[RFC] gluon based gateway project

[MoritzFago]

Please comment, this is a first draft, to get the idea out there. @sargon inspired this.

[lemoer]

As freifunk communities have a lot of different tools they prefer to use on their gateways (often due to external requirements), I doubt that this could work.

[rubo77]

I think this is an important project, so we only have one codebase to maintain and setting up a gateway will be as easy as setting up a node. Also we could work together on the "perfect" gateway that works for all gluon communities

[sargon]

@lemoer you are right, we all have different tools, community specific behaviors which we all implement into our gateway setups. On the other hand, we all do the same for our firmware, which we run on our nodes. We use our own community specific site repository, where we add special packages or just tune the base settings of our mesh networks. So we could do the same for our different tools and requirements for a gateway.

Now if you look at the basic gateway requirements, we all do one thing, permanently adopt behavior and requirements of gluon into our gateways. Starting with the right batman-adv or babel or ... mesh protocol version and continuing to the top where we adopt metric collection like respondd to be compatible with visualization tools for our nodes. So basically, we all burn manpower on the same basic setups.

Then we get to the threefold, with the disappearance of the german "Störerhaftung" our nodes miss on elementary feature, the ability to route traffic directly to the internet. An ability which make those nodes basically a gateway or supernode if you like.

[andibraeu]

what about this idea? should we merge it?

[lemoer]

implementaitions -> implementations

[lemoer]

I changed my opinion and I like the idea now. Even if the success and practicality of this project is unsure yet, it is not an argument for not trying it. If it works, it's definitely a benefit.

[andibraeu]

we need to add a possible mentor before merge :)