ArchiCAD is dying

[Bruce]

I know that's a controversial subject line, but I believe it's true. Not because I want it to be, but because Autodesk is an advancing monster; ArchiCAD firms are switching to Revit, and Revit-based firms are buying ArchiCAD firms...and switching them to Revit.

ArchiCAD is a great program, but if it keeps going the way it is, I fear it will gradually dwindle until it's finally gone. On a level playing field, it comes out more or less even with Revit (I have done a detailed analysis that has been vetted by Revit experts) - but it's not a level playing field.

In my opinion, Graphisoft needs to do a handful of things to even the odds (yes, I will compare to Revit, as that's the main competition):

1. Rebrand & revamp the UI: CAD is an obsolete term. Even though ArchiCAD was BIM way before the term was even coined, I think the "CAD" in the name does it a disservice. Also, the user interface is old and tired. Should it go to the ribbon? No way. Should it be brought into the 21st century? Absolutely - there are plenty of excellent examples out there. Blender, a free 3D program, is undergoing its second UI redesign in about 5 years. If Blender can do it, Graphisoft can.

2. Introduce type-based elements. At the moment, pretty much everything is instance based. If you place 100 doors 900mm wide throughout the project, you have to select and change every single instance (this is an example, so please don't tell me the workarounds - that misses the point). Essentially, this is extending the attributes database to other objects. This makes project-wide changes so much more consistent, with no fear of missing an element.

3. Easier creation of parametric custom content: A beginner user in Revit can create a basic parametric object by using geometry and dimensions. It is intuitive and accessible. This does have its limits, but GDL is completely inaccessible to any but the advanced user with a programming mind...something architects and drafties generally don't have - otherwise they'd be programmers. A mix of the two would be extremely powerful - maybe an interface similar to Visual Basic, or Grasshopper? Not only for 3D elements, but also for 2D labels.

4. Better labelling & keynote tools: At the moment it's one label per element per view. What if I want to tag more than the ID? What about material, thickness, height etc. Revit is excellent in this regard, and also in the ability to create your label format as specific as you please. Key notes are also critical.

These are only four key improvements that I think are critical. There are many others that I could list, but this post is already too long. I say the above not to criticise ArchiCAD, but to try and help (misguided however it may be).

I could be wrong - I would be happy to be wrong...but the Autodesk monster is advancing...

These changes should be done the Graphisoft way: not to match what Revit does, but to equal and better it.

[Anonymous]

Laszlo,
In Archicad you should avoid to complex SOE as it does slow down archicad. Sometimes a simple layer with the same number between two complex geometrically elements can slow you down significantly.
It's a good thing that Autodesk has issued the manual as it is good to know what might affect the performance of your tool.
Archicad should have the same. It does not have unlimited power. I have designed some really complex buildings in Archicad and came across several unsolvable situations. Not fun. You try to do a workarounds etc.
Regarding Dynamo. Have you played with it? This is a totally different experience when you work with native object and can automate the bunch of tasks for you directly in your authoring tool. Not just import of "dumb" objects. Imagine a situation. Large building a lot of different floor finishes. What to do? Hey lets run a script that will automatically read all room finishes and model all finishes for me as needed in few minutes. A 1000 room stadium? No problem 2 minutes. All done without mistakes. How cool it would be to do something like that in Archicad? And this is just a tip of the iceberg. There is basically no limit.
Teamwork, performance, modeling in 3D are currently the biggest strengths for me in Archicad.
Scripting, automation, parametrization are Revit's.

[Laszlo Nagy]

I am not saying ARCHICAD is perfect or does not have its own limitations. I am just saying Revit is hitting a brick wall much sooner.

The automation aspect of Dynamo is interesting.
Please see this post of mine:

http://archicad-talk.graphisoft.com/viewtopic.php?t=44079&start=130

So let us suppose that we will have Rhino-ARCHICAD bi-directional connection at some point in the future. The question is: will this enable us to do the same automation Dynamo can do? I hope it will.

