[racket] Post package installation

On Fri, Nov 8, 2013 at 12:01 AM, Laurent <laurent.orseau at gmail.com> wrote:
> On Fri, Nov 8, 2013 at 2:12 AM, Jay McCarthy <jay.mccarthy at gmail.com> wrote:
>>
>> There is not. I think you should add something into your documentation
>> that says
>>
>> "Run racket -e '(require laurent/init)'"
>
>
> Side question: Is there a difference with `racket -l laurent/init`?

I don't think so

>>
>>
>> after installing.
>
>
> Ok.
> To remove weight from the user's shoulders, l prefer to use
> `install-collection' and I'll try to avoid repeating the same questions on
> each setup.
>
>>
>> In my mind, this was not an oversight in the design,
>> but a purposeful thing.
>
>
> Do you currently have an explicit rationale for this, maybe?

I can't think of a use-case that isn't covered by having a dependency
or a slightly different packaged program.

For instance, if you need to compile some C code, then you should
compile it for your users and put it in the package (or as #:platform
dependenices) and when those fail on the platform use the ffi-lib fail
argument to throw an error like "Go to directory ... and type make"

For instance, if you need to have the user create a configuration file
so that when you run your main program, your program could check for
that config when it starts and build it there. (A good example in my
mind is s3cmd)

I like the idea of common file utilities like zip, tar, etc being able
to completely put package files into place and how we could smash
everything into a single collects directory. But this is a less
important thing to me.

Jay

-- 
Jay McCarthy <jay at cs.byu.edu>
Assistant Professor / Brigham Young University
http://faculty.cs.byu.edu/~jay

"The glory of God is Intelligence" - D&C 93