On Thu, Sep 18, 2003 at 12:24:25PM +1200, David McNab wrote: > On Thu, 2003-09-18 at 03:31, Toad wrote: > > > It is a critical node bug - a complete showstopper for anyone using > > > splitfiles over FCP. > > > > No it isn't. It works perfectly well locally > > Wrong - two other people are reporting the bug in localhost connections.
Hrrm. What versions? > > > and anyone advanced enough > > to be using FCP from another machine is probably capable of setting up > > an SSL tunnel. > > I'd use the Python remote object broker software I've hacked up. > Or other solutions. > > But, stepping back a little, why have I and others run into SEP-shields, > or received HTFS responses? Sorry, I do not know what HTFS means. SEP - no, it IS our problem, it's just that we have problems coming out of ours ears, and we have to deal with one thing at a time. > > <rant> > Freenet is a phenomenal, revolutionary concept. The project is now in > its 4th or 5th year. Yet it still seems to be sidelined and regarded as > a 'technical plaything'. Little by little, reactions of denial, 'HTFS', > 'NMP' etc tend to reinforce this reputation. > </rant> Bugs in Fred are my problem. > > When people confront me with bugs in my code that I can't reproduce on > my gear, I litter my code with all kinds of diagnostic logging and send > them a custom cut. They send me the logs. Repeat cycle till I know > what's going on, fix, get them to send patch, until bugs go away. Yes indeed, I am familiar with this technique. But for this particular bug it will be easiest to fix if I reproduce it locally, which should be easy enough. > > Why I feel all this to be an issue is that with (reportedly) build 5020, > all was fine, and with (reportedly) 5021, things broke. So someone broke > something, albeit in a way which doesn't manifest for all users. > > > > In the meantime, I might recommend that the 'official release' be > > > temporarily reverted to build 5020. > > > > That is completely overreacting. > > I'm not sure you'd feel that way if the bug was happening to you. > I still believe that FCP client programmers deserve at least to see a > link on the website mentioning the problem, and making available a > freenet.jar build which doesn't have that problem. Even if not that, > they deserve some kind of warning that 'some of the FCP FEC commands may > result in blocked sockets and send/receive failures'. Only if you run it from non-localhost. If you do that you're smart enough to set up a tunnel. > > -- > Cheers > David -- 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://dodo.freenetproject.org/cgi-bin/mailman/listinfo/devl
