Daniel Hahler

I think `pytest_collection_modifyitems` could always add an item. But I'm not sure it should be run when `-k` or args in general are used then.

Updated docs, and started to use new fixtures only internally, but it is quite messy. What also should be considered is using prefixes with the option names..

innodb has one failure: > E TransactionManagementError: Executing DDL statements while in a transaction on databases that can't perform a rollback is prohibited. myisam looked like the server was not...

Also Neovim does not handle `:.` to remove the drive on Windows: This is working with Vim, but not Neovim: ``` AssertEqual fnamemodify('C:/foo', ':.'), 'foo' AssertEqual fnamemodify('C:\foo', ':.'), 'foo' ```...

@betchern0t Note that the above could also "just work" (instead of improving the crash/error message). See https://github.com/pytest-dev/pytest/pull/6034 for something you might want to pick up for this.

In case you are interested: I am taking at stab at this in https://github.com/blueyed/pytest/pull/42, since I have "file:lnum" in summaries now (https://github.com/blueyed/pytest/pull/38).

@The-Compiler This is done automatically (Github bug/issue?) - maybe only for issues reported by myself? I will usually re-open them then if applicable (when going through my notifications).

Just for reference, since I've found this issue again: I'm using this myself since a while, and it is very nice (as expected ;) - I know why I've asked...

As OP I'm closing it again: it's fixed for me (literally).

SimpleCookie.__parsestring is just too strict: it also fails to handle expiration dates incorrectly (parses them based on the sane-cookie-date only), see https://bugs.python.org/issue35824#msg334392. (this means that aiohttp is affected by this...