Ugh, I really hate personal attacks (as they make my argument look weak) so I'll temper that with real world examples of why you are wrong as well.
First of all, I have seen you make some completley hairbrain assumptions on this list in the past (personaly, your hit/miss ratio with me is about 1/5) and I am definitly going to call this one hairbrain to the tune of 100%. Assuming that my internet connection is failing when I get RNF is retarded. Do you honestly expect that someone who can atleast code a little java is not able to detect the difference between a working internet connection and a broken one? I allmost feel insulted, or atleast I would if I did not consider you an allmost total retard. I assure you that when I get RNF my internet connection is working as I get messages to the tune of Number of nodes attempted: 45 Number of nodes that were unreachable: 40 Number of nodes that restarted: 2 Number of nodes that flat rejected: 3 Of course that is not cut/paste from my node's error page but you get the idea. Add to this the fact that I am constantly browsing and on irc and you can safely know that my internet connection does not fail on a regular bassis. If anyone else is really worried about the posibility of someone constantly refreshing a page when their node can't talk to the rest of the world it would be a simple mater to make the node not generate the refresh tag when it detects the "your request could not even make it off your node" state - however that happens in 2 scenarios, one where your internet connection is down and another where it is up but each and every node rejects your connection. Considering the fact that when your connection is down your node will send no data to the rest of the network is it really a big issue to have it refresh even up to 5 million times a second? Now consider the other posibility, every other node rejects your connection - in this case you DEFENITLY want this patch as you would be stuck sitting there clicking the retry button constantly. Your assertion that "If a page has DNFed at HTL=25 after 3 or 4 attempts,it should be not-retrievable" is also completely wrong. If you bothered to take my advice and try the patch out (btw, you should actualy try the patch before you go spout your mouth off about how bad it does not work. Number of times this has been stated now: 2) you would know that information can be "brought back from the dead" quite easily. As for your idea of the number of requests growing over time I agree with you. The number of requests will be anywhere between 1 every 2-10 minutes (on average) times the number of browser windows each individual has open times the number of individuals utilizing the patch minus the people that get their request fulfiled because of the refreshing (which will be alot). This number will grow over time but so will the number of nodes. I don't have the numbers to make a formula to see which will grow faster so I'll leave that up to Matthew T. to use his judgement seeing as I trust his idea of how freenet operates much beter to yours. Where are the problems? I don't see any.... Tyler PS: I doubt highly this list wants to see any more of our argument as I forsee it quickly degenerating to fluff and flame war so I'm not going to bother responding to your response unless it has some objective data that I feel is wrong. Your "gut" does not hold enough credibility with me to be considered as valid input when I leverage the past posts I have seen you make. --- Edgar Friendly <thelema at mercury.truman.edu> wrote: > Tyler Riddle <triddle_1999 at yahoo.com> writes: > > > This is only a bad idea if you enjoy having RNF > and > > DNF ruin your freenet using experience. I suggest > you > > actualy try the patch and see how effective it is > at > > pulling pages through that DNF at HTL 25 after 3 > or 4 > > attempts (just leave your page open over night and > you > > run a good chance of having it waiting for you in > the > > morning). > > > First, if your node is RNFing, retrying the request > is among the last > things you want to do. RNF can be a sign that your > internet > connection is down (and your node can't contact any > others to forward > requests to). When this is the case, > auto-requesting until you find a > key will not only never find the key, but it'll > empty your routing > table as every node in it fails connection > requirements time and > again. > > If a page has DNFed at HTL=25 after 3 or 4 attempts, > it should be > "not-retrievable", and given up on. If the key is > in freenet and > isn't being found, the solution is *not* to add a > auto-retry "feature" > into fproxy; that's just patching over the symptom > of the problem. > The real problem is that the key isn't being found, > and it's this that > needs to be worked on, either by tweaking routing > table settings so > that nodes (in general) use more information to > route, or by making > requests more lightweight so that higher HTLs can be > supported. > > > > Regarding the "flood the network" switch, we > allready > > have frost which does a plenty good job of > flooding > > the network. I'm no freenet expert but I suspect > the > > ammount of trafic this will generate (one request > > every couple to 10 minutes at high HTL) is minimal > > compared to the load that frost places on the > network. > > > > Tyler > > <SNIP> > > Frost is *not* distributed as part of the node. > Yes, many people are > using frost to flood the network, but that doesn't > mean we should make > people flood the network each time they browse to an > edition-based > site. Yes, I'm talking about the links to future > editions, which will > set fproxy working forever trying to find data > that's not in the > network. As for your characterization of "one > request ever couple to > 10 minutes", I see that number easily becoming > *much* larger after a > short amount of browsing. > > Thelema > -- > E-mail: thelema314 at swbell.net > Raabu and Piisu > GPG 1024D/36352AAB fpr:756D F615 B4F3 BFFC 02C7 > 84B7 D8D7 6ECE 3635 2AAB > > _______________________________________________ > devl mailing list > devl at freenetproject.org > http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/devl ===== AIM:rllybites Y! Messenger:triddle_1999 __________________________________________________ Do you Yahoo!? Yahoo! Mail Plus ? Powerful. Affordable. Sign up now. http://mailplus.yahoo.com _______________________________________________ devl mailing list devl at freenetproject.org http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/devl
