Jump to content
  • 2

Is top/plan view an evolutionary dead end?


line-weight

Question

So the basic idea of top/plan view is that it's a kind of symbolic representation of a 3d reality. It's not quite the same as a horizontal section, because drawing convention has it (mostly for good reasons) that in plan view some architectural elements like stairs or doors are shown in a way that's not quite a literal projection of what those things look like from "above".

In vectorworks we can (now) create a plan view of sorts by making a horizontal section.

Or we can go with the "top/plan" view which (in theory) creates much the same but with certain architectural elements show in the proper symbolic way. In reality this doesn't actually work though, as soon as you start dealing with anything a bit complicated. We're given tools like the Auto Hybrid to partly deal with this - effectively the Auto Hybrids let us say "this part of the 2D drawing shall be generated in much the same way as a horizontal section is". So what we end up with is a kind of mashup, where parts of the drawing are generated as a literal horizontal section, and parts are generated as 2D symbols which aren't literal projections. And these bits don't really join together properly, and there are all sorts of reasons why having certain things in these containers makes everything a bit difficult. So it seems basically inevitable that all sorts of things have to be patched up in 2D layers in order to create something presentable.

Essentially in my opinion, "top/plan" view is a mess and just doesn't really work. I don't really see how it can ever work properly in its current form.

Why can't we have a plan view that takes, as its starting point, geometry that's generated by literally cutting the 3D model. Then the symbolic elements like doors and so on are inserted into that in an intelligent way. In my mind it could be as simple (in principle) as a tick box in a viewport setting. So we just have one "plan view" which we can toggle between (a) a literal horizontal section of the 3D model and (b) the same but with things like doors replaced with conventional architectural symbols.

At the moment it seems to work in a completely backwards way - we start off with a 2D drawing that kind of generates the 3D stuff (but not very well) and then we go into 3D and draw all the other bits in a way that either feeds back to the 2D drawing in an unsatisfactory way, or which we just give up on drawing in such a way that will generate things properly in 2D, and chunks of the information end up getting drawn in parallel, once for the 3D model and once for the 2D output.

This just doesn't encourage model-centric drawing, which, I think, is what we're all trying to move to, isn't it?

So, anyway, ultimately my question is whether, in the long term, Vectorworks will move to something more like I describe above, or is the the current "top/plan" view approach here to stay?

Edited by col37400
  • Like 1
Link to comment

Recommended Posts

  • 0

I think you might get some push back from some folks who will tell you they just want to create 2D Plans, Elevations & Sections.

I'm not sure I agree with you when you say:

Essentially in my opinion, "top/plan" view is a mess and just doesn't really work. I don't really see how it can ever work properly in its current form.

I find the representation of TOP/PLAN really a good way to visualize, and communicates well to most consultants and trades for the vast majority of our work. My drafting textbook from school (first published in 1965!) tells me that a PLAN VIEW is like a slice cut midway through a building, generally speaking at 4' (1220).

Where I do agree is it would be very useful to have the ability to use the CLIP CUBE more creatively (CLIP 3D POLY?) to illustrate complex connections & assemblies without having to manage tons of CLASSES & Layers.

Link to comment
  • 0

An interesting corollary to this is a current discussion to include simplified or schematic Top/Plan like representation of objects in other orthogonal and even non-Cartesian views, so Top & Top/Plan, Front & Front/Elevation, Right & Right Elevation Cetc...

Far from going away I see it being integrated more deeply where the entire model becomes auto-hybrid with objects determining their individual 2d component on the fly based on reconfigured, by type, by object settings.

Link to comment
  • 0
I think you might get some push back from some folks who will tell you they just want to create 2D Plans, Elevations & Sections.

I'm not sure I agree with you when you say:

Essentially in my opinion, "top/plan" view is a mess and just doesn't really work. I don't really see how it can ever work properly in its current form.

I find the representation of TOP/PLAN really a good way to visualize, and communicates well to most consultants and trades for the vast majority of our work. My drafting textbook from school (first published in 1965!) tells me that a PLAN VIEW is like a slice cut midway through a building, generally speaking at 4' (1220).

Where I do agree is it would be very useful to have the ability to use the CLIP CUBE more creatively (CLIP 3D POLY?) to illustrate complex connections & assemblies without having to manage tons of CLASSES & Layers.

To clarify, I'm not arguing against the use of floorplans generally to explain a building. I'm talking about Vectorwork's implementation of the "top/plan" view.

Link to comment
  • 0

col37400, I agree with you 100%. Top/Plan makes a mess of presenting a decent plan view, but a 3D view rendering is no better.

I haven't tried a plan section, but maybe I will now - but I do see a lot of "fixing-up" on the viewport to show all the elements I need to show.

