Petri Sakkinen

Dave's got it right, I'm afraid! PIOs are not supported in VS export and there are indeed various other deficiencies. However, on the PC platform - at least with FileMaker Pro - it is possible to put entire VW files as 'objects' into a database. In short, my revision system goes like this: - the actual current file never changes its name - for each revision, I make a copy of the file and explicitly rename that (sequential numbers) - for each issued [set of] drawing, a copy is also created - a 'revision symbol' is added for each revision to each affected titleblock present in the VW document (one VW document can be used to generate all drawings - or at least several drawings) - a 'project management' document uses workggroup referencing to keep track of drawings and revisions Anyway, this requires a bit of thinking & planning & implementation. Each practice has its own peculiar ways of operating, so I won't offer to sell my CAD management system to you, although it works very well for me. (I don't use the VWA system at all.)

Dave's got it right, I'm afraid! PIOs are not supported in VS export and there are indeed various other deficiencies. However, on the PC platform - at least with FileMaker Pro - it is possible to put entire VW files as 'objects' into a database. In short, my revision system goes like this: - the actual current file never changes its name - for each revision, I make a copy of the file and explicitly rename that (sequential numbers) - for each issued [set of] drawing, a copy is also created - a 'revision symbol' is added for each revision to each affected titleblock present in the VW document (one VW document can be used to generate all drawings - or at least several drawings) - a 'project management' document uses workggroup referencing to keep track of drawings and revisions Anyway, this requires a bit of thinking & planning & implementation. Each practice has its own peculiar ways of operating, so I won't offer to sell my CAD management system to you, although it works very well for me. (I don't use the VWA system at all.)

So it seems... Have to try to remember.

Compared with a mailing list, this tech board is a nightmare. Takes ages to do anything with it. One improvement would be to get emailed notifications for replies to your posts - this is pretty much the standard in many web-based discussion groups. Better still: dump the whole idea and establish a mailing list.

Compared with a mailing list, this tech board is a nightmare. Takes ages to do anything with it. One improvement would be to get emailed notifications for replies to your posts - this is pretty much the standard in many web-based discussion groups. Better still: dump the whole idea and establish a mailing list.

Being a long-standing advocate of ODBC, I certainly support this! However, even more important for the future would be to have a more generic object database / object structure. The NNA official line is that STEP is dead, but I don't really see much signs of that. Complex - yes, and therefore takes a long time and a huge effort, but STEP (or something that evolves from it) will be the real future of data exchange. Today, Google returned 381,000 references to 'STEP ISO STANDARD' query. VectorWorks had 20,500 references... (STEP, the Standard for the Exchange of Product Model Data, is a comprehensive ISO standard (ISO 10303) that describes how to represent and exchange digital product information.)

If it is a PIO, you must use the appropriate tool. If it is a symbol, you must either import it into your current document or open the one it resides in and select it with object browser.

PLC, Chris, Thanks for the hints. You verified my worst fear: that there is no easy way. Well, this morning I wrote a crude log wall PIO which may eventually find its way to VectorDepot. As comes to window openings, a PIO can be used in a solid subtraction to punch a hole in the log wall. The 2D view is lost, of course, but is 'recoverable' by ungrouping, so I can produce reasonable plans & elevations/perspectives, from two separate layers (not really an ideal 'design' solution. but I can live it for one project.) Fortunately, I don't need much detail. Basic round logs, although theoretically unsatisfactory, are fine for this exercise. Having a real log profile with all the details (don't ask more - I haven't done a log building in 25 years and the last ones were in a different language! No idea what the things are acalled in English!) appears to be quite an exercise. Oddly enough, on the VectorScript mailing list, someone today wanted hints on using components of a symbol definition to be used in a PIO - which was what I thought to use to get a better representation of the log profile.

