I don't really understand what you are wishing for. How could anything automatically guess which things you want to be a different color, much less what color you want? Could you give a few examples of where random colors are useful?
One of the forum moderators AC 26 USA and earlier • macOS Ventura 13.6, MacBook Pro M2 Max 12CPU/30GPU cores, 32GB
Yes, there should be a more automated way to setup overrides for categorised or numeric data. Instead of having to create one rule for every value or range of a parameter it could be done by one rule which automatically divides the parameter to classes or ranges set by the user. For color overrides it could then either assign a color at random or based on a color ramp. Graphic overrides is a very intuitive aid when creating, auditing or presenting the model but there is stil a lot of obvious improvements that can be made to it, see this GS request for feedback.
In BIM collab zoom, it is possible to force colors and assign them automatically in a random pattern. This means that for each different value of a certain parameter, a color is automatically assigned.
For example, if we do this for the storey parameter, a color is assigned for each storey, which avoids creating a rule for each storey, or repeating the creation of a rule each time a storey is added.
It is typically used to check the parameters embedded in a BIM process
I also think automatic color classification could be useful.
Example 1: i have some walls with a list option property. The automatic color classification should evaluate the number of options, redistribuite them on a color gradient scale and then assign the right color to the specific element
Example 2: i have some walls with a generic number property. The automatic color classification should evaluate the min and max of the values, assign them to a color gradient scale boundaries and then assign the right color to the specific element.
I kind of get it, but don't see an easy or worthwhile implementation. In the likes of thermal imaging the colours have a natural logic of blue is to cold to red is hot. How does the assignment of an automatically generated (random) colour identify an element with a different parameter if you don't know what the colour indicates? If a colour range is used for wall thickness, will there be enough differentiation to highlight a wall is 5mm thicker than it should be? With GO's if I highlight all walls of a certain thickness to a set colour it becomes immediately obvious that there is an error in the walls that aren't highlighted.
For QA I like absolutes, either the checked value is correct or not and the GO is applied to suit.
If we are talking generalities like using Finite Element Analysis to give an indication of stress e.g. the percentage of glazing to a room or width of door openings based on occupancy for means of escape then I can see a value in that, but I think this is a whole new branch of GO. Perhaps a Building Performance Analysis tool that can, with appropriate input, create meaningful stepped or graded colour values based comparative percentage / difference / scale values is needed. The key word here is compare, which will require formula assessment. Personally I don't think GO has the capacity to deal with that and retain its relatively simple application.
Apple iMac macOS Ventura / AC26UKI (most recent builds)
This was what i had in mind. lets say you want to make a graphical overide for wall types. Currently you have to make a rule for each different wall type and keep increasing it if you have new walls.
With an automatic feature like you have in BIM collab you could make a rule where for every different wall a different colour would be automatically assigned and use a schedule to create a label of all the walls to place in the layout.
there are way more urgent things to fix with ArchiCAD but saving us time with repetitive tasks should also be somehow seen too.
Yes, the implementation would require some work but I don't think it would be conceptually difficult and I do think that it is worthwhile for the future. At the core it's about efficiency - why use multiple rules for something that easily can be achieved with one? But it is also a feature that is becoming common in other applications such as IFC viewers, Speckle and QGIS because it is a very effective and intuitive of way visualising and analysing spatial data.
As for the interface - yes the color functionality of AC would have to be enhanced (time to leave the 8-bit world) but it is quite easy to make it intuitive and effective - just look at QGIS where you easily choose criteria, which color ramp to use and the control the distribution.