search

panrg / path-properties

A Vocabulary of Path Properties
Other
1 stars 3 forks source link

Fix terminology #11

Closed renghardt closed 5 years ago

renghardt commented 5 years ago

Changed definitions according to the suggestions in #1.

Questions:

cyrill-k commented 5 years ago

Changed definitions according to the suggestions in #1.

Questions:

* With this, a path is always between two hosts, but a subpath can be between a host and a node. I think this makes sense, as we are talking about "endpoints" (hosts) a lot. Do you agree?

Yes, I think it might be a good idea to more clearly separate paths and subpaths.

* Is the part about the layering clear enough? I think a single path ultimately consists of just one sequence of path elements, just that some lower-layer path elements are hidden when you, e.g., only look at the transport layer. I'm not sure if "A path is defined on a specific layer" actually conveys this. Ideas? Opinions?

Maybe instead of saying: A path is defined on a specific layer.

We could say: A path can be viewed as an abstraction on a specific layer, omitting lower layer path elements.

renghardt commented 5 years ago

Thanks for the feedback and suggestions!

The node definition now talks about "processing" packets, but then gives a few examples. I'm still wondering if we need the "implements a function which" at all.

I like your proposal about the layering and I put everything layering-related into a single definition (the path) instead of having it spread across multiple definitions.

Also, a property is now defined on one or a sequence of path elements - maybe we can allow a sequence of length 1 and simplify our definition accordingly? But that would confuse me if we didn't make it explicit, so I left it like it is now.

What do you think?

cyrill-k commented 5 years ago

The node definition now talks about "processing" packets, but then gives a few examples. I'm still wondering if we need the "implements a function which" at all.

I simplified it to: An entity which processes packets, e.g., sends, receives, forwards, or modifies them.

I like your proposal about the layering and I put everything layering-related into a single definition (the path) instead of having it spread across multiple definitions.

The definitions looks cleaner now. I removed "Paths can be traversed by packets".

Also, a property is now defined on one or a sequence of path elements - maybe we can allow a sequence of length 1 and simplify our definition accordingly? But that would confuse me if we didn't make it explicit, so I left it like it is now.

What do you think?

I would also leave it as it is.

renghardt commented 5 years ago

Thanks for the changes! I just got feedback that "directly adjacent" can be reduced to "adjacent", which I changed in the draft. Otherwise, I think this is good to go.