Merge "SablierV2Linear" and "SablierV2Cliff"

[PaulRBerg]

Merges the two contracts by keeping the SablierV2Linear nomenclature.

Closes #86.

Depends upon #75, #78, #79, #82 and #83 being merged first.

[andreivladbrg]

When contract.create is used in the test function, that should be createdStream. When the streamId from the setUp function is used in the test function, that should be actualStream.

[andreivladbrg]

Issue

The actual SablierV2Linear contract is basically the former SablierV2Cliff. Comparing the gas report from this PR with the one I did here #47 (comment), I am surprised to find out that the gas cost for the create and withdraw functions has increased, but for the cancel function has decreased, as you can see:

Gas cost

Linear	Avg	Median	Max
cancel	33184	47612	55987
create	188803	202149	209449
createWithDuration	209407	209407	209407
withdraw	46043	59229	63579

[PaulRBerg]

When contract.create is used in the test function, that should be createdStream. When the streamId from the setUp function is used in the test function, that should be actualStream.

I see where you're coming from, but I suggest sticking with actualStream, because that would mean more consistency across the board. Even if the stream is created, in the context of running a test with assertEq, that is still an "actual" stream.

[PaulRBerg]

I was surprised to find out that the gas cost for the create and withdraw functions has increased, but for the cancel function has decreased.

This is why I suggested to you to reference the commit that you had used for generating the gas table. The best way to ascertain if the gas costs really increased is to reapply the changes made in this PR on top of that commit specifically - but of course this would be a hassle so not worth it to do.

There's a host of reasons for why the gas cost may have increased in the meantime - we may have added more tests, or we may have increased the complexity of the tests, or the REVM might have changes some of its operational gas costs.

I wouldn't dwell on this. There's a lot of benefits in merging the two contracts.

[andreivladbrg]

There's a host of reasons for why the gas cost may have increased in the meantime - we may have added more tests, or we may have increased the complexity of the tests, or the REVM might have changes some of its operational gas costs.

I haven't checked anything, but I am pretty sure it's about the complexity of the tests.