caused this problem to occur. Please advise. This file is approximately 25 MB. VWA 10.1 WinXP P4 2.2 Gh 1 G RAM [ 04-10-2003, 09:23 PM: Message edited by: P Retondo ]

thx, you cannot have too much processor speed for 3D modelling. I use a P4 2.2 Gh CPU with 1G RAM (Win XP or 2000). I've used up to 80% of available RAM for some operations, and if I could double my processor speed, I would. The system I have is adequate for a reasonably complex building model. As you add curving objects, the number of facets increases very rapidly -> a lot of calculations.

Zero, I also find that VW slows down after a memory intensive operation, and it helps to then quit and restart the program. It's been a while, but it may also help to reboot your computer to reset all the memory and page files and caches, etc.

I note that the author of "Clone PIO" is none other than Robert Anderson. Robert, VectorDepot says this is a VW 9.5 script, does it work with the VW 10 PIOs? Also, with respect to the updating of wall thickness, if you like I can send you a file with a door in a wall that will not update its "wall bits" thickness when the wall is made thicker.

Much appreciate knowing that the bugs are being worked on. BTW, out of curiousity, is "HTH" a person's intials or does it stand for something?

I think your biggest problem is not enough RAM. See if you can get your RAM up to 768, or better, 1G.

Re: the advantages of symbols, it's the fastest way to insert multiples of a given window or door. If we had a way to use the eyedropper tool to apply all settings from one instance of a PIO to another, this advantage of using a symbol would be obviated. The other thing that's great about symbols is that you can add custom muntin patterns to the 3D object, and that will appear in every instance of the symbol. This greatly reduces the time for putting such detail into a 3D model.

Robert, A couple of comments: if you change the thickness of a wall containing a PIO converted to a symbol, the jamb depth is modified, but not the symbol's representation of the wall in the area between casings. So, it appears to me that a symbol can "know" when the wall thickness has changed and make some appropriate adjustments. The code could be re-written to fix the "bits of wall" as well. Although the reasons you advance for having window and door symbols create their own little piece of a wall may be valid, I feel they are far outweighed by the disadvantages. Besides what has been mentioned here with regard to cavities and that extraneous line crossing the wall at one side of the window, if your interior trim is a different width from exterior trim, there is a gap in the 3D wall when viewed from the narrow-trim side. I think all of these are unacceptable flaws, and we should work to fix them. In the old window/door tools, there was an "implode loci" option, which I never understood until someone on this board explained that it provided the functionality we are asking for - it made it so that the wall was cut at the frame, rather than at the outer extent of casings. I think we are arguing that this is the preferred way to go, based on the unresolved problems and complications we are seeing.

By the way, in the example cited above where the wall thickness does not update properly with a change in wall thickness, the window or door jamb DOES!

My setting is at "75%" of always. I have tried various settings on that scale, and none makes any difference. The only thing that works is finding that small area of the screen where my initial click on a move retains the rendered model. By the way, I can then release the mouse button to stop and click anywhere to start motion, and the model is always retained. But when I go back to some other view, and then return to a rendered view, I have to find that special location again to get the tool to work properly.

If I create a window from the "Window" insertion tool, then change the thickness of the wall (requiring that I reposition my wall cavity lines!), the portions of the wall between the casings (that the PIO creates) do not update. I.e., the cavity lines are incorrect. Also, if I convert a PIO window to a symbol and go through the same process, I cannot get even the wall thickness to update (in that portion of the wall between the casings). The PIO tools need to be rewritten so that this problem is taken care of - or better yet, so that there is no need in the first place to create that little chunk of wall between the jamb and extent of casings!

