From paul@darkstar.bos.locus.com Tue Sep 19 00:24:28 CDT 1995 Article: 116298 of alt.folklore.computers Xref: uchinews alt.folklore.computers:116298 alt.sys.pdp10:782 Path: uchinews!vixen.cso.uiuc.edu!howland.reston.ans.net!news.sprintlink.net!in1.uu.net!info-server.bbn.com!news5.ner.bbnplanet.net!news3.near.net!news-server.bos.locus.com!not-for-mail From: paul@darkstar.bos.locus.com (Paul Cantrell) Newsgroups: alt.folklore.computers,alt.sys.pdp10,alt.lang.teco Subject: Re: TECO on non DEC platforms Date: 18 Sep 1995 17:52:06 -0400 Organization: Locus Computing Corp Lines: 57 Message-ID: <43kpm6$nku@darkstar.bos.locus.com> References: <42ikac$e2p$1@sydney.DIALix.oz.au> <42tc2p$l92@dns.plano.net> <431t93$6lc$1@perth.dialix.oz.au> NNTP-Posting-Host: darkstar.bos.locus.com Status: RO In article <431t93$6lc$1@perth.dialix.oz.au>, Skipton Ryper wrote: >Did teco ever get ported to other platforms (no DEC)? >Is there a public domain version? My Video TECO runs on almost any Unix system you might care to name, plus VMS. It really needs virtual memory, so I've never ported it to personal computers. If you want to look at some documentation for it, look at the following url: http://world.std.com/~paulc/teco.html This document is huge, I have not had the time to break it up into smaller web pages, so be aware that it will take a little time to load. But if you look at "1.4. Screen Layout" there is a GIF file showing you what the screen looks like when you are editing. Basically, this is a TECO optimized for screen editing, allowing you to edit multiple files at once, and files can be very large assuming you have the virtual memory to support that :-) The main characteristic that sets this editor apart is that you edit in TECO commands (which echo at the bottom of the screen) and you simultaneously see the changes to the file. For instance, if you are editing and you type the letters "h" and "k", you see "hk" echoed at the bottom of the screen, and you see the contents of the buffer dissappear. If you rubout the "k", the contents of the buffer re-appear. The level of undo is unlimited, back to the last double escape (and double escape is almost never required - I edit for many hours at a time without $$ so that in theory I can undo back through the entire session). It is not absolutely faithful to original TECO, so old macros are unlikely to run, but it is actually based on the DEC TECO manual, so most commands are 100% faithful. However, whenever it came to being faithful versus working the way I wanted, I changed it to work the way I wanted. That was the whole reason I wrote my own TECO in the first place - so I could make it work the way I wanted! The biggest change is that lines are terminated with a single rather than . This made sense because I was writing this under unix. I like it that way, but some people may object. You can split the screen into multiple panes which can point into the same buffer, or different buffers (look at section 7.10 for an example). "n" searches cross buffer boundaries, so you can load the entire source of the unix kernel into the editor, and then a single "n" search will search through all the files until it finds your string. Q registers are implemented as normal edit buffers, so you can jump into them and edit the contents of the Q-registers. Anyway, if you are interested check out the www page. I plan on putting the source on the web server in the near future, it's currently available by anonymous ftp from ftp://ftp.wimsey.com/pub/teco/cantrell-teco Paul -- THIS | Paul Cantrell paul@bos.locus.com BRAIN | URL: http://world.std.com/~paulc INTENTIONALLY | DoD #1144 LEFT BLANK | HELO/ASEL/ZX10 Pilot From clements@bbn.com Tue Sep 26 15:01:39 CDT 1995 Article: 117202 of alt.folklore.computers Xref: uchinews alt.folklore.computers:117202 alt.sys.pdp10:812 Path: uchinews!vixen.cso.uiuc.edu!uwm.edu!spool.mu.edu!bloom-beacon.mit.edu!newsserver.pixel.kodak.com!news.sprintlink.net!in2.uu.net!info-server.bbn.com!clements From: clements@bbn.com (Bob Clements) Newsgroups: alt.folklore.computers,alt.sys.pdp10,alt.lang.teco Subject: Re: TECO on non DEC platforms Date: 25 Sep 1995 17:59:09 GMT Organization: Bolt Beranek and Newman (BBN) Lines: 32 Message-ID: <446qld$oto@info-server.bbn.com> References: <43kpm6$nku@darkstar.bos.locus.com> NNTP-Posting-Host: lion.bbn.com Status: R In article (a fine long Message-ID if ever there was one) Mark Crispin writes: >TECO was always this way at MIT, from the first PDP-1 >implementation that Dan Murphy wrote. Nobody there ever thought that TECO >would be useful without a display. >Bob Clements generated the first Teletype TECO for the DEC PDP-6 monitor; >I don't know why. Perhaps they didn't let him have a display monitor... Pretty much true. I TOPS-6-ized the MIT TECO in my spare time while sitting around at SAIL during the installation of their PDP-6. Mainly to have a better editor than EDLIN (or whatever) to use while I was there. I happened to have the TECO source with me at the time. That would have been -- guessing -- August 1966? The only terminals available to me at the time were TTY-33s and TTY-35s. And on another point, someone recently spouted a silly rumor that a "JRST @." would hang a KA-10 and asked for an old-timer to refute it. I (project engineer for the KA-10) refute it, of course. I posted to that effect a week or so ago, but posts from here weren't getting out for a while. >-- Mark -- /Rcc Bob Clements, K1BC, clements@bbn.com, (w) +1 617 USE K1BC From jebright@magnus.acs.ohio-state.edu Mon Oct 2 11:28:14 CDT 1995 Article: 117634 of alt.folklore.computers Xref: uchinews alt.folklore.computers:117634 Path: uchinews!vixen.cso.uiuc.edu!sdd.hp.com!usc!math.ohio-state.edu!magnus.acs.ohio-state.edu!jebright From: jebright@magnus.acs.ohio-state.edu (James R Ebright) Newsgroups: alt.folklore.computers Subject: QED wes:Re: Illiac II & 7094 music Date: 29 Sep 1995 05:01:34 GMT Organization: The Ohio State University Lines: 60 Distribution: na Message-ID: <44fuje$lkm@charm.magnus.acs.ohio-state.edu> NNTP-Posting-Host: top.magnus.acs.ohio-state.edu Status: R Oven in another thread... Douglas W. Jones,201H MLH,3193350740,3193382879 wrote: ... >For those not in the know, the SDS 940 was the result of some work at >Berkeley by folks like Butler Lampson. They took an off-the-shelf SDS 930 >and added virtual memory addressing to it. The 930 and 940 were wonderful >24 bit single-accumulator single-index register machines, and the software >that came out of Berkeley for the 940 included the Berkeley Timesharing >System, the QED text editor, a version of the DDT debugger, a FORTRAN II >compiler, a LISP interpreter, the CAL programming language (The California >Algorithmic Language), and perhaps XTRAN (a superset of a subset of >FORTRAN IV). CTSS also had QED... 1,$s/vi/qed/ which was an editor written by Ken Thompson of Bell Labs. I 'imprinted' on QED :) Ken supposedly wrote the editor in one of those fits of continuious programming that so often characterizes the hacker. I am still amazed at the number of things 'modern' editors can't do that were easy (or, at least, more obvious) in QED. It was a good editor, especially for the times. (Somewhere in the attic I still have the Project TIP QED Manual... one of the more useful documents Project TIP produced for the MIT community.) One of our folks, Lew Morton (last known doing contract work for a UK firm in .fi) tried to fix some bugs one day... A quick but dirty job of programming is what he found... ...Interesting that it was ported to the SDS line. I bet QED is not running anywhere today... But, amazing as it seems, I just today found 'a piece of code' as old as QED that is still running. MIT's Civil Engineering (Course One) Department had CESIL (?? Civil Engineering Systems Integration Lab??) under Daniel Roos. It came up with a 'language' for doing survey plots. COGO (COordinate GeOmetry) was almost ideal for laying out subdivisions. I found it still being used as an extra on a $285 AutoCad-based Windoze package... I remarked to my fiancee who was with me in the bookstore and who knows nothing about computers that when I was at MIT it took a million dollar computer (IBM 360/30 then later /40) to do a simple plot... now it is a 'throwaway add-on' for PCs... a durable package! :) "I don't know what languages will be like in the twenty-first century. All I know is one of them will be called COBOL" - Capt. Grace Hopper. -- A/~~\A 'moo2u from osu' Jim Ebright e-mail: ebright@bronze.coil.com ((0 0))_______ "'Eternal Vigilance Is The Price of Liberty' used