Fix bug with snapshot thinning.

[NotSpecial]

In the current version, the snapshot thinning does not work properly because the n%10 != 0 check, intended to always add the final snapshot, suffers from an off-by-one error.

Concretely, the main loop in the function appends snapshots at index 0, 10, 20, 30, ...; i.e. the first element and every tenth after. Second, the final check tests whether the list of snapshots is a multiple of 10, and if not adds the final element. And this is where the problem lies: the loop does not take every tenth element, it takes the first one and then every tenth after, i.e. the 11th, 21st, 31st, etc -- it is essentially shifted by 1, and thus the n%10 check is incorrect. This leads to the following behavior: If the list has a length that is a multiple of 10, the final element will be excluded. If it has a length of a multiple of 10 plus 1, the final element will be duplicated. Here's an illustration:

IN:  []
OUT: []
---
IN:  [0]
OUT: [0, 0]
---
IN:  [0, 1]
OUT: [0, 1]
---
IN:  [0, 1, 2]
OUT: [0, 2]
---
IN:  [0, 1, 2, 3]
OUT: [0, 3]
---
IN:  [0, 1, 2, 3, 4]
OUT: [0, 4]
---
IN:  [0, 1, 2, 3, 4, 5]
OUT: [0, 5]
---
IN:  [0, 1, 2, 3, 4, 5, 6]
OUT: [0, 6]
---
IN:  [0, 1, 2, 3, 4, 5, 6, 7]
OUT: [0, 7]
---
IN:  [0, 1, 2, 3, 4, 5, 6, 7, 8]
OUT: [0, 8]
---
IN:  [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
OUT: [0]
---
IN:  [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
OUT: [0, 10, 10]
---
IN:  [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11]
OUT: [0, 10, 11]

In this PR, I simplify the code somewhat to avoid the index check: The main loop only considers elements up to, but excluding the last one (i < n-1). Afterwards, the last element is always added.

---

This change is 

[NotSpecial]

TODO: A unittest would be nice. I don't have much experience with unittess in go, but I can give it a shot next week.

[coveralls]

Pull Request Test Coverage Report for Build 7416

• 4 of 6 (66.67%) changed or added relevant lines in 1 file are covered.
• 156 unchanged lines in 2 files lost coverage.
• Overall coverage decreased (-3.9%) to 63.267%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
parser/tcpinfo.go	4	6	66.67%
<!--	Total:	4	6	66.67%	-->

Files with Coverage Reduction	New Missed Lines	%
active/active.go	2	90.63%
storage/storage.go	154	0%
<!--	Total:	156		-->

Totals
Change from base Build 7409:	-3.9%
Covered Lines:	3009
Relevant Lines:	4756

---

💛 - Coveralls

[NotSpecial]

Mh, just by looking at the test file, I cannot tell where the expected number of 1588 snapshots comes from. Probably needs some extra investigation.

What I want to say: As the counting has been wrong, I am not sure whether 1588 is the correct number to expect.

[stephen-soltesz]

This change was incorporated into https://github.com/m-lab/etl/pull/1105