On Mon, Oct 08, 2001 at 09:05:09AM -0500, thelema wrote:
< >
> Yup. All correct. What I'm proposing doesn't go against any of this.
> I'm suggesting that *for only the ref-tester*, that it be seperate from
> the node. The reference tester only needs to access the same classes as
> a real node, so that its method of handshaking and such is consistent
> with a real node's method. It doesn't need access to any internal data
> structures, you just give it a reference, and it connects to those
> nodes. (Maybe if your own node's info is required for handshaking, it'd
> have to read the store to get at that data, but it'd be using *all
> existing code* to do this.) It'd just take the results of that code and
> handle it differently.
Well, it seems to me that the best way to check if the node can connect
to a reference is to have the node attempt to connect to it. In practice
it may seem that just making the connection from the same machine covers
all the bases, but I don't like depending on the abribrary
characteristics of current networking protocols - it's better to do
stuff in a manner that is axiomatically certain.
> If there's more information required to do the simple handshaking and
> authentication, I'd like to know so I can revise my position on this,
> but from what I see, there's no need to make it part of the node, so
> there's no need to bolt it on.
At the moment nodes accept connections from anyone, so a test would not
really require using the actual nodes identity, but that may not always
be the case.
--
Though here at journey's end I lie
In darkness buried deep, above all shadows rides the Sun
beyond all towers strong and high, and the Stars forever dwell:
beyond all mountains steep, I will not say the Day is done,
nor bid the Stars farewell.
(JRRT)
Oskar Sandberg
oskar at freenetproject.org
_______________________________________________
Devl mailing list
Devl at freenetproject.org
http://lists.freenetproject.org/mailman/listinfo/devl
