Talk:O26 (text editor)

Who Wrote O26?

I have a copy of the source code for the SCOPE version of O26. The name at the very top of the listing is none other than Greg Mansfield (who wrote a lot of CDCs early OS software), and is dated December 9th, 1967. Dave Cahlander (another prominent CDC OS developer) added enhancements in the early revisions as well. It is all written in PP COMPASS to directly drive the 6612 Operators Console.

I have now read the Irons paper. He describes a completely different editor, which ran on completely different terminals (CDC 210s - specially modified with extra function keys), and is implemented completely differently from O26. It also ran on IDAs homegrown OS. --Wws 15:56, 12 September 2007 (UTC)

That's good - but it would be nice if there were some reference to point to, e.g., a published manual (even out of print) and a date. The original CDC topic, by the way, cited 1966, which I thought referred to the Irons implementation. (There often are dates cited in web discussion which I know to be incorrect - would be nice if Wikipedia at least can be made correct) Tedickey 16:59, 12 September 2007 (UTC)

Documentation for O26 was hard to come by. After all, it was an 'operators console', and programmers were not supposed to use it during 'production time'. (It tied up a PP and a big block of central memory.) So we usually used it during 'systems time', or for quick edits.  ;-) The later KRONOS and NOS operating systems had a manual called the Systems Programmers Instant - which had a couple of (very small) pages documenting the O26 commands. KRONOS came out around 1970 or so. I am not sure where it was documented in earlier SCOPE systems, other than the source code itself. --Wws 17:45, 12 September 2007 (UTC)

Well, thanks for the update. I only noticed O26 because I was looking for information to describe the SED program (so I could see why it was listed as a historical entry) Tedickey 17:52, 12 September 2007 (UTC)

Actually, thank you. I had never heard about the Irons editor before. So that was quite interesting.--Wws 18:01, 12 September 2007 (UTC)

Not first

The IDA editor is not first, contrary to what reference 1 claims. TECO predates it by at least two years, as shown by an MIT Project MAC memory from 1965. —Preceding unsigned comment added by Paul Koning (talk • contribs) 16:54, 12 September 2007 (UTC)

oh - did TECO run, displaying full-screen on a CDC machine in 1965? Tedickey 16:59, 12 September 2007 (UTC)

No, on DEC equipment. So that's not directly relevant to Cyber editors. My point is that the reference claims that the IDA editor was the first full-screen editor anywhere, and that is clearly false. Paul Koning 17:11, 12 September 2007 (UTC)

I'd qualify it as the first published implementation for a CDC machine. By the way, the TECO article on Wikipedia doesn't give a date (or reference...) for when it was able to edit full-screen. When I used TECO in the 70's, while it may have had full-screen capabilities, no one seemed to know that Tedickey 17:16, 12 September 2007 (UTC)

Noting the update for TECO - I only see the "v" command mentioned in the new reference, which might be what you're talking about. The description talks about a static display rather than a dynamic one (that one could edit a minibuffer). It would help if you pointed to the part that you're citing Tedickey 17:29, 12 September 2007 (UTC)

I'll add that to the TECO article when I get a chance. It's on page 9 of the MAC-M-250 memo, which says that it displays the region around the pointer. That makes it a full screen editor, because it tracks the state of the text as it changes. TECO-11 did the same (on a GT40) around 1974, and O26 did it too. All of these are command driven editors, but with a dynamic screen displays that always reflects the current state of the text buffer. Paul Koning 17:50, 12 September 2007 (UTC)

oh. I did use a GT40 (different editor), and agree it's a different thing than the Beehive that I used TECO on. For the Beehive, TECO definitely wasn't full-screen. But the GT40 has more capability - and requires less of the application to maintain the display. By the way, the comment that TECO ran on RT-11 was incorrect - it was a program with similar syntax but different commands, perhaps inspired by TECO (it was a nuisance since I used both concurrently) Tedickey 18:16, 12 September 2007 (UTC)

I define "full screen editor" as an editor that uses a CRT-type display to show the state of the editing buffer, updating it automatically to reflect the state of the editing buffer. In that sense TECO is one on at least the DEC VT52/VT100. Probably not Beehive, since that used a different screen command syntax which TECO would not know. TECO-6 is a full screen editor by that definition; I've updated the TECO article with references from 1964 and 1965 that say so. As for RT-11, you're mistaken; TECO most definitely was available there. Perhaps not supported by DEC, but I remember using it in college. The TECO pocket guide (from DEC, see the TECO article) clearly mentions RT-11. Paul Koning 18:44, 12 September 2007 (UTC)

I see (google says so too). I wouldn't have seen it, since it was unsupported. DEC's supported program had the same syntax, but wasn't TECO Tedickey 18:50, 12 September 2007 (UTC)

I have a copy of TECO 6 running in simulation using a 340 display, so you can try it for yourself a decide whether it's a full-screen editor or not.

Also note that the MIT AI lab updated their PDP-1 TECO to use a Type 30 display. This editor was the inspiration for TECO 6, or indeed ported rather directly from the PDP-1 to the PDP-6. Lars Brinkhoff (talk) 19:32, 14 March 2019 (UTC)