Just for information, my work involves custom kitchens and commercial millwork.

Link to comment
  • 0

Top/Plan is incredibly useful and highly tuned to a large amount of work. It will be around as long as we use paper plans.

If you don't like it, work in 3D, cut views or something else. Clip cube and Auto Hybrid need years of development to be anywhere near as polished as Plan. Perhaps an Auto Hybrid view will make things work for you. I'm very open to ideas for new ways to work, but don't take my main one away from me.

  • Like 1
Link to comment
  • 0
Clip cube and Auto Hybrid need years of development to be anywhere near as polished as Plan.

This is the problem though - there are all sorts of things that are simply impossible* to get drawn properly in top/plan view - Auto Hybrid is an attempt to fill in those gaps but as you say it's just not polished enough.

*When I say impossible, I don't mean it's impossible to draw anything in 2D, I mean it's impossible to draw so that they are correct both in 3D and in top/plan view, without having to do manual 2D patch-up work to fix things.

Link to comment
  • 0
col37400, I agree with you 100%. Top/Plan makes a mess of presenting a decent plan view, but a 3D view rendering is no better.

I haven't tried a plan section, but maybe I will now - but I do see a lot of "fixing-up" on the viewport to show all the elements I need to show.

I guess the question is which needs the least "fixing-up" - a top/plan view output or a plan section output.

The answer is probably different for different people depending what kind of work they do.

Link to comment
  • 0

Top Plan was fine for 2D drawing with manual stacking orders.

Even fine in 3D times for 2D people drawing 2.1D by using 3D Walls/Windows/Doors

without caring about Z heights.

For me it does hardly work in 3D after you had to leave the 3D BIM Tools and do

some normal Solids geometry like Extrudes which aren't represented the same way.

It is disturbing that after drawing your Walls and such, you insert a floor and

everything will be hidden behind that Floor, although it is clear that all other geometry

including tables and stairs are exceeding the Floors Z height.

I think that the Top Plan for the Drawing Design Layer has to be handled separate from

a Top Plan Mode for representation Viewports.

Generally I prefer to draw in Wireframe Plans.

Sometimes it is nice to have some parts like a few walls in colored mode though to

make complex designs better understandable.

As my main part of the geometry normally consists of Solids and Extrudes, that are

wireframed anyway, that normally works well for me.

If there are too much things like filled slabs that hide the story below, I mostly tend to

set temporary Class Transparencies anyway, to allow overview and snapping.

Link to comment
  • 0

It is disturbing that after drawing your Walls and such, you insert a floor and

everything will be hidden behind that Floor, although it is clear that all other geometry

including tables and stairs are exceeding the Floors Z height.

You know you can use the "send to back" (and send forward, etc) tools to solve this?

It's still hardly ideal though; it shouldn't be necessary to do this manually. The things have Z positions so the software should be able to stack them properly automatically.

I think this is partly the reason why the VW suggested set-up seems to be to have separate layers for floor slabs, walls etc. But for me that just generates too many layers.

Link to comment
  • 0

There is a right way to use VW and a wrong way. The right way is 3d hybrid. I hear and accept that Top/Plan view does not work really well for that.

Personally, I think the wrong way is the right way, and for that Top/Plan view is perfect.

I should qualify that. It is the right way for the kind of work that I do, which is mostly work, often very complicated and detailed, that can pretty much be done by one architect.

It's not like I never drank the Kool-Aid either. I was a drinker from Minicad 3/4 on. It's just that eventually I realized that my vision of CAD utopia wasn't going to happen in my lifetime. I was probably on the wagon for good around MC/VW 8.

Then the question became, how to do the most professional, un compromised work?

For my work the solution was to pull the hybrid apart, to its rightful constituent elements: a 3D modeling environment for design and visualization studies; and, a 2D drawing environment to describe and communicate what is to be built and how.

In that solution, there was no need for the two parts to be the same software, or to have a "conversation" in two different languages in the same drawing package. I went for quality and usability in each part, and for that, VW gets the 2D nod. It's got a bruise or too, but still by far, The Champ.

So long live the top/plan view.

Link to comment
  • 0

Somehow I think the whole concept of "hybrid" is outdated or flawed.

It ought simply to be that if I create the geometry correctly in 3D, the software converts that information to a 2D floorplan using whatever drawing conventions are desired. This, essentially is what happens in my brain when I draw a floorplan - I create a 2D representation of what will be a 3D reality. Having decided on certain graphical conventions, there is only one correct way to draw the plan. This should be an ideal task for a computer.

