Thanks for that explanation, Guy. It doesn't sound too fun to contemplate working on. I'm not all that familiar with sockets, but assuming they work relatively similar you're still probably talking a lot of cluttering #ifdefs in the code - both emacs and emacsclient.
(And Jason, I'd missed seeing the -eval flag in emacsclient and thought it only dealt with sending files whereas gnuclient could send lisp expressions to evaluate, but emacsclient does do that. And it has the moderately cool "alternate editor" flag.) But, now that I've got the newer tcp/ip gnuclient working with the CVS 23.0.0.1 ntemacs build, I'm happy. I work with software all day so the fact that gnuclient isn't an official part of Emacs and needs installed separately doesn't bother me (though projects like Lennart's EmacsW32 make that pain even less I think). So if the big problems are that gnuclient's not officially part of Emacs and hard to get it so, and emacsclient won't work without a fair amount of effort, I'm guessing there's not a lot of "pull" to make Windows developers want to make emacsclient work, and I suspect Windows isn't a really important OS target for Emacs developers so there's not a lot of "push" to make them want it done either. Lennart, you mentioned a windows-port list - is that another email list for developers working on a new port or just maintaining and extending the current Windows-code in Emacs? Rob
