Wishes
Post your wishes about Graphisoft products: Archicad, BIMx, BIMcloud, and DDScad.

The new working environment, was Open two stories ...

Anonymous
Not applicable
I would like to be able to open multiple plan view windows, but it would almost certainly involve a major overhaul of the way the program works. ArchiCAD (despite inventing the BIM concept) is plan centered for historical reasons (it allowed the program to run on the much slower machines of twenty years ago). Prior to version 5.1 the plan view was the only way to edit the model with 3D and section views as strictly 2D output.

To allow multiple plan views would (IMHO) require the major change of going from the historically plan centered to a model centered approach. The primary interface to the project would become the 3D window and everything else would behave as specialized views of the model.

I don't mean by this to imply that I object in any way to the idea of multiple plan views. In fact it is part of my ultimate fantasy of how ArchiCAD should operate.

What I am imagining (perhaps in ArchiCAD 11?) is:

- The 3D window is the primary and central interface for developing and navigating through the model.

- 3D navigation and editing with the ease of SketchUp and the power and accuracy of ArchiCAD.

- Instant 3D cut-away views of any plan, section or detail in the main model window. This means that the model is cut as defined in the particular view but is still be free to move and edit normally in 3D.

- The option to ghost the 2D elements in the above cut-aways. (I'm not sure how useful this would be, but it sure would look cool.)

- View types: Plan (floor & ceiling), Section & Elevation, Interior Elevation (that knows what room it's in, and not a kludgy add-on), Details, and 3D Views. (This is really just the current types with the addition of Plan views)

- 3D views would be separate windows for axon or perspective views with the ability to annotate in the model or the view plane (ie. dimensions oriented to the model in an isometric projection and notes and titles oriented to the page).

- Plan views: In addition to the variable cut height also share the section tool's ability to offset so that split levels or other conditions could be easily and accurately drawn. Setting the plan's cut height also eliminates the annoying split walls we have to use to model transom windows etc.

- Ceiling Plans: Like the floor plan views only looking up. I probably don't have to tell anyone all the workarounds this eliminates. Imagine, easy ceiling plans that look exactly the way you want them to.

- Plan boundaries: The plans are (optionally?) bounded allowing the tool to be used for enlarged plans which are still live model views. (This is a major limitation of the detail tool for this purpose.)

- Sections: The section tool is visible and editable in the main model window (as is the plan tool; perhaps only when the view is activated in the 3D window). The section tool has the ability to offset both vertically and horizontally allowing different section locations through each floor (and ending up at the ridge if desired).

- Details: These are still 2D drawings (derived from the model) as far as I am concerned. I find that if I want to correct the model to make the detail work there is much more to it than just the limited area of the detail.

- General: All views have the option to display as line drawings, shaded views, or rendered by any available engines including any combination of these with or without shadows as appropriate (and feasible). Imagine a site plan with notes and dimensions rendered with shadows in a watercolor style for the Planning Board hearing.

What do you think folks?

Would you buy this upgrade?

PS: I know this probably belongs in the wish list. I got a bit carried away beyond multiple story windows or live details.
Perhaps this whole topic should move to the wish list. (What do you think Djordje?)
31 REPLIES 31
Anonymous
Not applicable
Djordje,

I don't see how to add a poll to an existing topic. I assumed that this was only possible when creating a new topic.

Tell me how and I will be glad to add the poll. Otherwise please feel free to add it. The responses could be: 1. Great stuff, 2. Sounds OK, 3. Maybe, 4. Yuck, 5. No way.


Note:
This started as a response to Adalbert's topic about opening multiple plan windows at:
http://archicad-talk.graphisoft.com/viewtopic.php?t=846
Anonymous
Not applicable
Matthew wrote:
Tell me how and I will be glad to add the poll. Otherwise please feel free to add it. The responses could be: 1. Great stuff, 2. Sounds OK, 3. Maybe, 4. Yuck, 5. No way.
I might have not expressed myself correctly, but 2 or more plan view is what I was talking about in this poll :
http://archicad-talk.graphisoft.com/viewtopic.php?t=836
Djordje
Ace
Matthew wrote:
I don't see how to add a poll to an existing topic. I assumed that this was only possible when creating a new topic.
Edit the post; then you can add the poll. Like I did ...
Djordje



ArchiCAD since 4.55 ... 1995
HP Omen
__archiben
Booster
hmm. quite a bit to think about matthew! your notes provoked an immediate reaction from me and i felt i had to jump in, although i'm still not sure which way i'm going to come down on this. . .

upon seeing sketchup demo'ed at the archicad university in the summer, my first thoughts (apart from the initial 'wow' factor!) were: 'this is the future of CAD', and then 'this is how CAD would have started had the hardware of the time been capable'.

