I won't revert a patch to add FCP noderef adding/listing support... but I still think if it's just for the proposed matchmaking site or the public ubernode, then it's a mistake to do it this early.
On Sun, Jun 18, 2006 at 10:34:26AM +0100, Dave Baker wrote: > I believe I probably speak for most of the Freenet hacker community when I > say > that I just think the devlopment effort that you're putting in to this would > be so much better spent on having the functionality available over FCP, > rather than building a, as you say yourself, slightly hackish library over > the TMCI. > > All the features you describe sound useful, so I'd encourage you to write the > functionality you need into the node instead, which will be much nicer, more > stable, and we'll all benefit from your code. Especially so as you say you'd > be writing it in Java anyway. I appreciate that it involves extra time > finding your way around the Freenet codebase, but it's not that hard, and > ultimately worthwhile. > > > All I'm trying to do is to add a noderef to my OWN node.. I would > > hope that that wouldn't be so tough ;) > > Totally. I'm also of the opinion that this should be part of the FCP, but > there is debate on this, so if you do decide to implement this in FCP, I'd > have a chat with Toad first. > > > Dave > > > > On Sunday 18 June 2006 09:46, Colin Davis wrote: > > I want to create this for several projects that I'm thinking about > > internally, and I figured it might be useful to other people as well. > > > > > > I understand that you're busy, that's why I figured writing a bridge > > from Java to the TMCI would be helpful. I'm not asking you to do it- > > I'm trying to figure out the best way for ME to - I want to be clear > > on that point ;) > > > > I just figured some other people might find it useful. > > > > > > > > > > > > > > > > One- If could export the peer list to Java easily, I could keep > > track of how often a particular peer is connected, versus backed off, > > or the like. This would allow me to create a Peer rating, similar to > > the Seti ratings. Then I could see which peers were often around, > > which weren't/ > > > > Two- I run a DYNDNS client on my home PC, but it doesn't work very > > well. When I want to use terminal services to connect to it, I find > > that the IP associated with the hostname is often wrong. That said, > > It does run a freenet node, and so does my server. The IP address > > that my server has for my home node (determined via ARK) is /better/ > > than the DYNDNS address. > > Using TMCI, I can get this address, so I can connect to my home > > machine... I realized that I could expand this concept- I could allow > > anyone who wanted to to register with me, and I'd track their IP > > address, as long as their node was connected. > > > > Three- When you say Node Autoloader, I'm not sure you understand what > > I'm doing... I'm not dealing with SinnerG style autoloaders. > > I want to add a noderef to my own node, via a PHP script. Right > > now, the best way to do this is via a curl request to fproxy... > > That's insane! As you've pointed out, fproxy isn't a stable > > interface. I've had to modify the script every few days to continue > > to have it work.. > > All I'm trying to do is to add a noderef to my OWN node.. I would > > hope that that wouldn't be so tough ;) > > > > Four- I'd like to use the noderef data to script my node to > > automatically remove nodes that haven't connected in a week.. I don't > > want to have people that have given up on Freenet wasting resources. > > I'm doing that manually now, but I'm happy to script it using Java/ > > TMCI... > > > > > > Again, I'm just working on some personal projects, but I thought it > > might be useful to other people.. Before I got started, I thought it > > best to make sure no one else was doing it. > > > > I know my reasons probably aren't the best ones for exciting > > developers, but I'm trying to build things on freenet... And these > > are the tools I need to build them.. I can't do it via my primary > > choice, which is a Library... So I'm falling back to doing it via > > TMCI, and writing a wrapper lib around that... > > If that fails, I'll probably write a lib around HTTP scraping, the > > way I'm doing it now... > > > > I know it's messy, but unless you tell me that what I'm doing is > > hurting freenet, I'd rather keep going. > > If you think I'm hurting the network, I'm willing to discuss it, and > > ensure I'm not doing further damage. > > > > -Colin > > > > > > > > On Jun 18, 2006, at 4:07 AM, Lars Juel Nielsen wrote: > > > > > On 6/18/06, Florent Daigni?re (NextGen$) > > > <[EMAIL PROTECTED]> wrote: > > >> * Colin Davis <[EMAIL PROTECTED]> [2006-06-18 02:27:29]: > > >> > > >> > I'm considering starting a project to create a new library, but > > >> > wanted to make sure no one else was doing it first. > > >> > > > >> > Currently, there exist a number of options that exist via TMCI, but > > >> > not via FCP. Examples include getting a list of connected nodes, > > >> > adding peers, etc.. > > >> > > > >> > I'm interesting in creating a new library which emulated these > > >> > functions in Java, by connecting to the TMCI server, and then > > >> > formatting them, and returning them. > > >> > > > >> > For example, I could (pseudo code follows) > > >> > > > >> > Node = new connection(127.0.0.1,2323); > > >> > Array NodeNames[] = Node.getListOfPeers(NodeName); > > >> > Array IPAddresses[] = Node.getListOfPeers(IP); > > >> > > > >> > Node.AddConnection(URL, "http://ubernode.org/ref/ref.txt";); > > >> > Node.DeleteConnection(NodeName, "sanity-darknet"); > > >> > > > >> > > > >> > > > >> > etc... > > >> > > > >> > > > >> > Is anyone working on this presently? Is there any reason I > > >> shouldn't? > > >> > > > >> > > >> It should be done in FCP, we all agree on that, including toad. The > > >> thing is we don't want to ease the development of references auto- > > >> adder > > >> so the priority of the task 'implementing the FCP feature of exposing > > >> the peer list through FCP' is low. > > >> > > >> Exposing the configuration framework through FCP is also > > >> planned ... If > > >> you want us^toad to raise the priority of that task, find good > > >> arguments, keeping in mind that anything with 'auto-adder' in the > > >> name > > >> won't please developpers. > > >> > > >> Developping YetAnotherLayer/Library isn't a good idea imho > > >> > > >> NextGen$ > > >> > > >> > > >> -----BEGIN PGP SIGNATURE----- > > >> Version: GnuPG v1.4.1 (GNU/Linux) > > >> > > >> iD8DBQFElQhvU/Z/dHFfxtcRAlqKAKD9ilmqsL9Qo+iTuee9w8/59160ZACgyNgh > > >> 4XQ1sqlQriHDzyyDZ38gf6g= > > >> =fL56 > > >> -----END PGP SIGNATURE----- > > >> > > >> > > >> _______________________________________________ > > >> Devl mailing list > > >> [email protected] > > >> http://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl > > >> > > >> > > > > > > Not to mention that the TMCI interface won't be there forever. > > > _______________________________________________ > > > Devl mailing list > > > [email protected] > > > http://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl > > > > _______________________________________________ > > Devl mailing list > > [email protected] > > http://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl > > > _______________________________________________ > Devl mailing list > [email protected] > http://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl > -- Matthew J Toseland - [EMAIL PROTECTED] Freenet Project Official Codemonkey - http://freenetproject.org/ ICTHUS - Nothing is impossible. Our Boss says so.
signature.asc
Description: Digital signature
_______________________________________________ Devl mailing list [email protected] http://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl
