Thu May 30 20:58:21 UTC 2002
> From: Vernon Schryver > > From: <firstname.lastname@example.org> > > ... > > If anyone else is getting errors from those IPs please let me know. > `cdcc "host dcc.rhyolite.com; stats all"` suggests that all of the > servers with dcc.rhyolite.com A records are getting lots of them. > The most recent in my log is from 12:41:16 PDT. Sorry, I should have said: "If anyone continues to get errors..." > The current problem looks like a byte-order bug. 16777216 is 0x1000000. Ya, just noticed that. Fixed now. Thanks. > > As I mentioned before, I am doing a Windows port. Apparently I haven't got > > the packet format quite right yet. > > ... > As I've understood the word "port" when I've been paid to do it to > something, it involves lttle related to understanding and implementing > a protocol, not to mention misplacing an ntohl() in code that works > without changes on big- and little-endian systems. Perhaps I'm playing a little fast and loose with the term "port". I don't expect people are interested in the details but I've ported the checksum generation code and place it in a DLL. The network code I re-wrote. > The WIN32 port that I've had on my very long term to-do list since > the beginning involves approximately adding a modest amount of junk > to deal with Redmond's embrace-and-extend notions of sockets, a few > ifdefs, and a simplistic shell script to take the place of the autoconf > script and generate a fixed dcc_config.h file for the cross compiling. > Some of the worst of the work would probably be dealing with InstallShield > to make a tolerable package (as opposed to typical PC, trash-the-whole- > registry-and-c:/windows tactic). > Well, there's also the mess of dealing with WIN32 shared files instead > of POSIX for the client map. The "port" I have done is pretty specific to my needs. I am writing an email processing program that runs on the end user's computer, intercepts their mail and processes it (similar to Norton). This is a bit different from the server based system that you created. I am sure your code would port quite nicely. It's just that that would not fit my needs in this case. In any case: I have fixed the byte-order bug. If anyone continues to get bad packets please let me know. It is not my intention to make a pain of myself. Sean.
More information about the DCC