my reaction to your comments here come more from a hardware capability/project size perspective, something that is also a big part of my thinking for a response to your teamwork topic (which i am still trying to get together!), and my love-hate relationship with the virtual building model concept.

i think that the reality, as far as my experience suggests, is that a typical/average workstation still falls short on processing power than is needed for large buildings. as much as i hate the idea, i don't know how we would have managed a few of the jobs here without quite a bit of 'unlinking' taking place. i am still sceptical about our ability to take these jobs through into the full construction phase without a bank of arrayed G5's at each workstation!!! sure there is an IT budget, but a majority of our machines here are only less than a year old. try selling the idea to the directors that these 'new' machines just can't handle the job! doesn't work . . . and we wait for the next round of fees that could justify major hardware upgrades.

also, by "archiCAD 11", i would imagine that the fields held in the database for each element, and the additional features included from the wish list would have grown too, putting equal if not more strain on tomorrows processors as today! (the hardware-software neverending upgrade cycle!)

you refer to the historical plan-centric model as a way of enabling the program to run on much slower machines, and i think that you still have to allow for this in your proposal. but this doesn't necessarily mean that a model-centric approach cannot be used. i think that in this re-write of the program you suggest, could possibly be included a more efficient dynamic handling of the information required from the database to perform each operation.

i'm betting that much of the element information loaded into memory is not necessarily needed for simple editing, (please somebody correct me if i'm wrong here), and managing just when certain information is needed could be a way of going model-centric without killing big project files. for example, moving elements around on plan requires only specific information from a processing point of view, whereas editing it's parameters in a dialogue box could 'load-in' extra information at the point at which it is needed. this is also something that i've been thinking about with regard to the teamwork topic, and certainly not something that i have clear in my mind yet. there are flaws - direct editing in the info box?

i still believe that unless hardware advances dramatically in relation to it's software counterpart that the ability to 'unlink' information in order to efficiently work in larger jobs must be retained.

i think that despite all of this, i am going to be coming down on your side with regard to the direction archiCAD should take. as i said, mine is a love-hate relationship with the BIM: i do want to see it working well and in a manner that is efficient and intuitive to use, but sometimes i think that i need the pinch of salt to help me take a reality check. murphy couldn't have been an optimist all the time, eh djordje?

i am imagining your archiCAD 11 . . . and it looks just like sketchup 7!

~/archiben
b e n f r o s t
b f [a t ] p l a n b a r c h i t e c t u r e [d o t] n z
archicad | sketchup! | coffeecup
Geoff Briggs
Mentor
Holy hanging chad Batman, how do you vote on this one

Is this really about multiple plan views or about the future of ArchiCAD and our collective vision of BIM?

Some random thoughts:

We have multiple plan views (with boundries) now using the detail tool. This tool could evolve to encompass more of the things Matthew is looking for. Live 2-way linking and detail-in-detail would be a good first step. User definable Z coordinate is another.

The option to manually control when a window is rebuilt (the subject of a recent poll) would probably be necessary to address the inevitable hardware issues Ben raises.

Which leads to another question: Shouldn't ArchiCAD retain distinct view settings (layer combination, view options, etc.) for each open window? Without addressing the 3D window specifically, this would make ArchiCAD more model-centric. Each window would represent it's own snapshot of the virtual model. That's the way it works when you save out to PlotMaker, why not when you're working?

As to the bigger question of whether the 3D window should become the primary interface, that makes sense in the design phase of a project. Thus all the comparisons to SketchUp, a product focusing on the design phase. Who wouldn't want to see their groovy interface come to the ArchiCAD 3D environment?

But we all know the difference between these products. SketchUp is a surface modeler, ArchiCAD is a database. A very special kind of database where most windows act as both records (input) and reports (output). Many of my own wishes speak directly to this. Things that reduce 2D drafting, like slabs with associated fills, or proper composite intersections in section, are more than time savers. They represent the intelligence required for the assemblies we model to report themselves correctly no matter the view.

All views reveal something about the whole. As the project progresses it becomes more technical. Plans, sections and details become valuable places to refine the design and the best area to focus our effort as we insure that our concepts can be built. Reduce not the power of the plan window, but bring parity to S/E and detail views.

