chris@mimsy.UUCP (Chris Torek) (12/17/87)
Completion is nice. It lets one abbreviate; one need not remember the exact spelling of some command or option; and so forth. Completion is bad. The moment there is a new command or option, the old subnames no longer work. (Emacs---Gosling and GNU, at least---does completion, and I expect a number of people have experienced both the good and the bad.) -- In-Real-Life: Chris Torek, Univ of MD Comp Sci Dept (+1 301 454 7690) Domain: chris@mimsy.umd.edu Path: uunet!mimsy!chris
ralphw@TEMP.IUS.CS.CMU.EDU (Ralph Hyre) (12/17/87)
In article <9827@mimsy.UUCP> chris@mimsy.UUCP (Chris Torek) writes: >Completion is nice. It lets one abbreviate; one need not remember >the exact spelling of some command or option; and so forth. Agreed > >Completion is bad. The moment there is a new command or option, >the old subnames no longer work. This is why I always hit <ESC> or <space> as part of the abbreviation. If there's a new command which makes the abbreviation ambiguous, the system beeps at me. In any event, having completion is ALWAYS better than not having it. Without completion, you don't even have the choice of having abbreviations, unless you want to define an alias for everything. You still have the new command problem, though. It's interesting to thinkg about completion interacting with a spelling-correcting (DWIM, type 'sl' and you get 'ls') front-end.
howie@cunixc.columbia.edu (Howie Kaye) (12/20/87)
Completion should allow abbreviations to be set(in the command tabes), so that old commands and their abbreviations will comtinue to work. COMND on the TOPS20 did this. CCMD does also. ------------------------------------------------------------ Howie Kaye howie@columbia.edu Columbia University hlkcu@cuvma.bitnet Systems Group ...!rutgers!columbia!howie
simon@its63b.ed.ac.uk (ECSC68 S Brown CS) (01/11/88)
In article <535@PT.CS.CMU.EDU> ralphw@TEMP.IUS.CS.CMU.EDU (Ralph Hyre) writes: >It's interesting to thinkg about completion interacting with a >spelling-correcting (DWIM, type 'sl' and you get 'ls') front-end. I guess that a completion would have to be a lot more intelligent - in addition to adding extra stuff on the end, it could change stuff already typed to get a "better match". Of course, you might well like it to change some of the already-typed data even though this is in fact already a "valid" initial part of a word, because it's not one you use at all often. So, you really need *all* the exciting user-interface toys to interact with each other for maximum affect - hopefully in a programmable way so that you can define in advance the various inter-relations between them. A Really Good interface would have completion, alternatives-listing, history, job-control, windows, help, DWIM, etc... all mixed up together, rather than existing as completely separate abstractions - but without losing any structure. --Simon. -- -------------------------------------------------- | Simon Brown | | Laboratory for Foundations of Computer Science | | Department of Computer Science | | University of Edinburgh, Scotland, UK. | -------------------------------------------------- UUCP: uunet!mcvax!ukc!lfcs!simon ARPA: simon%lfcs.ed@nss.cs.ucl.ac.uk "Life's like that, you know" JANET: simon@uk.ac.ed.lfcs