mg@unipress.UUCP (required by law) (04/04/88)
I had sent this a couple of weeks ago but it apparently never made it... Yes, I'll put in a plug for Emacs, too. It's definitely the most powerful development tool for NeWS (or for most anything else). Although some people balk at having such a heavyweight behind something like a file browser (Emacs is *not* tiny), I think it's well worth it. All you need is one Emacs session running to handle many such things at once. Besides, the sort of people who run NeWS in the first place are the sort who would run Emacs as well. I literally consider a machine to be unusable if it doesn't have Emacs. (Emacs has a pig reputation in some circles, dating from the days of 40-user VAX/780s; if someone disparages Emacs, ask them when they last used it and where.) Emacs is a very nice environment for developing under NeWS. You can run a NeWS server listener in a window, which is akin to running psh in a psterm window, except that listeners are editable with command recall, etc. You can edit PostScript code in another editor window, and load it (or some part of it, such as the nearest 'def') into the NeWS server listener with two keystrokes. This allows very fast prototyping (as long as you don't confuse the server too much with bogus code). The editor has a mode designed for editing PostScript code, which is sensitive to comment/code contexts, so you can fill a paragraph within a comment, for example. There is a database of descriptions of the PostScript and NeWS operators, so you can see a description of something by mousing it. You can add descriptions of your own things very easily. In itself, Emacs is a programmable working environment which is very effective at gluing programs together. It provides a ready-made, user-customizable/extensible front-end and command dispatcher. It includes a set of subsystems, including a customizable on-line documentation reader and a file-system browser, that are useful in many applications. It is possible, for example, to build a sophisticated front end to a dumb program (i.e., one with a stone-age user interface, like raw dbx) very quickly. If there is any interest, we'll provide an ed(1) interface later on. You can write programs in (around?) Emacs, using it as a "user interface toolkit". It provides text editor windows and listener windows, which present an interactive program (such as a shell) in an editable window, so you can review the session, recall previous commands (a la Korn shell), cut and paste between other listener/editor windows, write listener sessions to files, stuff files to the programs, etc. Since it's Emacs, you can extend it with keyboard macros (somewhat like event journaling but much easier to handle; journals become commands that you can bind to keys) and extension language code. If you do things right, your program can be usable from a dumb terminal as well as under a window system without your having to rewrite a lot of UI code. There are a few programs around these days that call themselves Emacs. The comments above apply to the bigger ones (not so much to Jove or uEmacs). We (UniPress) sell a commercial version of Emacs originally written by James Gosling, who also wrote NeWS. GnuEmacs is a free (but not public domain) version also descended from Gosling's Emacs. The two have diverged considerably in the 3-4 years since then, but they have roughly equaled each other in capability (although you can easily start religious wars over that). The last time I looked ours had more native window-system support (e.g. using multiple window-system frames), a somewhat fancier text editor, and was smaller and a bit faster. I don't know what's happened to GnuEmacs since then -- anyone care to fill that in? (Pardon my natural bias, but I also find the UniPress code more readable and easier to work with.) However, GnuEmacs has a real Lisp as its extension language, with all that implies, because its post-Gosling developers were heavily under the influence of MIT. (The MLisp extension language in UniPress Emacs resembles Lisp but isn't.) GnuEmacs doesn't directly cost much money if you know where to get it (I think $150 from the Free Software Foundation). If either will do what you want, GnuEmacs is good if you want a free Emacs or if you want real Lisp; UniPress Emacs is good if you don't mind paying (it's $395 binary+MLisp source, $995 for that plus C-source), or if you want someone