This is getting ridiculous! Almost every time I try to use a VWA PIO, I find them next to useless. WindowWall (Straight): 1. Surely we should be able to have more than 3 horizontal dividers (mullions) in our designs 2. More often than not, the number of dividers is a result of max pane width or height OR number of desired panes. Well, in MY designs at least. 3. There does not seem to be any way of assigning a colour/class/texture to the frames/mullions only - if a colour/class is assigned to the entire PIO, this is reflected in the plan view so that the entire PIO is coloured. Has the creator of this little beauty ever had any lessons in architectural drafting? (Definitely not in architecture...) Why on earth did I spend the money to get VWA + Landmark? What did I get: a bunch of useless trade show pony PIOs, a DTM that does not work, a drawing management system I cannot/will not use, a few crappy utilities here and there. I was hoping to cease to be a 'meta-programmer,' but that did not eventuate. Well, I am now able to use some VectorScript procedured not available for mere mortals, but apart from that: wasted money! Well, as an Aussie, I did get a bunch of Julian's stuff. That JUST may justify the cost - but the rest of the world: tough luck, you are suckers.

I never thought I would do a log construction, but life is strange indeed. Can anyone give hints of the best modeling approach? I'm afraid I need line drawing type elevations as well as rendered images. I am thinking of writing a PIO, but I'd rather not as I hope this won't happen again so other techniques are preferable. It is not a huge project - some 350 sqm gross floor area - and I don't need to worry about working drawings as an old-fashioned drawingboard person will be engaged to do those.

As a person from a non-English speaking background, I dare to say this: Smart curser - oh, indeed! I curse it every day! Just bare with me... Sorry, sorry - no offence intended. Just to point out the problems of the English language - which never would have been chosen as the international 'lingua Franca' by anyone with even the basic understanding of communications theory. (not to mention the monk-Latin phrase 'lingua Franca')

Interesting! I do the exact opposite: the current version never changes its filename. Instead, every day (or even more often) I make a copy of the file in Finder and give it a sequential filename extension. This works OK in most cases, but WGR (WorkgGroup Referencing) occasionallly chooses to use the previous version on the Mac. (Because of the intricacies of the 'alias' system, I believe.) The copies go to a specific [versions] folder, so the scheme is workable also when using Windows. (copy -paste). (Fancy that - one of my QA proceddres is workable on Windows! But, hey - why don't we Mac users have the copy/paste option for entire files!!!!!!!) As comes to official issued versions, I also make a copy and rename it - not change the name of the current workimg version. BTW, the same approach is used in our firm for all files. We have some recalcitrant people who want to call the latest version by names like 'Design - current - latest - uptodate - May 2002' but, with the exception of my business/life partner, they are taken out and shot at dawn behind the cemetery stone wall.... Anyway, whatever works for you. Just be careful for the bullets.

Tried that, but because, apart from measuring, there is no way to know what the PIO thinks the landing should be, I didn't really get anywhere. Also, values less than zero are not accepted - the stair in question is almost 3 metres wide but the landing definitely should not be anything like that, rather, a '2 step' landing (not being a 'ceremonial' stair.) Also, using two single stairs is awkward because there is no numeric feedback from the PIO.

The point here is primarily about straight stairs with landing - and the length of the latter. In U-shapes, the landing width indeed normally is the same as stair width (although even then, I tend to be a bit more generous - even 200mm makes quite a difference.) Anyway, I politely disagree with your statement that stairs should only be designed to comply with a standard or a code - to me, they are a key element in the spatial experience. I believe my great compatriote Alvar Aalto would agree - his stairs are wonderful experiences! Well, maybe you don't subscribe to the pragmatic ideology, either - but the current PIO is based on this kind of thinking.

...especially as 'Code' can mean anything in the 85 (or whatever) countries VW is used in. However, the principle I elaborated on is, I think, rather widely used and promoted in at least the two books on stair design I happen to own. (Not to mention my professors at the Helsinki University of Technology in the 70s...)

Thanks, Matthew! I'll try that.

Let's add optional step on the top level to the wish list. (One of the modifications I made to the old PIO was to add a 3D locus to the top level 'theoretical' step edge. At least made it easier to figure out where the slab should go. Also, there were some outright bugs that I fixed: if I remember correctly, the hand rails were all over the place.)

