Re: Platform Dependent Code

["J.J. Sestrich" <jsestri2@uiuc.edu>]

dhruva wrote:
> Hello,
>
> On Sat, Jun 21, 2008 at 4:56 AM, Chong Yidong <cyd@stupidchicken.com> wrote:
>   
>> <jsestri2@uiuc.edu> writes:
>>
>>     
>>> How much of emacs is platform dependent code? I am wondering how much
>>> code has to be replaced if you wanted to port emacs to a new platform
>>> essentially: How much of emacs is native c code / libraries and how
>>> much is done through lisp code?
>>>       
>> Almost all the platform-dependent code is in C and concerns
>> terminal/window system communications.  See the w32* and mac* files in
>> the source tree for an example.  The rest of the work is a matter of
>> juggling compilation switches, identifying system call names, etc.
>>
>> (If you provide some information about what platform you are interested
>> in, maybe someone could be more helpful.)
>>     
>
> I feel the most complicated would be the generating the emacs
> executable by dumping a process with elisp code loaded
> (bootstrapping). I remember when working in the OpenVMS labs, on
> Itanium platform, that was the only missing piece in getting a full
> blown working emacs.
>
> -dhruva
>
>   

I was thinking about making a javascript implementation that could be 
used in a browser. It would use some cgi scripts to do file I/O and 
window stuff would probably be easy enough to do with html/javascript. 
At this point, I'm just trying to gauge the level of work that would 
need to be done to make it happen. I would imagine that in addition to 
the windowing and file I/O I'd have to re-write the lisp interpreter, 
but this might be easier to do in javascript than C. I don't really know 
enough about this to judge anything though. Does anyone have any insight 
on these ideas: Size of code to be re-written? Difficulty of code that 
has to be re-done?

J.J.