Lars> "John A. Martin" <jam(a)jamux.com&gt; writes: Lars> Er -- no. You mean not bad? The parody was supposed to illustrate application of Richard's principle of not allowing non-free software atop free OSes. Lars> I think that Richard is right in his assessment that the Lars> inclusion of well-advertised functionality for using ssh in Lars> Emacs will remove some impetus towards making a free version Lars> of ssh. I didn't even know that ssh wasn't free, and I Lars> doubt that many people do, precisely because it has become Lars> the de facto tool for encryption in free software circles. Well, I think it is cutting of our nose to spite our face. I don't know what can be done about people installing software without reading the copyright and license except as has been suggested in this thread to put up some unequivocal banner when they install the Emacs wrapper. However, we should also perhaps note that ssh1 is to all intents and purposes freely distribute and freely usable but cannot be offered for sale. Probably everyone on the list reading this agrees that that is evil. I happen to believe however that it is an even greater evil to put roadblocks in the way of easy use of strong encryption and especially when such roadblocks will inevitably encourage even more restrictive solutions including perhaps proprietary protocols and crypto software without source. Crypto applications without accessible source are of the very highest order of evil trampling not only upon free software but potentially also upon civil liberties. Lars> I don't quite understand what you wish to achieve by this Lars> escalation in rhetorics over a point of *strategy*. I did not intend an escalation but merely to identify the framework of the parody by which I hopped to cover a lot of ground without a lot of elaboration. Please see also my response to rms. Lars> And if not, they are not. Exactly. We run it up and see who responds. Lars> Some people are of the opinion that, say, support for audio Lars> devices is as important as freedom. I don't think so. Some causes are more equal than others. :-) Lars> The latter people. I think the opposite. This is based in part upon a reading of the ssh mailing list and the reception there to the restrictive ssh2 license. There is also a parallel with pgp2, pgp5, OpenPGP (rfc2440), and GnuPG. The pgp issues are more complicated but there remain many who eschew pgp5 that will almost certainly go to an OpenPGP when a conforming implementation is ready for casual use (GnuPG is pretty close). My guess is that most of the pgp2 hold-outs will migrate toward GnuPG while many of those that have embraced commercial versions of pgp5 may move toward the commercial implementations of OpenPGP that will presumably appear. I speculate that some users of "Free for Personal Use" pgp5 may be perhaps lean toward the commercial option because they are already comfortable doing crypto with something other than Free Software and with GAK/CAK (called "Big Brother Inside" by slb) to boot. jam