Space tool resizes empty pool vertically

[nikku]

Describe the Bug

An empty pool has a dedicated, fixed height shape to distinguish it from lanes / expanded pools. Using the space tool it gets resized however:

Steps to Reproduce

1. Model an empty pool
2. Use space tool on top of empty pool

(Can be reproduced on https://demo.bpmn.io/s/application-processing).

Expected Behavior

Empty pool does not resize (vertically).

Environment

• Browser: Any
• OS: Any
• Library version: v10.x, https://demo.bpmn.io

[MaverickDe]

Hi, can I get assigned to this issue, Thanks

[nikku]

@MaverickDe If you're interested go ahead and file a PR.

[jyzib]

Hi, can I get assigned to this issue, Thanks

[nikku]

@jyzib You don't need to be assigned. If you want to contribute, please open a PR that fixes this issue (or attempts a fix).

[hkupitz]

I have a fix ready for the space-tool behavior, but I wonder if it generally should be impossible to resize an empty pool vertically.

The bug description by @nikku says:

An empty pool has a dedicated, fixed height shape to distinguish it from lanes / expanded pools. Using the space tool it gets resized however:

Maybe I am misunderstanding this, but should the following be possible or not?

[nikku]

@hkupitz That is an interesting question. I'd argue: A default size empty pool makes it simpler to recognize it and not confuse it with a lane. So 👍 from my side.

[nikku]

On the other hand making something not resizable is a bold move; we may not be aware of all scenarios, and we may not be ready for on-direction resizing either, nor won't be our users. I.e. I don't know this to be a thing from a drawing tool.

So let's decouple this fix from the general "should be resizable" discussion. In terms of resizing we are well covered; essentially we want empty pools to behave like text annotations; these do not resize either (in any direction):

[hkupitz]

I think using the spacing tool along the x-axis should be allowed as empty pools quite often stretch along another accompanied pools:

When disabling both axes, this leads to - in my opinion - unexpected behavior:

What do you think, @nikku?

[nikku]

100% agreed.