The thing is, as far as I can see we aren't so far off doing exactly this. What is preventing it working seems to be this "hybrid" concept because all the time I'm trying to create a floorplan from a combination of elements using two different systems. The auto-hybrid tool is capable of creating a correct section from non-parametric geometry. And the parametric type tools are capable of creating 3D objects that mostly display correctly in 2D according to drawing conventions. But they don't always work because they seem unable to do certain things that the auto-hybrid tool can, like creating a correct section cut through walls of varying heights. Why have we got these parallel systems, which between them seem able to do everything we need, and yet somehow they can't work together?

It's like a drawing office with two draughtsmen - one of them can do you an accurate section cut through complex geometry but has never been taught about architectural drawing conventions. The other one is good on those drawing conventions but has never been taught how to project a section cut from 3D information. They don't talk to each other, work in different rooms, and the boss doesn't want to teach either of them the skills that the other one has. And when you send them the info for your design, you have to decide which bits to send to each guy, and they independently send you back two floorplans with bits missing which you then have to assemble together using scissors and sellotape.

Link to comment
  • 0
  • Vectorworks, Inc Employee

It's like a drawing office with two draughtsmen - one of them can do you an accurate section cut through complex geometry but has never been taught about architectural drawing conventions. The other one is good on those drawing conventions but has never been taught how to project a section cut from 3D information. They don't talk to each other, work in different rooms, and the boss doesn't want to teach either of them the skills that the other one has. And when you send them the info for your design, you have to decide which bits to send to each guy, and they independently send you back two floorplans with bits missing which you then have to assemble together using scissors and sellotape.

This is easily the best analogy for the Plan/3D issues that I have ever read.

Link to comment
  • 0

This is easily the best analogy for the Plan/3D issues that I have ever read.

Any clues as to whether there's any intention of seriously addressing them in the mid/long term?

From a practical point of view, I'd like to know if I be investing in developing my own systems of workarounds on the assumption that I'll be using them for some time to come.

Link to comment
  • 0

<

When we think of what CAD ought to be, in a perfect world, when it is far from easily done, when it amounts to a new experience we want but can't yet have, when it seems only right that what isn't should be, it raises the question: is this an engineering problem or a utopian idea?

One thing that sets utopian ideas apart, or defines them, is that fact that there is some present barrier that makes them not attainable. Merging the two drafters (US side of the pond here) is a bit like a sex change where both receive the sex of the other and keep their original parts.

Another quality of an utopian or highly desired idea, is that "sorta works" is not good enough. Were that not true, would we be having this conversation? What we have sorta works.

We have great engineers. Why isn't this just an engineering problem? I think it is because when you start to unpack this problem, the complexity mushrooms beyond belief. Both drafters are engaged in a process of selective approximation. They are discarding information from consideration.

Each can find the essential information for their purpose, which in one case to describe form and the other to describe assemblies. Each makes very approximate, (or discards) information dear to the other.

The utopian merge would require, I think, information without approximation and a higher entity to manage it. Right now, the two drafters have a hard time knowing for sure that something in their world doesn't work in the other. Sections are a good example of this.

This higher entity would be a form of intelligence, not just a pre-programmed series of logical steps. Information without approximation is an enormous data set. It's not hard to imagine that technological advances will make these two things possible. But right now, it is the stuff of dreams.

So now, for those on the front lines of architecture and the like, and particularly those in smaller organizations, there are, vis-a-vis this conversation, two options:

1. Sorta good and always getting sorta better and much better than it used to be hybrid.

2. Separate 2D and 3D.

I chose number 2, but that's just because I'm a bit picky. In most cases number 1 will get the job done.

Link to comment
  • 0
  • Vectorworks, Inc Employee

A few things:

1) Yes this is absolutely acknowledged as an issue.

2) No, a full solution will not be something that is implemented extremely soon.

3) A small part of what I think is the best way to address this is slated to be included in Vectorworks already.

4) No, unfortunately I can not discuss details about it yet. Both because it is still under NDA as well as the issue still not really done being worked out to significant levels of satisfaction.

Link to comment
  • 0

I still would prefer that auto hybrid be a function of the view and not the geometry. In my vision of the future, everything is 3d and 2d things are always stuck to a plane (layer plane or some Z distance if created in an orthographic view).

Enabling top/plan simply invokes a universal auto hybrid very much like the clip cube but with multiple "clips". One for the cut plane, one for above, one for below. The user can interactively change those clip heights, and can set a default line type for each, or override.

Let the viewport have an override for it as well.

With regards to the convention of certain things in plan being more symbolic rather than a true horizontal section...that's tricky. But perhaps we need a clip cube that can recognize a door, and when the door is clipped, rotate it about it's hinge point for us. When not clipped, it doesn't.