Matthew, It may well have worked that way - I just haven't used much it because I found it rather disappointing. Thus, I wrote my own, which on the other hand is much more limited. I also made some improvements to the standard issue so that it was more useful - and if the NNA PIOs would not be protected, I would do it again. Anyway, this PIO behaviour is idiotic and, unfortunately, symptomatic. Many add-ons are rather half-hearted, looking good in trade shows and demonstrations, but failing miserably in real work. Having spent the money, I now obviously have to spend the time as well - either to more manual labour or to improve my own PIO.

I have just realised that changing the WIDTH of a 'straight stair with landing' changes the LENGTH of the landing. Hello? Has the person behind the PIO ever designed stairs? The landing length should of course be user-defined, perhaps prompting the correct convenient length. This, as all architects surely know, is based on the step length of the stair in question - 2 risers + 1 tread - so that similar steps take your foot to the end of the landing and the left-right rhythm is kept, ie. if you stepped on the landing with left foot, you proceed with right foot. OK, it is not a major problem if the rhythm changes if you take at least three steps on the landing. Anyway: STAIR WIDTH HAS NOTHING TO DO WITH LANDING LENGTH!!!!!! Right - back to using one's own stair PIO - but one womders why one spent the money to buy VWA in the first place.

The polylines came from a MapInfo cadastral base via DXF. Nothing was moved or rotated (would have defeated the purpose!)

Well, OpenGL seems to work in OS X, so I may just be able to survive. Fortunately, this is a small model.

Things are going from bad to worse. At present, my main Mac is at the hospital so I'm using my PB Ti 400 as the workstation. When I try to do an Open GL rendering, I only get noise - the entire screen painte with random colour patterns / lines. Reinstalled MacOS 9.2.2 - no change. Got latest (April 9) ATI drivers - no change. Reinstalled VW - no change. Reinstalled the new ATI drivers - no change. Changed Extension Manager settings to 'MacOS 9.2.2 All' - no change. Any ideas? I need to do a live presentation on Tuesday! Did not do a clean install (that takes a day! MacOS 9.2.2 G4 400640MB RAM 400,000K allocated for VW (tried with less, no change)

The protractor tool reports wrong angles. Polyline 1 Angle between edges 90.78 dgrees (verified visually, by dimensioning & by constraint), reported as 146.8 degrees Polyline 2 Angle between edges 90.21 dgrees (verified visually, by dimensioning & by constraint), reported as 57.0 degrees Some other polylines are reported OK (at least by visual assessment.) Copied and pasted to a new file, no change. No duplicate vertices. VW 9.5.0MacOS 9.2.2 (Sent also to bugsumbit, but also here to warn others who might rely on the tool.)

To my disappointment, INSYMBOL worksheet database criterion does not work for PIOs. I appreciate that there may be technical difficulties, but it would still be extremely useful to eg. count chairs in a Seating Layout as a part of the normal furniture schedule without exploding the PIO. With my own PIOs, I can perhaps devise workarounds (which, however, necessitate changes in record format structures), but with the Bonnet-Cannot-Be-Opened cars provided by NNA, even this is not possible.

The philosophical background of XREFs and WGR are different. The WGR mechanism often needs to be complemented by layer links, which you can move around. (And even duplicate - just unlock the beast first.) In most cases there should not be a need to reset the origin of the coordinate system (well, in VW 9, that is) - but in case you want to fo it (and I can indeed think of several scenarios), the location of the origin is easily (?) determined with VectorScript: PROCEDURE OriginCoords; { ? Petri Sakkinen 2002 } VAR xo, yo : REAL; BEGIN GETORIGIN(xo, yo); MESSAGE('x: ', xo, ' - y: ', yo); END; RUN(OriginCoords); With a little (??) bit of work you can perhaps write a script that writes the values into a text file and another one that moves everything in the other drawing by the same x/y values so you don't need to jot them down. Now, the problem we still have is this: If the origin has already been moved, we have no way of finding out what the PREVIOUS origin was. Maybe your script should in fact use two text files, compare the x/y values and move objects accordingly. Dunno - I'd rather establish a local coordinate system and stick to that.