Re: NoCatSplash v0.90 released (finally)
Schuyler Erle <[email protected]>
| Newsgroups | gmane.network.nocat |
|---|---|
| Message-ID | <[email protected]> |
* On 22-Nov-2004 at 8:20PM PST, Scott C. Lemon said: > > I am *very* interested in the Passive Mode operation. You seem to be aware > of potential problems ... Can you give me any ideas of why you feel it is > "probably broken"? What is it that makes you suspect? Have you tested it > much? Well, it still forks out to GnuPG. So you need to have GnuPG on the same platform. I think all the code is written, but I never got done testing it, really, and I suspect it just needs to be debugged. Also, it blocks waiting to hear back from GnuPG, which you don't want to do in a single-threaded network server, but there are a couple of examples of setting up event callbacks with g_io_add_watch() scattered throughout the code base. > I just downloaded the code and am going to start looking ... I'm hoping that > the compile-time option is obvious! ;-) ./configure --with-mode=passive The README may be of service, as well as ./configure --help. > Also ... Are you suggesting that I run this against a v0.82 auth server? That was the idea. Probably someone should write a different data sharing protocol (at worst just using an external stunnel or something) and do away with GnuPG entirely. GnuPG might have been a nice idea when we hoped people would have gateways talk to multiple auth services. But OTOH we can still do that with outgoing HTTPS connections, or whatever. I didn't want to require the authserver to maintain state, but that was as much laziness on my part as anything, I think. SDE