Re: Re: DogBone
By: Wilfred van Velzen to Tommi Koivula on Mon Jun 21 2021 11:06:49
I don't quite understand the idea of this server... What is (or will
be?) the purpose of collecting this kind of data?
I'm not sure either. Currently it just seems to be a proof of concept. But why register nodes/bbs's when we already have the nodelist?
As I stated in my intitial message, I have some errors in my approach, so this is a proof of concept. The actual concept is this. I made a few statements about moving forward with FidoNet and was taken to task by various net.gods, and told, rather snidely, that if I wanted things to be different, then I should go ahead and write it myself... so that's what I'm doing. I have a concept in mind that I think could be useful. One which encapsulates a number of different aspects. Just one of those is auto updates to the nodelist. That's the simplest and least important aspect, I think. Extending out from that is auto-discovery and routing, shared (distributed) registry of nodes and users, registry of echos and which are carried by which nodes etc.
I wrote a server and a client that would share this info, but realised that convincing sysops to install yet another piece of software would be a real struggle until/unless the concept was proven.. and even then, it would be much MUCH simpler to first convince the BBS writers to include the functionality into the BBS software suite. So I decided to take the tools already available, and build something that works to make that proof of concept tangible.
I decided that writing a netmail based "robot" that would allow sysops and users to register and/or update their information was both in-line with current/old Fidonet procedures/technology and required nothing more than what everyone on FTN already has access to ie Netmail. Thus this proof of concept I requested testing of.
The current model is hugely flawed. I suspect that a better model will end up having the nodelist data imported into it's own table, and any changes to it only allowed with a password. Then as a secondary table, have user data that people can choose to submit. This would function something like a "white pages" for netmail addresses. One aspect of that would be the AKA funtion as most of us have multiple addresses across the FTN networks for example, simply registering on a new BBS gives you a new netmail address on every FTN and QWK network that the new BBS is a part of.
A next stage might be for sysops to opt in to register which Echos they carry. Additional info might be to add their up/down links. From that from the user perspective, it would be easy enough to find a place to pick up the conferences which they are interested in, especially those that don't have a wide distribution. It would also allow for some auto mapping of routes and flows of echomail that it not dependent on wading through all the various available echos.
There are numerous aspects to this that I have only barely speculated on, for example, cross referencing the echos with the Elist that Vincent maintains would be really helpful as well as maintaining cross-network links of echos and gateways.
As I said. Proof of concept. Putting my money where my mouth is. Exploring the options. Figuring out what works for users and what doesn't. YMMV. Caveat Emptor.
Net ter inligting, ek hou baie daarvan om met Nederlandse mense te praat in Afrikaans.. al voel dit vir julle as of die taal heel vreemd voorkom ;-)
All the best
John
===
* El Gato de Fuego (The Fire Cat) 4:920/69 * Pedasi, Panama
... Curiosity killed the cat, but for a while I was a suspect.
--- SBBSecho 3.14-Win32
* Origin: El Gato de Fuego - Pedasi, Panama (4:920/69)