So bring on the improvements to the 3D environment. Anything to improve the design and presentation capabilities of ArchiCAD. But don't banish the other windows to the second class citizenship of being mere extractions of the model. Nay, improve them too!
Regards,
Geoff Briggs
I & I Design, Seattle, USA
AC7-27, M1 Mac, OS 14.x
Anonymous
Not applicable
Geoff wrote:
Holy hanging chad Batman, how do you vote on this one
I agree that this topic is a bit large for a simple vote.
Is this really about multiple plan views or about the future of ArchiCAD and our collective vision of BIM?
The original post was about multiple plan views. My response went a bit beyond this.
Some random thoughts:

We have multiple plan views (with boundries) now using the detail tool. This tool could evolve to encompass more of the things Matthew is looking for. Live 2-way linking and detail-in-detail would be a good first step. User definable Z coordinate is another.
I would rather see a different plan view type and keep the detail tool as it is. I think in the overall scheme it is better to maintain a distinction between plans and details. I presently use the detail tool as you describe for enlarged floor plans, but I would rather have them in a live model view and I think of them as a fundamentally different type of drawing. The enlarged plan is a drawing of one specific place in the building whereas detail drawings are usually of typical conditions which may occur in many locations. This is why I don't think the detail tool needs to be a live model view (perhaps I should introduce this as a new poll topic).
The option to manually control when a window is rebuilt (the subject of a recent poll) would probably be necessary to address the inevitable hardware issues Ben raises.
This would be very useful now and, you're right Geoff, would be absolutely essential in the model-centric vision. (The lack of this ability is the main, perhaps only, reason I still unlink the elevations.) I don't agree that hardware performance is such a big issue. The plan views in my scenario would behave similarly to the section views and I haven't found editing in section to be particulary tedious. The big job from a programming standpoint would be keeping the immediate feedback as it is in plan presently (and adding this in section?) but this doesn't seem to me to be a hardware performance issue.
Which leads to another question: Shouldn't ArchiCAD retain distinct view settings (layer combination, view options, etc.) for each open window? Without addressing the 3D window specifically, this would make ArchiCAD more model-centric. Each window would represent it's own snapshot of the virtual model. That's the way it works when you save out to PlotMaker, why not when you're working?
I don't see much of a change here. The present approach in sections and details seems fine. The distinction between the project map with the basic window list and the multiple associated view sets would work fine for the plan windows. The complex part is how to organize the plan views in navigator. Perhaps the existing story views would remain and would become subsets with multiple plan views stored within. The thing I haven't got a clear image of is how going up and down the story views would work; perhaps it would depend on whether a particular plan view crossed multiple stories (talk about recursion problems ).
As to the bigger question of whether the 3D window should become the primary interface, that makes sense in the design phase of a project. Thus all the comparisons to SketchUp, a product focusing on the design phase. Who wouldn't want to see their groovy interface come to the ArchiCAD 3D environment?

But we all know the difference between these products. SketchUp is a surface modeler, ArchiCAD is a database. A very special kind of database where most windows act as both records (input) and reports (output). Many of my own wishes speak directly to this. Things that reduce 2D drafting, like slabs with associated fills, or proper composite intersections in section, are more than time savers. They represent the intelligence required for the assemblies we model to report themselves correctly no matter the view.
I don't see a major issue here. I don't think the ArchiCAD 3D interface should be just like Sketch Up; only adapt some of the ideas as appropriate. Perhaps a new massing study tool would be nice (virtual clay). A model-centric approach would allow the possibility of 3D only tools like this. A quick (3D) space bar click and the massings could become zones with walls and floors. Perhaps the massings could have optional relationships between each other and with their contents (now THAT might require some serious hardware).
All views reveal something about the whole. As the project progresses it becomes more technical. Plans, sections and details become valuable places to refine the design and the best area to focus our effort as we insure that our concepts can be built. Reduce not the power of the plan window, but bring parity to S/E and detail views.

So bring on the improvements to the 3D environment. Anything to improve the design and presentation capabilities of ArchiCAD. But don't banish the other windows to the second class citizenship of being mere extractions of the model. Nay, improve them too!
I wholeheartedly agree. Part of my conception is to bring improved function to all the view types while keeping the unique qualities that each requires. The idea is not to diminish the plan views in preference to the 3D but to enhance their function by unburdening them from being the central means of relating to the project.
Geoff Briggs
Mentor
Matthew wrote:
I would rather see a different plan view type and keep the detail tool as it is. I think in the overall scheme it is better to maintain a distinction between plans and details. I presently use the detail tool as you describe for enlarged floor plans, but I would rather have them in a live model view and I think of them as a fundamentally different type of drawing
Isn't that just semantics. What if it were called the Partial Plan by Polygon Boundary tool? I'm all for the functionality you describe and have no problem with a new window type, unless that prospect keeps it from happening.

Maybe we already have the tool in place if only it were enhanced, as the section tool was, by allowing it to be linked or unlinked. The current section tool can be used for sections of course, but also elevations, wall sections, and sectional details. An enhanced detail tool would be just as versatile.
The complex part is how to organize the plan views in navigator.
Views may appear under the detail part of the project map, but can be put wherever you want in your view sets.
I don't see much of a change here. The present approach in sections and details seems fine.
Wouldn't it be more true to the virtual model concept if each view (window) retained it's own settings even through a rebuild?

