The Origin of XEmacs -- Richard Stallman XEmacs is a forked version of Emacs, maintained by a group that does not cooperate with the GNU Project. The roots of XEmacs go back to 1990, when the programmer in charge of maintaining Emacs told me he was leaving, giving 2 weeks notice. He had been hired away by Lucid, which wanted him to add certain features to Emacs. The job he walked away from was getting Emacs 19 ready for release. I was too busy at the time to do it myself, so I had to move someone else onto the job. That was Jim Blandy, who had started a few weeks before to write a GUI desktop for the GNU system. (This was our first attempt to start such a desktop project. GNOME, started in 1997, was our third.) Emacs is a large program, and Jim was not familiar with the code. He needed a year to learn to work on it efficiently, and he needed my help, which I could only give occasionally. He eventually became good at the job, but ultimately the Emacs 19 release was delayed until 1993.(1) Meanwhile, the previous Emacs maintainer brought a copy of the unreleased Emacs 19 sources to Lucid, and started to add the features Lucid wanted; but he left (I don't know why) without finishing the job. In 1991, a Lucid executive asked another group of Lucid's programmers to add features to Emacs. They started with the unreleased Emacs 19 sources which they already had. They did not even inform me that they had started working on major Emacs changes; I found out about their project only through gossip. Around January 1992, I phoned Jamie Zawinski, who was one of the group, and he described what they were doing. Then I said, "Let's discuss these design decisions," but he said it was too late for discussion; they had already halfway implemented the ideas, they had a deadline, and they intended to finish the project without ever accepting input from me. He continued by saying that they wanted me to put all of their additions into Emacs, just as they had written them. But if I wanted to use only some of their extensions, or wanted to change aspects of it, they were not going to help with the adaptation. He said we would have to do that on our own. In effect, they had decided to barge in on Emacs development and declare themselves in charge, and they sincerely expected me to meekly accept this. I rejected that, of course, but I tried not to let that prejudice me against their work. I agreed with some of their ideas, and disagreed with others. I found some parts of their code clear and understandable, while other parts I found incomprehensible. We did not use their entire changes "as is". True to their word, they refused to help us merge parts of their code into Emacs. But Lucid did sign the legal papers our lawyers say we should ask all contributors to sign, clearing the way for us to merge in the parts we liked from their code. So we merged some of that code, and we found volunteers to adapt some other parts. Some of those parts are in Emacs today, though some have been changed greatly since. A few months after that phone call, the team at Lucid released a modified version of our Emacs 19. Their release came before ours. Based partly on their features, and partly on the Emacs 19 features that they had got with our unreleased sources, they started aggressively asking users to switch to Lucid Emacs (subsequently renamed as XEmacs). They treated the GNU Project as a competitor, and have treated us thus ever since. Zawinski describes all this as "attempts to help the FSF" (http://www.jwz.org/doc/lemacs.html). If helping us was their intention, they disregarded the first principle of trying to help a project: to cooperate with the people already working on it. His account of the history is misleading when it says they "reached the point where [they] just couldn't wait any longer" so they "bundled up" and released what they had. Those words seem to imply a period of trying to work with us, and followed by a decision to make a separate release only at the end of that period. The facts, stated above, are otherwise. They may have envisioned that we would adopt their entire changes, but the full extent of their cooperation with the FSF, during the entire period of development before the release of Lucid Emacs, consisted of explaining the design plans to me during the phone call that I had initiated. Some users argue that the competition between Emacs and XEmacs is beneficial, resulting in more progress on Emacs features. That argument ignores the fact that the human resources have been taken away from other work. What if Jim Blandy had provided the GNU/Linux system with a user-friendly graphical desktop in 1995 or so--how would that have have altered our community's history? What if I could have written some other program, or helped the fight against repressive laws such as the Digital Millenium Copyright Act, instead of maintaining Emacs? That might have done some good. Overall, we would have accomplished much more by working together, so it is unfortunate that the Lucid team decided not to do this.