Convert series of NET_PACKET_* defines into a typedef enum

[GrayHatter]

This change is

[iphydf]

I think this comment can go away, since it's already present at the bottom, where it's important. If you really must have a comment here, make it say that the enum should be sorted, which forces people to observe the comment at the bottom if they ever think about adding an enumerator > MAX.

[GrayHatter]

I added that comment to be a 'reminder' that this would be a perfect place for an assert. Maybe we should tackle the use of asserts in toxcore after this?

[iphydf]

This is totally not a perfect place for an assert, because an assert would be a syntax error at this point. I'm not sure I understand what you mean. The comment is redundant, and whatever it's a reminder of, is unclear.

[JFreegman]

Where did these two types come from?

[iphydf]

They are for #200. Better remove them from this PR.

[iphydf]

Can you remove the extra spaces?

/* See: docs/Prevent_Tracking.txt and onion.{c,h} */

We've replaced github reviews back to reviewable

[GrayHatter]

Review status: 0 of 1 files reviewed at latest revision, 3 unresolved discussions, some commit checks broke.

---

toxcore/network.h, line 119 at r1 (raw file):

Previously, iphydf wrote…

They are for #200. Better remove them from this PR.

Done.

---

toxcore/network.h, line 121 at r1 (raw file):

Previously, iphydf wrote…

Can you remove the extra spaces?

/* See: docs/Prevent_Tracking.txt and onion.{c,h} */

Done.

---

Comments from Reviewable

[iphydf]

---

Reviewed 1 of 1 files at r2.
Review status: all files reviewed at latest revision, 3 unresolved discussions, some commit checks broke.

---

Comments from Reviewable

[nurupo]

@so, and the reason for converting these typedefs to enums is that them having type different from integer
can help catch implicit type coercion? Just trying to understand why this PR was made.

Are sizeofs before and after this PR preserved? e.g. sizeof(typedef_value) == sizeof(enum_value). Would be bad if they are not and some code depends on this, especially if such sizeof is used to determine the size of data to throw in a socket, that would make the network protocol incompatible with the one before this PR. Based on the 255 restriction I assume this is not an issue, it probably uses sizeof(uint8_t) or just 1 for sizeofs, but that's just an assumption, it doesn't hurt to double-check.

[iphydf]

@nurupo enumeration constants have type int, so that doesn't change anything.

[nurupo]

You are right, not sure what I was even thinking about, it's so obvious too.

---

Reviewed 1 of 1 files at r2.
Review status: all files reviewed at latest revision, 3 unresolved discussions, some commit checks failed.

---

Comments from Reviewable

[nurupo]

---

Review status: all files reviewed at latest revision, 3 unresolved discussions, some commit checks failed.

---

Comments from Reviewable

[JFreegman]

---

Reviewed 1 of 1 files at r2.
Review status: all files reviewed at latest revision, 3 unresolved discussions, some commit checks failed.

---

Comments from Reviewable