On Saturday 02 February 2008, Grant wrote: > You're right, access to the printer can be given only to certain > hosts. So simply using 'lpr file.pdf' on the remote machine doesn't > strike you as a bad idea?
Lets look at this from the perspective of what is really going on. You have a process on one machine that opens a high numbered port to knock on a low numbered port on another machine and conduct a TCP/IP session. Data moves up and down blah blah blah. The process on the first machine just happens to be lpr, and the port on your machine just happens to be 631. Here's another scenario: You have a process on one machine (which just happens to be Firefox) that opens a high numbered port to knock on a low numbered port (which just happens to be port 80) on another machine and conduct a TCP/IP session with the process listening on port 80 which just happens to be Apache. Data moves up and down blah blah blah. How are these two things different in any fundamental way? They are not. Gladly setting up say Apache and also being hesitant about setting up a print server is totally inconsistent (and yet you would be amazed at the amount of clueless knuckleheads out there advising exactly this attitude). The only reason I would not do that remote printing setup is if I knew of specific weaknesses/exploits in lpr and CUPS. But I don't. -- Alan McKinnon alan dot mckinnon at gmail dot com -- gentoo-user@lists.gentoo.org mailing list
