.PS 55,78,5,0.Title Runoff Future Plans
.C;^&FUTURE RUNOFF PLANS\&
.S2
.AP
The current version of RNO is currently lacking a number of features which
are in DSR or which the user community would dearly like to have.
Here is a partial list which may be added in the future, or which
have been contemplated and are rejected for the immediate future.
.s
.list 1
.LE;PROPORTIONAL SPACING - This is where each character occupies a different
spacing. This is a really big ticket item and while I have
considered it, the problems are large.  If supported it might not
be possible to do it for any arbitrary printer.  It is definitely not planned
as an enhancement in the immediate future.  Some of the pieces necessary for
it are already included, but implementation of this feature might not be
device independent.
.LE;Hyphenation dictionary.  This is the only means of properly
hyphenating all words. This will only be available if I have help in creating
a hyphenated dictionary. 
.le;Better hyphenation algorithm.  This would be nice if improvements
could be made without too much extra memory needed.  I am open
to suggestions from other users.
.LE;The remaining missing DSR commands are either conveniences or unnecessary
bells and whistles.  Some or all could be added with little overhead, but
since they can generally be done without they just might be.  They include:
.list 0
.display elements ll,')'
.LE;AUTOJUSTIFY - Maybe, but why ?
.LE;AUTOTABLE - Probably not
.LE;CONTROL CHARACTERS - Not needed at all
.LE;DATE / NO DATE - Probably not (may be simulated)
.LE;NO SUBTITLE - Maybe (same as .SUBTITLE with no text)
.LE;NUMBER RUNNING - This is irrelevant and unnecessary!!
.LE;ENABLE/DISABLE BOLDING + FLAGS BOLD - Too much overhead.
.LE;FLAGS PERIOD - Probably not
.LE;FLAGS COMMENT - Probably not
.LE;REPEAT - Probably not
.LE;SET DATE/TIME - Probably not (simulated by _.DELETE/DEFINE SUBSTITUTE)
.LE;SET LEVEL - Maybe
.LE;XLOWER/XUPPER - Maybe
.LE;Double column index - Only as general multiple column package.
This is difficult to mix with escape sequences, equations and half spacing.
Printer independence would be difficult to maintain.
Several ways of doing this are possible.  Each has distasteful consequences.
.list 0
.display elements RL
.LE;Coordinated postprocessor - The post processor must know all about
possible escape sequences.
This project would be a good one for someone else in the user community.
.LE;Backspace to top of document for each column - This would not
be available for ordinary printers.  Some printers can not handle
such violent paper movement.
.LE;Keep internal 1 page buffer - This would be a major logic change
which would require too much work.
.els0
.ELS1
.le;Expanded _.DEFINE COMMAND might be considered.  Because of the
various levels of nesting, parameter passing while nice might be
very difficult to implement.  The original define command was essentially
a freebe as part of _.DEFINE SUBSTITUTE.    Parameters if passed would 
initially be numeric or string and would be labeled %1,%2...%9.  String params
would have to be enclosed in quotes, and text lines would be passed
only as an entire line ie. %A for whole line.
.le;A set of options may be added to _.TRACE to aid in debugging RNO
input files.
.le;A /VARIANT switch would probably be a difficult item.  This
may be accomplished other ways.  For example input files containing
the appropriate variant definitions may be inserted before the main file.
.els
.text
.r;John Clement
.r;Bonner Lab
.r;Rice University
.r;Houston, Tx.
.end text