Imagine seeing the completed exterior of your project in 3D (no interior parts), a framing plan in 2D (structural assemblies only), an enlarged plan in a detail, er partial plan window (composites, casework, furniture) and a section in an S/E window (composites, casework, no furniture). Each window is a live model view. Make changes anywhere, rebuild the lot and see how things fit, both in terms of design and documentation. As it is now all the views switch to the same layer and display option settings. To get the same feedback you have to cycle through all your views, rebuilding each in turn.
The idea is not to diminish the plan views in preference to the 3D but to enhance their function by unburdening them from being the central means of relating to the project.
I agree completely.
Regards,
Geoff Briggs
I & I Design, Seattle, USA
AC7-27, M1 Mac, OS 14.x
Anonymous
Not applicable
Geoff wrote:
Isn't that just semantics. What if it were called the Partial Plan by Polygon Boundary tool? I'm all for the functionality you describe and have no problem with a new window type, unless that prospect keeps it from happening.

Maybe we already have the tool in place if only it were enhanced, as the section tool was, by allowing it to be linked or unlinked. The current section tool can be used for sections of course, but also elevations, wall sections, and sectional details. An enhanced detail tool would be just as versatile.
Just semantics? Semantics, along with syntax and grammar, are how we organize the expression of our thoughts and as such I would say they are quite important. The changes we are discussing are about how we organize the creation and expression of our design work and I suspect that we agree that the details of how this is done are very important (ergo our conversation here).

What I am proposing is a structure in which the different types of drawings each have their own environment which is finely tuned to their specific requirements. These drawing types, as I see it, are:

1. Plans; including floor plans, enlarged plans, site plans, ceiling plans, structural plans, etc.

2. Sections; including elevations, building sections and wall sections (etc.?).

3. Interior Elevations: this is separate from the section type because its reference annotation requirements are significantly different and it should have the ability to know what zone(s) it is in.

4. Details; including derived details (those with a source view in the model), drawn details (2D drawings specific to the project) and various other details (company standards, vendor details, shop drawings)

5. 3D Views; including axonimetric and perspective views with 3D and 2D annotation tools.

6. Schedules & Specifications; including door, window & finish schedules, product specifications, general notes, etc. (these would be in a word processor, spreadsheet and/or database format either internal to ArchiCAD or live linked through add-ons to external apps.)

This arrangement, I believe, is superior to trying to make one tool do multiple functions. Certainly the detail tool can be used to make floor plans and the section tool can make interior elevations (I use them for these purposes now) but I would prefer an appropriate and unique enviroment for each of these primary functions. This also makes maintaining standards and training staff much easier. I would prefer to tell someone "use the plan tool to make plans, the detail tool for details..." and so on, than to say "use the section tool this way for an interior elevation," and "floor plans are drawn in the main plan view (along with ceiling plans, electrical plans, etc.) but enlarged plans are drawn using the detail tool." Each of these methods (workarounds, as far as I am concerned) requires substantially more documentation and training to maintain standards in an office with more than two or three architects.

Another big attraction to me is that my proposal allows all annotations to be done on one layer which is the same in each drawing. I presently (as I m sure most others do) use separate layers for annotating floor plans, ceiling plans, electrical plans and so on. It is the limitation of having only a single plan/story view that requires all these layers. (I suppose that I could do all my floor plans, etc with the detail tool now, but this not very appealing for obvious reasons.) I love the idea of setting all the 2D tools to one layer and never have to change them again.

The way I see it the is that these drawing (and schedule) views are the output/production part of the program. Design work is done in the model and story views (the latter is my new conception of the evolution the current plan view freed of the burden of annotations). The various drawing views could be made visible as reference keys or ghosted images in both the story view and 3D model.

Wouldn't it be more true to the virtual model concept if each view (window) retained it's own settings even through a rebuild?
Imagine seeing the completed exterior of your project in 3D (no interior parts), a framing plan in 2D (structural assemblies only), an enlarged plan in a detail, er partial plan window (composites, casework, furniture) and a section in an S/E window (composites, casework, no furniture). Each window is a live model view. Make changes anywhere, rebuild the lot and see how things fit, both in terms of design and documentation. As it is now all the views switch to the same layer and display option settings. To get the same feedback you have to cycle through all your views, rebuilding each in turn.
This is a great idea. To be able to have multiple windows open each with their own independent layer, display options and dimension preferences etc. would be terrific.
Geoff Briggs
Mentor
Matthew, it's hard to argue with the elegance of your logic or the eloquence of you argument. You have my vote.
Regards,
Geoff Briggs
I & I Design, Seattle, USA
AC7-27, M1 Mac, OS 14.x