to mean \ / the \ we watched the government - not the other way around." (--)\ OSU | - Bill Stewart From tom_van_vleck@taligent.com Mon Oct 2 11:28:57 CDT 1995 Article: 117672 of alt.folklore.computers Xref: uchinews alt.folklore.computers:117672 Path: uchinews!vixen.cso.uiuc.edu!howland.reston.ans.net!tank.news.pipex.net!pipex!news.sprintlink.net!in1.uu.net!taligent!tvv.taligent.com!user From: tom_van_vleck@taligent.com (Tom Van Vleck) Newsgroups: alt.folklore.computers Subject: Re: QED wes:Re: Illiac II & 7094 music Date: Fri, 29 Sep 1995 09:18:38 -0700 Organization: Multicians Lines: 30 Distribution: na Message-ID: References: <44fuje$lkm@charm.magnus.acs.ohio-state.edu> NNTP-Posting-Host: tvv.taligent.com Status: R James R Ebright wrote: >CTSS also had QED... Ken Thompson's QED had commands based on 940 QED, but Ken added regular expressions to the search and substitute commands, and added many commands. The CTSS editors that came before QED began with line-number oriented editing, in which you could only replace a whole line by typing its number, and had to exit the editor to look at the file being edited. Art Samuel, the IBM researcher famous for an early checkers-playing program, visited Project MAC in 1963 and wrote a much better editor called CTEST9 (don't ask) which had a "peek" command that let you look at the file while editing it. Another much better command called EDIT was then produced by the MAC folks; and then QED came along and became the power users' editor. QED's descendants survive in UNIX: ed and sed. The big difference between QED and ed is that QED supported multiple "buffers," each containing a file, and you could insert the contents of one buffer into another. And a buffer could contain editor commands, which functioned as macros. A search that failed exited a buffer expansion, so data-dependent branching was possible. Writing "programs" in this macro language was possible: you could do arithmetic by character substitution. Somebody at BTL, maybe Doug McIlroy, wrote a set of macros to play tic-tac-toe. Martin Richards visited Project MAC and brought BCPL with him, and several Bell Labs folks got interested in the language, including Dennis Ritchie and Rudd Canaday. A version of QED was written in BCPL, and was ported to the fledgling Multics system. It was pretty slow: enough so that Bob Daley wrote a cut-down version in PL/I called qedx, which is still in use today at the surviving Multics sites. More on Multics editors at http://www.best.com/~thvv/mgq.html#QED From atbowler@thinkage.on.ca Wed Oct 4 00:09:56 CDT 1995 Article: 118148 of alt.folklore.computers Xref: uchinews alt.folklore.computers:118148 Newsgroups: alt.folklore.computers Path: uchinews!vixen.cso.uiuc.edu!news.cts.eiu.edu!news.ecn.bgu.edu!news.moneng.mei.com!daily-planet.execpc.com!spool.mu.edu!howland.reston.ans.net!plug.news.pipex.net!pipex!dish.news.pipex.net!pipex!tank.news.pipex.net!pipex!oleane!simtel!news.sprintlink.net!news3.insinc.net!grumpy.insinc.net!thinkage.on.ca!atbowler From: atbowler@thinkage.on.ca (Alan Bowler) Subject: Re: QED wes:Re: Illiac II & 7094 music Message-ID: Sender: news@thinkage.on.ca Organization: Thinkage Ltd. References: <44fuje$lkm@charm.magnus.acs.ohio-state.edu> Distribution: na Date: Tue, 3 Oct 1995 18:47:11 GMT Lines: 41 Status: R In article <44fuje$lkm@charm.magnus.acs.ohio-state.edu> jebright@magnus.acs.ohio-state.edu (James R Ebright) writes: >Oven in another thread... > >Douglas W. Jones,201H MLH,3193350740,3193382879 wrote: >... >>For those not in the know, the SDS 940 was the result of some work at >>Berkeley by folks like Butler Lampson. They took an off-the-shelf SDS 930 >>and added virtual memory addressing to it. The 930 and 940 were wonderful >>24 bit single-accumulator single-index register machines, and the software >>that came out of Berkeley for the 940 included the Berkeley Timesharing >>System, the QED text editor, a version of the DDT debugger, a FORTRAN II >>compiler, a LISP interpreter, the CAL programming language (The California >>Algorithmic Language), and perhaps XTRAN (a superset of a subset of >>FORTRAN IV). > >CTSS also had QED... >1,$s/vi/qed/ > >which was an editor written by Ken Thompson of Bell Labs. I 'imprinted' >on QED :) > According to Butler Lampson, QED originated