[Anonymous]

Only if Rhino will understand Archicad elements. I would prefer this to be part of Archicad. Vectorworks can do it. Why not Archicad.
https://www.youtube.com/watch?v=8V6l7zqUG7E

[Anonymous]

First thing AC needs to fix up is the modelling kernel and geometry, then they should focus on the antiquated GDL.
(As for) Vws has whole lot of separate issues/problems to deal with ie user base, credibility...?
Optimization, lower cost, simulations (simplified proofs) and better building solutions are what my clients are asking for..Here this video may help

https://plus.google.com/u/1/communities/115195196112205285216

[Anonymous]

Miki wrote:
Regarding Dynamo. Have you played with it? This is a totally different experience when you work with native object and can automate the bunch of tasks for you directly in your authoring tool. Not just import of "dumb" objects. Imagine a situation. Large building a lot of different floor finishes. What to do? Hey lets run a script that will automatically read all room finishes and model all finishes for me as needed in few minutes. A 1000 room stadium? No problem 2 minutes. All done without mistakes. How cool it would be to do something like that in Archicad? And this is just a tip of the iceberg. There is basically no limit.
Teamwork, performance, modeling in 3D are currently the biggest strengths for me in Archicad.
Scripting, automation, parametrization are Revit's.

I second this.
I think the Rhino-Archicad, Archicad-Rhino plugin connection will only cover the "cool" looking shapes part. Not the automation one.

[Bricklyne Clarence]

TMA_80 wrote:
Just a thought:
Is it the fact that ArchiCAD lacks parametric relationships that make it "faster" and able to handle bigger files ?
As ArchiCAD is implenting slowly but surely some parametric relationships, I wonder how would ArchiCAD respond if it went full parametrics . would it always handle the big projects ?
(is it the same automated vs semi-automated "Boing and the airbus" debate ?! ).

As for the generative design nothing to say more, we've been asking this for years, and even more ( scripting ) . However, the move from GS to implent a bi-directional grasshopper plugin is ( or will be ) , imho, an intelligent one.

Based on my experience working with both, I would categorically suggest that going full parametrics is a BAD idea, and I hope one that GS never pursue.

There's a good reason why AutoTable has been drawing back from Revit's relentless parametric engine and outright warns users to try to be economical with their constraints usage.

Four words : Cyclic Constraints Dependency error

Anyone who's ever worked with Revit and received this error message right before it crashed and took down your entire file with it, knows the pain of working with a program that relies too much on constraints and parametrics to a degree that it gets too many of them for even itself to keep track of.
You don't even need to go that far to see how damaging full parametrics would be to a typical ArchiCAD workflow.

If you've ever had a project with too many SEO's (which are in effect, a form of Constraints and parametric relationships), you know how the program slows down and drags to a crawl.
I think it's the sort of thing that all programs tend to struggle with (and frankly, speaking there's no program I've ever used with handles live boolean relationships that well and easily And this includes even higher-end modellers like 3ds Max and Maya).

Maybe it's the reason ArchiCAD is able to more fluidly handle larger projects of sizes that cause Revit to choke up.

It would be great to have the option to have some more constraints and parametric relationships - BUT only as an option (by the user) rather than automated, and also with a warning and caveat from the program that performance may be affected if they get too complex or many.

[abdelaziz]

I like this interface
https://www.youtube.com/watch?v=mah76xQ2SIg

[Anonymous]

Ok, there is some potentially great news about Archicad.
There is a new plugin soon to be launched which will create a direct dynamic
link between Grasshopper - Archicad and Rhino - Archicad

https://youtu.be/GatmUYbvaKQ

They say that it will be possible to transfer Any geometry from GS to Archicad BIM elements. If this is true it could be a huge thing. I hope it can work well with the curtain wall/facade panelling tool.
I am curious about the limits but it surely seems something that many have been waiting for.

[Anonymous]

Haven't been here on AC Talk for a while, but I guess AC may be dying in certain locales - like Canada, where AC enjoyed only niche usage and whose market share here has shrunk over the years to the point that I don't even know if or where it is used any more in the great white north.

Megaprojects are indeed done in Revit but if you don't force many alignments, and use linked files (site > building, complex shell > building), you'll get along fine. RAM is still the main thing you need and 16 GB will handle almost any project, and only the largest (say, a major airport) require more, say 24 GB.

While I teach Revit as BCIT I'd still love to see a resurgence of AC in North America in general and in Canada in particular.

[Anonymous]

metanoia wrote:
.
While I teach Revit as BCIT I'd still love to see a resurgence of AC in North America in general and in Canada in particular.

Yep, Revit is also used on Large projects Down Under (Lack of choice!), but here is the kicker, when things start going wrong they couldnt care less about which software is used to solve those problems..
So I can use what ever software I choose and no one gives a %$^# and
I make sure I point out to the project owner that specific (software) experts means Zero on Big projects without the right qualifications, knowledge, Data, Simulations,etc,etc
Goodluck, with teaching the uninformed & novices ...