So if i drop status on a parametric 3D cell it does two different things.
1) If it has variations it blows up into unusable garbage.
2) If it has no variations and it's just parametric and drop status it's usable and i can modify it however i need, no longer parametric.
Now why does this mater? I must drop status in order to move the parametric cell to a level of my choosing and change it's other attributes like line weight, color etc. In fact after discovering this flaw with variations we've gone backwards to make parametric cells Graphic cells so that we can get work done. I've already planted the Idea for a solution but no idea if/when it'll ever happen. Till then our in house working solution was going to be dropping status and at least we saved some time up front. Now we have to dump variations all together which makes some cells super un-usable.
I have a working theory if pressed i could code a pick list in an item type to change variable values based on nested if/then statements.... pretty sure i can make this work but its going to take me days to write those expressions and debug them so it's faster to just make them graphic cells. I've gotten pretty good at using Note Pad ++ to write expressions for parametric modeling easier but it's still clumsy.
Am i missing something here or is this just an over-looked bug? I know talking to some Bentley folks a while back the existing parametric solid/shared cell behavior of following the same level/line weight etc. is intended "for now".
Using Microstation only "mode" of OBD-U5.
-Grant
Grant Wood said:I've gotten pretty good at using Notepad++ to write expressions for parametric modeling easier
Can you tell us more about Notepad++ and how it helps with expression composition?
Regards, Jon Summers LA Solutions
Do you have any example files?
Grant Wood said: must drop status in order to move the parametric cell to a level of my choosing and change it's other attributes like line weight, color etc
For symbology in the cell definition if the profiles for features exist on a specific level that symbology should be retained when placed. ie a hole profile is place on leval "holes" with color set to "red" when placed that symbology should be honored.
If the profiles are placed on the default level then the symbology used is the active symbology during placement.
Following posts may offer some insight as well:
https://communities.bentley.com/products/microstation/f/microstation-forum/175553/ce-u11-parametric-cell-messed-up-if-not-on-default-level/508906#508906
https://communities.bentley.com/products/microstation/f/microstation-forum/126172/levels-in-parametric-cells/383911#383911
It's just note pad but for better suited for programming and scripting. So in the example below i adjusted the size of the window to wordwrap and i can see my entire expression cleanly. It'll retain it's aspect ratio un-like the Bentley expression builder that defaults to impossibly small. I can easily change expression values from say "round" to "ceil" or "floor" if the expression isn't working as i intended because i missed a potential solution result. I started using it to make feature cell expressions in v8i because Boolean Logic isn't something i'm formally trained in but had to write.
Sadly i'm still trying to find the actual fault for exploding cells i think it's missing constraints, and/or constraints not behaving universally as expected. I spent 1-1/2 days building a very well defined ladder cell (screen shot below) and it seems to work okay with a few variations but a stair cell I've been working on isn't working 100% and some other cells built within the office by others are not working with variations.
Thus I'm willing to conclude it might be user error, If we never use variations even a poorly constrained model is drop-able... but with variations it isn't. The tutorials on how to create parametric cells are very simple for what we're attempting to create. I still wish there was a defined set of basic rules regarding when/how to use the 2D vs 3D constraints and the more optimal way to build a part i can create the cells 100 different ways but some ways work better then others. It feels like in this regard the product is a sandbox, and it's up to us users to figure out how to really use the tools for anything complicated. This makes for a hard choice when we consider profitability, in the same time-frame i could create a parametric cell i can model all 100 variations as graphic cells and know they will work without question. Thus at the present for any purchased standard part/product we're just making them graphic cells. The only parametric cells are building tools that we "hope" will pay off in the long run for design.
I actually regret saying this, but feature cells in v8i worked far more predictable and better then the CE tools thus far. The 2nd screen shot below is a classifier built with material thickness in 5 minutes you could have a functional design done, modify it if you needed in the future and detail it for fabrication. For the sake of comparison the classifier took around 4-5 hours to model with feature tools in v8i. Tho i will give credit where it's due, the expression builder is light years ahead of v8i.
I'll have to keep playing with it and try to isolate the failure point.
To address the link 2D parametric cells are great but it's the 3D i'm concerned with in our work flow we need to be able to move 3D parametric cells to any level and we never let the level define the symbology of a part we control the attributes (line weight, style, color) on each element. Thus current parametric 3D cells are only helpful to start a design we have to drop status and if the design changes we just manually fix the cells as required.