to be accountable for fixing any bugs you find, or if you want to embed it in a commercial product (the Gnu license forbids the latter). Also, unlike FSF, we LIKE NeWS. The Emacs we're shipping now (V2.15) has only rudimentary but adequate NeWS support; you can use the mouse to point at things and drag modelines and text around, but it uses only one NeWS window for its editor "windows". We are currently developing a version of Emacs that uses NeWS to advantage, using multiple NeWS frames (i.e., windows) and many other neat interface objects (if you're proficient with the keys you can use Emacs as a window manager, and seldom need to use the mouse for anything at all). It's in its first round of user testing now, but it won't be generally available until this summer (that's "software developer" summer; sorry I don't have a more exact date). We'll be at Usenix in June, where there's bound to be a NeWS BOF. ------ Mike Gallaher Emacs Hacker Boss UniPress Software
randy@NCIFCRF.GOV (04/04/88)
> (I think $150 from the Free Software Foundation). If either will do > what you want, GnuEmacs is good if you want a free Emacs or if you > want real Lisp; UniPress Emacs is good if you don't mind paying > (it's $395 binary+MLisp source, $995 for that plus C-source), or if > you want someone to be accountable for fixing any bugs you find, or > if you want to embed it in a commercial product (the Gnu license > forbids the latter). Quick clarification: The GNU license does *not* forbid embedding gnu-emacs in a commercial product. What it does say is that if you ship gnu-emacs with your operating system, you must either 1) Ship the source code, or 2) Make your customers aware that they can get the source code for a *small* (tape handling) fee. (You also cannot restrict redistribution *of gnu-emacs itself*). If you develop a program starting from the gnu-emacs code, it stays covered by the provisions I mentioned above, but other software shipped with gnu-emacs does not. > Also, unlike FSF, we LIKE NeWS. Now *that* is a major plus for Unipress! > ------ > Mike Gallaher > Emacs Hacker Boss > UniPress Software -- Randy Smith (I in no official way speak for Freesoft and the above simply represents my understanding of the situation. If you are interested in doing this, talk to them and their lawyers. If you need to sue somebody over what I've said, sue me, not them) Randy Smith @ NCI Supercomputer Facility c/o PRI, Inc. Phone: (301) 698-5660 PO Box B, Bldng. 430 Uucp: ...!uunet!ncifcrf.gov!randy Frederick, MD 21701 Arpa: randy@ncifcrf.gov
bob@allosaur.cis.ohio-state.edu (Bob Sutterfield) (04/07/88)
In article <8803291504.AA08952@unipress.uucp> mg@unipress.UUCP (required by law) writes: > [...a bunch of other glowing stuff nobody can really argue with :-) > [...also some impressivly balanced comments about how his own > [...product compares to GNU Emacs...] >We (UniPress) sell a commercial version of Emacs originally written >by James Gosling, who also wrote NeWS. GnuEmacs is a free (but not >public domain) version also descended from Gosling's Emacs. All Gosling code has been excised in recent versions. >...GnuEmacs doesn't directly cost much money if you know where to get >it (I think $150 from the Free Software Foundation). Repeat the mantra: "That $150 is a tape copying fee." If you don't want to pay FSF to spin tapes for you, you can get it via anonymous FTP from prep.ai.mit.edu, or via anonymous UUCP from osu-cis. >Also, unlike FSF, we LIKE NeWS. This is a FSF attitude that I entirely disagree with. Chris Miao added support for NeWS 1.0 in Emacs 18.49, and Clayton Elwell fixed some stuff up for NeWS 1.1 in Emacs 18.50. I understand that Chris's changes were offered back to FSF as a contribution to their efforts, but were turned down because of NeWS' proprietary status. RMS, why not accept the changes on the same basis that VMS and SunView are included in the distribution: work that willing contributors offered, in order to give the product a wider audience, in those mean, nasty, proprietary non-standard environments? Yes, I know that RMS doesn't read this newsgroup, but if we get everybody singing the first four bars... he'll think it's a movement! >Mike Gallaher, Emacs Hacker Boss, UniPress Software -=- Bob Sutterfield, Department of Computer and Information Science The Ohio State University; 2036 Neil Ave. Columbus OH USA 43210-1277 bob@cis.ohio-state.edu or ...!cbosgd!osu-cis!bob