Re: LZ: Shells


[Prev][Next][Index][Thread]

Re: LZ: Shells



Damien wrote:
> 
> Jeffrey S. Sharp, Crystal Baird, and Mark Sharp wrote:
> 
>      Michael Wyman wrote:
>      >
>      > I was
>      > thinking about setting up an IRC time to meet... About
>      making sure that
>      > all shells have a few common grounds
> 
>      OS-85, SuperNova, ZShell, CShell, OShell, whatever-shell,
>      blah blah
>      blah.  What we have here is a *failure* *to* *communicate*.
>      IRC chat
>      sessions would be great.  But what we really need is to make
>      a
>      standard.
> 
>      Here's an idea.  Let's imagine a new and wonderful shell.
>      Call it, say,
>      OpenZ.  Here's how everything would work:
> 
>      (1) You have *one* shell.  No one will write another shell.
> 
>      (2) You have *one* standard for shell-writing on the 85.
>      Call it, say,
>      The OpenZ Standard.
> 
>      (3) The OpenZ shell will include *all* options set forth in
>      the
>      standard.  This includes APD, contrast, an Expnader SF
>      filesystem,
>      function keys, interrupt piggybacking routines, everything.
> 
>      (4) OpenZ is distributed as _commented_source_code_.  Also
>      included in
>      the distribution packet will be an assembler and an install
>      program
>      which will automate *everything*.
> 
>      (5) The commented source code will have preprocessor
>      directives that
>      divide OpenZ into several modular parts.
> 
>      (6) At install time, the user selects which components he
>      wants.
> 
>      (7) The install routine will, using equates to activate the
>      preprocessor
>      directives, assemble a shell with only the components the
>      user selected.
> 
>      (8) The installer will transfer the shell to the calc.
> 
>      (9) Anybody can add to or modify the shell as long as it
>      stays
>      backwards-compatible.  It's an open standard.  When someone
>      does this,
>      he or she should release an updated standard.  A
>      version-control method
>      would be nice here.
True, but it sounds like you're forgetting that this is a CALCULATOR;
_NOT_ an entire computer system.  You've got some good ideas, but you're
also forgetting the main method of transport for these programs--THE
TI-85.  VERY few people have a GRAPHLINK, and those that do are not ALL
as technical as we are.
Bottom Line?--WE HAVE TO KEEP THIS SIMPLE AND EASY-TO-USE.

Once again, I think that a Shell-Developers list would be helpful...

Cheers,
-- 
~Keith
TSK3000@Prodigy.Net


Follow-Ups: References: