Cells - and headaches

Coming from the Autocad world some time ago, none of the areas has caused bigger headaches at our office, than working with cells. In Autocad you have blocks. Blocks are a part of the drawing, nesting usually is no problem. Pretty straightforward, in fact.

In Microstation you have a couple of varieties.

  1. normal cells
  2. shared cells
  3. compound cells
  4. feature cells
  5. parametric cells (done by PCS)
  6. cells done by the group command (unnamed cells)
  7. ...and things I don't want to know yet (generative components ?)

So what's the problem with having so many different types?

In the first run it's simply not easy to explain to a newbie at the office, of what he/she should use for a specific task. As I have found out in the last years, at least one thing is sure.

Never use shared cells!

Shared cells have been the reason of many problems here.

  • They are extremly difficult to update.
  • They are tricky when being nested.

A better way?

In my opinion Bentley must come up with a better solution on this matter.

  • Normal cells and shared cells can have the same name in a drawing.
  • This should not be allowed.
  • Cells usually need an external cell library.
  • It should be possible to have normal and shared cells be stored in a drawing. It should be possible to attach a drawing file as a cell library, thus enabling a workflow similar to Autocad.
  • Through discussions with Inga Mozoroff I finally found out, that cells, models and references are more or less the same thing of one matter. Using those nevertheless leads to different results.
  • Report functionality should be improved a lot. At this moment I often use the "cellcounter" utility, but this gives no results on nested cells. Triformas report utility doesn't really support cells at all, everything is listed as if it were a non-shared part of the drawing.

Let's stop here.

Hey - it's my first blog anyway :-)