Link to comment
  • 0

I still would prefer that auto hybrid be a function of the view and not the geometry. In my vision of the future, everything is 3d and 2d things are always stuck to a plane (layer plane or some Z distance if created in an orthographic view).

Enabling top/plan simply invokes a universal auto hybrid very much like the clip cube but with multiple "clips". One for the cut plane, one for above, one for below. The user can interactively change those clip heights, and can set a default line type for each, or override.

Let the viewport have an override for it as well.

With regards to the convention of certain things in plan being more symbolic rather than a true horizontal section...that's tricky. But perhaps we need a clip cube that can recognize a door, and when the door is clipped, rotate it about it's hinge point for us. When not clipped, it doesn't.

Link to comment
  • 0
A few things:

1) Yes this is absolutely acknowledged as an issue.

2) No, a full solution will not be something that is implemented extremely soon.

3) A small part of what I think is the best way to address this is slated to be included in Vectorworks already.

4) No, unfortunately I can not discuss details about it yet. Both because it is still under NDA as well as the issue still not really done being worked out to significant levels of satisfaction.

Ok. Thanks for the reply on this.

Link to comment
  • 0
I still would prefer that auto hybrid be a function of the view and not the geometry. In my vision of the future, everything is 3d and 2d things are always stuck to a plane (layer plane or some Z distance if created in an orthographic view).

Enabling top/plan simply invokes a universal auto hybrid very much like the clip cube but with multiple "clips". One for the cut plane, one for above, one for below. The user can interactively change those clip heights, and can set a default line type for each, or override.

Let the viewport have an override for it as well.

With regards to the convention of certain things in plan being more symbolic rather than a true horizontal section...that's tricky. But perhaps we need a clip cube that can recognize a door, and when the door is clipped, rotate it about it's hinge point for us. When not clipped, it doesn't.

I think this approach makes a lot of sense.

The main question is how it would work when you were editing/drawing a floorplan. Because at the moment top/plan is an editable drawing mode as well as a type of "view".

Link to comment
  • 0
<

When we think of what CAD ought to be, in a perfect world, when it is far from easily done, when it amounts to a new experience we want but can't yet have, when it seems only right that what isn't should be, it raises the question: is this an engineering problem or a utopian idea?

I think it's an engineering problem now. I can't see that it is technically impossible to do. Plus, if BIM is really going to happen properly it's a problem that *has* to be solved.

Partly why I'm curious how other packages deal with it. What about Revit for example?

Recently I considered moving completely to Sketchup. With a couple of plugins it seems that sketchup can pretty much generate a completely correct floorplan from a 3d model. I chickened out of making the switch for various other reasons but I was tempted.

My guess is that what's stopping it happening in VW for now is how to work out a way of implementing it without changing things so fundamentally that a large proportion of the current VW userbase can't carry on working in the way they are accustomed to. That, I can see, must be a big and difficult engineering problem.

Link to comment
  • 0

I'm not sure if everyone posting here realizes that hybrid 2d/3d objects, such as walls, were the founding innovation of VectorWorks, then MiniCAD. When Richard Diehl launched this software, AutoCAD was far and away the dominant market player (it still is). There were a few startups in the CAD world that tried to tackle 3d, and I believe that VW and Archicad are the only real survivors. Revit came much later and was absorbed by Autodesk.

Back then, 2d drafting was the thing a CAD program HAD to do. The idea that an object could be both 2d and 3d held a lot of promise, and despite the glitches it's still a workhorse for us. If we want to convert something purely 3d to some kind of plan representation, getting the clip cube to work with fills and so that it can be viewported to a sheet is the best idea I've heard. I don't see any huge engineering hurdle there. But let's refine, not dismantle, the Top/Plan view.

  • Like 1
Link to comment
  • 0

Partly why I'm curious how other packages deal with it. What about Revit for example?

I'm on VW2012 so I can't comment specifically on how auto hybrid's work.

Revit controls views via a View Range. You can select the height of the cutting plane and the top and bottom height limits of the view. If I remember correctly (last time I used Revit was 2010, other than the demo I have) objects within the view i.e. furniture can be excluded from the section, shown in section at the specified cutting plane, or shown in a top view.

Revit View Range

  • Like 1
Link to comment
  • 0

Partly why I'm curious how other packages deal with it. What about Revit for example?

I'm on VW2012 so I can't comment specifically on how auto hybrid's work.

Revit controls views via a View Range. You can select the height of the cutting plane and the top and bottom height limits of the view. If I remember correctly (last time I used Revit was 2010, other than the demo I have) objects within the view i.e. furniture can be excluded from the section, shown in section at the specified cutting plane, or shown in a top view.

Revit View Range

We need something like that, basically.

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Answer this question...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...