As far as I am aware, you need to create an extruded solid to do this. Obviously this is not a parametric object. But you can create a wall of zero height to go along with this, so that in plan view you have a parametric object. If you have windows and openings, I'm afraid you will have to construct everything in 3D. It might be possible to construct a custom window symbol that has a tilted 3D frame and other tilted parts. [ 04-06-2003, 07:59 PM: Message edited by: P Retondo ]

If you want the ceiling to be a different color from the floor above, I create an extrusion the thickness of gyp board and place it at the desired height. I generally create a special layer for ceilings to do this, but the object could just as easily live on the slab layer.

quote: Originally posted by iboymatt: what's strange is eyedropper tool already picks up some wall attributes but not others. it picks up the cavities, but not width. Not so strange that the tool behaves this way. Actually, it is convenient and I would guess was intentionally designed. You can quickly make your 2x4 and 2x6 walls look the same with the eyedropper tool, without worrying about having your 2x6's change to 2x4's!

I wonder if anyone else has experienced this problem: in OpenGL render mode, the walkthrough tool has in the past maintained rendering while moving around the buiding/object. I find that with v 10, I have to click on one particular point in the rendering, otherwise as soon as I commence movement the rendering reverts to wireframe and stays that way until I stop. If I happen to click in that "sweet spot" near the center, I can move around at will and stay in OpenGL rendering. Is there any explanation for this behavior? VWA 10.1 WinXP/2K P4 2.2Gh 1G RAM GeForce 4 Ti 200 128mb

Runtime, I can confirm that with a system almost identical to yours my export image function works fine with VWA 10.1.

Peter, look at the top of your screen with the 2D selection tool active. On the right of the "mode" bar you will see a little icon that looks like a pair of tables, one faint, with an arrow on the right. Clicking this on and off enables the ability to move a connected wall, dragging all wall connections with your move. This only works when you drag with the mouse. It doesn't work with the move command.

Katie, what I guess I was trying to say was that I hoped OLE would work both ways, so that a VW worksheet could be linked to Word or Excel. What I understand by OLE is that an object could be inserted into a VW document so that, when it is changed, it can be automatically re-inserted in updated fashion.

I'm bringing this topic to the front again since it hasn't been addressed. I'm surprised no one at NNA has requested a file showing this potential bug.

Dan, et al., jan15 turned me on to the easy way to deselect all objects: hit "x" twice. I use it all the time. Just be thankful that in VW, when you select an object, previously selected objects are deselected automatically, unlike in AutoCAD, where you have to hit "esc" twice before you select the object you want. I have to agree with others that leaving the object selected would be my preference, so that I don't have to reselect the object when performing multiple operations. It strikes me, however, that there could be a program preference to toggle this behavior, thus making everyone happy! I have posted a request for a single keystroke deselect command. I would like to add that this command should deselect all objects on all layers, active or not. "x+x" leaves objects selected in non-visible and inactive layers. [ 03-31-2003, 12:42 PM: Message edited by: P Retondo ]

I created a DTM, but forgot to set a boundary first. I had already created and modified a proposed site model object. I then added a boundary object, attempted to regenerate the DTM and the program crashed. I then deleted the DTM symbols and used the "Delete" command to delete the DTM internal info (boundary object and modifier objects classes set to "none"). When I attempted to re-create the DTM, I received an error message ("A miscellaneous error has occurred error codes 20,1"). I get the same message when I attempt to check the 3D polys. I have run into this before. If I completely rebuild the file, recreating all layers and objects, I can then generate the DTM. It appears to me that the "delete" command is not working properly. Is this a known situation? Is there an easy workaround?

Based on what I've been reading here, and my own experience, I would guess that the two most popular addtions to VW would be these: 1) OLE capability, allowing us to embed linked Excel and Work files into VW. This would take care of all the complaints about text editing. If we could export Worksheet data to an Excel file, perhaps that would take care of all the Worksheet formatting problems people experience. 2) Automated generation of 2D sections from a 3D model. This would update the outline of all solid objects in a saved section view whenever the model is changed. All you need to do is come up with some algorithms that can find the surface, or outline, of all solid objects cut by the section plane. I think this is the easiest of the many requests for 2D-3D interactivity you have received. If it works seamlessly, and we can assign classes with line weights to the elements of the original section that would be preseved on update, it would be a very exciting and time-saving capability.