[net.sources.d] Request for suggestions for next "less" release

casey@csustan.UUCP (Casey Leedom) (08/21/86)

Well, I appear to have been elected as less' new owner by a bunch of
people.  Since Rich Saltz posted my latest release of less I've received
several suggestions for new/slightly modified features and bug fixes.
Almost all of them were definite plusses.

What I'd like to do is install a bunch of these patches and then just have
Rich Saltz distribute a diff shar.  So if any of you out there have any
particular favorite patches, fixes, or suggestions, please send them
directly to me.  I won't hold a discussion on the net about what are and
what aren't Good Mods, but my judgement is pretty good.  I won't, for
instance, install an interactive sub-shell facility, nor will I make a Lisp
Eval environment available. :-)

Here is the current list of patches, bug fixes, and suggestions I've
received and thought about on my own.  Almost all of them *will* find their
way in unless someone just sends me a terribly convincing argument to the
contrary.  I will credit the source of any ideas when I make the next
distribution.  Thank you for your attention.

1. In certain circumstances when less' input is piped from another command
   and it's interrupted at the wrong time, less will continue running in the
   background sucking up large amounts of CPU time.

2. When less is given an invalid flag it should print a "usage" line.

3. When less is being applied to a single input (either piped or only one
   file is specified), it should simply exit immediately if the output is
   less than one screen.  Perhaps this should be a flag option "-E"?  Anyone
   have any negative feelings about this one?

4. When the "-e" flag is in effect ("exit less when the end of file is bumped
   into twice"), less should *not* "ring the bell" when you excersise that
   option.

5. When a file can't be opened for one reason or another less should say
   *why* it couldn't open it rather than simply printing "Cannot open ...".

6. Interrupt should kill less.  Perhaps controlled by a flag option "-I"?
   That is, interrupt should act as a preemptive quit ("q") command.  Anyone
   have any strong feelings on this one?
-- 
Leith (Casey) Leedom				lll-crg.arpa!csustan!casey
Computer Science Department			work: (209) 667-3185
California State University, Stanislaus		home: (209) 634-2775
Turlock, CA  95380