POSIX_ME_HARDER
karmak.org
karmak.org
* goes back to reading *
I'm only following about 1/3 of this, sadly. Anyone care to explain a bit?
It's particularly funny when Stallman takes pleasure in GNU Emacs highlighting some text, in the latest version.
All jokes aside, this is a fun read, and it gives you an almost softer side of Stallman, because no one is trying to attack him (except for one brief moment mid-article when they scuffle over Perl vs. Lisp and the fundament of lists relative to the languages.
This reminds one that Stallman is a great hacker and not just a campaigner.
He took over the original version EMACS after Guy Steele in particular negotiated a common set of Editor MACroS and key bindings, and in the long term he added some serious value to it.
The current EMACS we use (GNU or X) started from a fully functional version of an EMACS written for the VAX in C by this obscure guy named James Gosling. Even had a bytecode that its Mocklisp compiled into.
RMS started with a copy of that under ... questionable circumstances ... and substantially improved it, e.g. giving it a full although unfortunately dynamically scoped LISP.
I'm not sure who wrote the first Lisp Machine EMACS, named EINE (Eine Is Not Emacs, a popular naming scheme at the time), but as I recall Dan Weinreb did a lot of ZWEI (Zwei Was Eine Initially). It's likely RMS contributed to it, he made quite a few contributions to the Lisp Machine's software.
So what you mention can be summarized: Steele was also significant in early history (before Unix implementations) Gosling made the first Unix implementation (but proprietary one) and there were also Lisp Machine implementations.
But Stallman was not only involved in Emacs before Unix, he was the guy who managed to keep Emacs as widely usable as we know it today by creating the free license and the free (GNU) Emacs.
So I still don't understand what your "not any version of it" was supposed to mean.
Many people contributed to TECO macros, to the point where there were three or more major versions and people had the problem of not being able to necessarily sit down at a colleague's terminal and type into their editor.
Guy Steele spend a week or so obtaining order out of chaos (biggest issue was negotiating the common key bindings, something I assure you RMS is not diplomatic enough to do :-) and near the end of this Stallman joined and then took over the project and most especially its leadership. For Steele this was just a problem to be solved and once it was in good hands he had other fish to fry (e.g. Scheme).
Another semi-correction to Wikipedia is that there were one or more ersatz UNIX EMACS versions floating around before Goslings' (ersatz by RMS's terminology; they had the basic command set and key bindings but if you needed to change or add anything you had to dive into the C code. This was a necessary limitation for running on the PDP-11 with its severe 64K of code and 56K of data limits). There was also a serious 32 bit version built on top of ex as I remember, but it stayed commercial and therefore died.
My "not any version of it" refers to your "As far as I understand, the guy made Emacs ... practically alone initially." Which to me implies that he started the projects or at least was an initial team member, which was not true for either the TECO or 32 UNIX versions. Instead, he took over existing healthy versions and made them much better. By the time I showed up on the scene, he was emphasizing self-documentation and extensibility, the latter of which was by and large never offered except through a LISP. (Perhaps a bit pedantic, but I'm a scientist by training so getting attributions of this sort correct is important to me.)
SINE may be an early exception since it was a custom LISP like language for building its EMACS, but its community was so small (the Media Lab precursor) I'd doubt any normal mortals played with it. Whereas Bernie Greenberg was intensely proud about how he managed to make MULTICS EMACS so accessible that MIT secretaries were doing their own extensions, generally started from a key sequence capture.
Bernie was big on "LISP for the masses" (along with a number of others in this period); as I recall, he was the author of a very good medium length (less than 100 pages) general introduction to the language, Notes on the programming language LISP and frequently taught a course during January based on it ("LISP is is for 'interesting' programs, as opposed to, say, accounts receivable." (rough paraphrase of the start of his motivation for learning it; everyone knew it was the cool language at the time and he was great at expressing why and why you might find it useful)).
Note that he's talking about Emacs running in a terminal, not the windowed version.