[plt-scheme] PlaneT, plts and distributing software.

From: Erich Rast (erich at snafu.de)
Date: Thu Oct 27 03:54:39 EDT 2005

While were at this topic, I'd like to ask a question that might have 
already been answered, but the answer may be somewhat burried in the 
depths of the current thread. Could someone who has the time perhaps 
summarize how to fulfill the following requirements:

Scenario: I have developed a MrEd program that I want to distribute as 
an independent executable for OS X and possibly other platforms. 

1.  The executable may never access the net or download anything else, 
i.e. the distribution is completely self-sufficient.

2. It must be assured that the distributed executable never loads any 
code of an eventual Plt installation, planet caches, collects, or any 
other code that has not been deployed with the distribution itself.

3. The original source code uses various Planet packages, i.e. there 
are planet requires scattered around in the source code at a vast 
number of places.

4. Installation effort for the end-user should be as minimal as 
possible and preferably not require admin privileges. In the ideal 
case, it would be an OS X `package'  binary that behaves like one file 
and can be installed via drag&drop.

5. Building the distribution from the source code should require 
minimum effort for me.

I understand that for OS X I'll have to distribute the frameworks as 
well and I might use a custom installer, but of course I still want to 
keep installation as minimal as possible. Furthermore, all source code 
or byte-compiled code of my app should remain in one directory. As for 
step 5, I'd like to avoid having to manipulate source code files 
extensively before sending them to mzc. Point 1 and 2 are mandatory 
security requirements.

So far I was assuming that using mzc --gui-exe would put all collection 
code including any planet code into the executable, but perhaps I 
wasn't reading the manual closely enough... ? :-O

Best regards,


Posted on the users mailing list.