Add ignored test case for conflicting deps in unrelated ws members

[maciektr]

Stack:

• Add ignored test case for conflicting deps in unrelated ws members #1007 ⬅
• Add ignored test case for dev-deps propagation #1030

⚠️ Part of a stack created by spr. Do not merge manually using the UI - doing so may have unexpected results.

[szymmis]

What are those tests related to, especially in the context that they are ignored?

[maciektr]

@szymmis basically bugs that we should sort out eventually

[mkaput]

I do understand the reasoning. I am afraid how this will play out with PubGrub, as I believe the canonical algorithm will attempt to unify some_dep in this scenario.

Did some user request this behaviour? Do we need to commit to it?

[maciektr]

Did some user request this behaviour? Do we need to commit to it?

Nope, just a thought experiment

I am afraid how this will play out with PubGrub

Well, I do not know as well. Though it seems that having same dep with different versions may actually make sense even in cairo - when the graph is disconnected.

[mkaput]

Though it seems that having same dep with different versions may actually make sense even in cairo - when the graph is disconnected.

That's a highly philosophical argument. One could want instead to use a single version of dep throughout the workspace because it's well -- a workspace. Given a workspace of multiple contracts, I'm not sure the author will want the contract A to have dependency v0, while contract B -- v1.

I think we have to ask users on Telegram.