Diff for /gforth/ToDo between versions 1.3 and 1.8

version 1.3, 1995/02/02 18:12:58 version 1.8, 1997/05/21 20:39:16
Line 13  direct/indirect, NEXT splitting, keeping Line 13  direct/indirect, NEXT splitting, keeping
 **make it easy to put the right variation for each processor into the  **make it easy to put the right variation for each processor into the
 configuration. I.e., on installation all combinations of options  configuration. I.e., on installation all combinations of options
 should be measured and the fastest chosen. Knowing OS and architecture  should be measured and the fastest chosen. Knowing OS and architecture
 is not enough, the best otions depend more on the processor and the  is not enough, the best options depend more on the processor and the
 compiler version.  compiler version.
   
 * ANSI Forth  * ANSI Forth
Line 37  Use gcc-generated assembly on machines w Line 37  Use gcc-generated assembly on machines w
 processors supported by gcc  processors supported by gcc
   
 *Foreign Language Interface  *Foreign Language Interface
 If anybody wants to do this, mail me  If anybody wants to do this, take a look at
 (anton@mips.complang.tuwien.ac.at). I have given some thought to this  ftp://ftp.complang.tuwien.ac.at/pub/forth/foreign.ds
 and will type in my notes if you are interested.  Some of the problems are discussed there,
 **C  **C
   Stuart Ramsden is doing a bit here.
 **FORTRAN  **FORTRAN
 **C++  **C++
   
Line 53  Issues: Convenience, portability across Line 54  Issues: Convenience, portability across
 existing tools (Emacs, F-PC)  existing tools (Emacs, F-PC)
 ** Decompiler and Debugger  ** Decompiler and Debugger
 need debugging  need debugging
   ** Profiling
   The way this (and perhaps also debugging features) could work is this:
   On compilation all code fields are remembered somewhere (using a
   special hook like etags). If the user now decides to profile part of
   the code, the corresponding code fields are replaced by fields
   pointing to code that performs the measurement (or whatever else is
   intended).
 ** emacs support  ** emacs support
 can be improved  can be improved
 **prefix file generator  **prefix file generator
 A tool for generating a prefix file for a program that explains in  extend or complement ans-report.fs to provide Forth definitions for
 what way the program conforms to ANSI (i.e., which wordsets are used)  the simple non-ANSI words.
 and contains Forth definitions for the simple non-ANSI words.  
 ** rightcase  ** rightcase
 A tool that converts all uses of words in a source text to the exact  A tool that converts all uses of words in a source text to the exact
 case of the definition. There's something like this out there on the  case of the definition. There's something like this out there on the
 net (Joerg Plewe has posted a reference), but I think a program that  net (Joerg Plewe has posted a reference), but I think a program that
 wires itself into the compiler (like etags.fs) is harder to fool by  wires itself into the compiler (like etags.fs) is harder to fool by
 search order tricks etc.  search order tricks etc. EVALUATE is a problem.
   
 *Object-Oriented Extensions  
 John Hayes has a portable package  
   
 *Documentation  *Documentation
 A texinfo file  A texinfo file
   
 **glossaries of all wordsets.  **glossaries of all wordsets.
 *** Inclusion of stack comments, glossary comments, and wordset comments  *** Inclusion of stack comments, glossary comments, and wordset comments
 in all source files.  in all source files.
   
 * Distribution and Announcements  * Distribution and Announcements
 ** add copyright notices to all the source files  
 ** Write articles for (general-purpose) magazines  ** Write articles for (general-purpose) magazines

Removed from v.1.3  
changed lines
  Added in v.1.8


FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>