>>301623
>Ah, okay. You don't know what link archetecture is.
Or…, I'm answering to someone who seemed to have an understanding of the word 'internet' is limited to it being a function on phones, sold by cell providers in units named 'mb'.
This would explain bringing in cell tower operators as a supposably relevant party in a wifi network. By assuming it is provided by your cell provider because it is on your phone, so probably done though the cell tower; it could make sense. But now it seems you may have a different reason for this.
Following the same argument from ignorance, conflating them with internet providers by having them charge people for 'wifi' (persumably 'internet', because wifi is internet, only free) traffic. The cell tower you connect to probably has the same company owning it as your cell provider. Both have the word 'cell' in it. So obviously they are also entitled to money.
Hence I did an apparently shitty job of trying to explain that neither of these technologies are required to have an internets.
>>Of course the internet is built on an archetecture, several archetectures even. Even wireless communications uses archetecture.
You are correct. But there is no overall architecture for what is the internet. The internet consists of several hetrogenous networks put together. It can even be argued that no complete architectual map could be made as every leaf node on pretty much every sub-network could have a network running with undocumented or secret protocols, mapping back to standard internet protocols at said leaf node.
>If you were thinking about just having nodes polling every other node themselves and forwarding all the requests they receive, then that shit is going to be unusable.
Here you are haphazardly describing some sort of simplefied networking protocol, but I am unsure why. I do not see the significance with it in regard to the internet working with or without cell towers, and at this point I dont know what you think i was arguing for instead.
>Even when it does work, it will be slow as fuck. Do you not remember Freenet .5? And that was using cables.
With 'Freenet .5', do you mean the 0.5 release of the software to run freenet as in https://freenetproject.org/ ? In that case, interesting you mention a version number. A screenshot on that site mentions a 0.7-a2-pre release so perhaps some of the problems you may have had have been solved or mitigated by now. Though, the main way this program was used was to visit normal websites one can also access without it. S
ince the program itself facilitates this by going over the same internet via different locations with numerous jumps, this is most likely going to invoke extra latency over the direct route. This is of course assuming the direct route is actually available and not blocked by your provider for some reason. That this is not always the case is probably part of the reason this program exists.
By the way, interesting you mention cables, but I will go in to that later below.
>https://en.wikipedia.org/wiki/Data_link_layer
Ah yes. The data link layer. the accords we struck on how we say things went well, things went bad, who is A, who is B, how to get to A, how to get to B, and who is allowed to at a given time, with regards to the bits and symbols sent to and received from the physical layer.
If you are wondering which protocols to use in a wifi mesh network, it would depend on the specific network itself. It is possible, and in small scales relatively efficient, to just use the standard ethernet protocols for this.