with the 940 project. The QED's on other systems are descendents (reimplemenations) of that one. It was a "Quick EDitor" because it was originally quickly put together because the planned editor was not ready when it was needed by other members of the project. This was reimplemented in BCPL for CTSS, Multics, Gcos. The Gcos version was converted to assember, and the use of regular expressions for patterns was added. (Gcos QED compiled machine code for pattern matching.) The Gcos QED evolved over time and piece by piece had a complete rewrite including a replacement with a faster pattern matcher not bassed on the finite auatomaton approach used in the Thompson/Ritchie version. This editor is still alive an well on Gcos today and called FRED. Unix "ed" and "ex" are sort of pale reimplementations of the Gcos QED lacking the full regular expressions and programming capability. Tom Duff (of "Duff's device") did write a Unix QED in C, that does have comparable pattern matching and programming capabilities. If this had bee available sooner it is not clear that SED and AWK would ever have evolved. When those type of tasks come up on Gcos, they are almost always done with a Fred program. From dnichols@d-and-d.com Sat Apr 29 12:29:42 CDT 1995 Article: 100386 of alt.folklore.computers Xref: uchinews alt.folklore.computers:100386 alt.sys.pdp10:566 comp.sys.acorn:50138 Newsgroups: alt.sys.pdp10,alt.folklore.computers,comp.sys.acorn Path: uchinews!vixen.cso.uiuc.edu!uwm.edu!news.alpha.net!news.mathworks.com!gatech!newsfeed.pitt.edu!uunet!ceilidh!dnichols From: dnichols@d-and-d.com (DoN. Nichols) Subject: Re: Belling the cat (was Dumb Command Names (was Retro-Computing!)) Message-ID: <1995Apr28.023030.9627@d-and-d.com> Sender: usenet@d-and-d.com (Usenet) Nntp-Posting-Host: windigo Reply-To: dnichols@d-and-d.com Organization: D and D Data, Vienna VA References: <1995Apr20.071429.14317@ivax> <3njqtn$8tr@nntp4.u.washington.edu> Date: Fri, 28 Apr 1995 02:30:30 GMT Lines: 47 Status: R In article <3njqtn$8tr@nntp4.u.washington.edu>, David W. Barts wrote: [ ... ] >Guessing how to use the editor was more difficult; the name itself -- >ed -- wasn't hard to guess, but it was hard to tell I had guessed it >and that it was waiting for editing commands to be input without >prompting for them. What a yucky design decision not to prompt for >input. The worst use of ed(1) that I have encountered was in the AT&T Unix-PC, a 68010-based machine with a bitmapped monochrome display, and an early windowing front end aimed at suits. Normal purchasers would get the "Foundation Set", and purchase the applications which they intended to use with it. The development set was an extra-cost option. However, the manuals provided with the foundation set were feel-good "click on this and watch this happen" type manuals. The *real* unix manuals were shipped with the development set. (*And* on-line manuals were *not* offered for that machine. Since the first of them shipped with a 10MB HD, this was perhaps semi-reasonable. Vi(1) (and ex(1)), were also in the development set, as part of the "advanced editors" package. However, if you *didn't* have the development set, it would not bother asking you about your default editor, since ed(1) was all there was. Now the box was set up to do reasonable things when you clicked on the menu line for some item you were interested in -- in *most* cases. However, if you clicked on a plain text file, without the execute bit set, its idea of what the *reasonable* thing was was to fire up ed(1) on that file. Nowhere in the "manuals" which came with the foundation set is even a *mention* of ed(1), much less how to use it, or even recognize when you were in it. It just started a window, ran ed(1) in it, and loaded in the file you had clicked on. Then, if you had no previous experience with unix, (and remember that this was marketed for suits), you had the fun of bashing at the keyboard till you accidentally hit something which might get you out, usually saving the mess that you had made of the file in the process. I wonder how many people this soured on unix through the years? :-) -- Email: | ...!uunet!ceilidh!dnichols Donald Nichols (DoN.) | Voice (Days): (703) 704-2280 (Eves): (703) 938-4564 --- Black Holes are where God is dividing by zero ---