Re: sources and scripts for generated files

[Andreas Röhler]

On 26.05.2014 11:23, Stephen J. Turnbull wrote:
> Andreas Röhler writes:
>
>   > assume you are kidding.
>
> No, he's not.  GPL "sources" are defined as the preferred form for
> editing if you wish to make changes to the software.  If the .h file
> is as easy to edit as the other, it might be the preferred form for
> local changes (or at least acceptable).  On the other hand, it's
> converted by an automatic process *and* the script that does so is
> part of Emacs (AIUI).  So there's a strong presumption that the file
> from ICU is the source.
>
>   > If not kidding, I'm still wondering, resp. asking: please make the
>   > GPL a text every Emacs core developer may understand without
>   > undergoing expensive law-studies.
>
> That cannot be done; the GPL's primary purpose is to protect various
> interests in the covered Work, and readability to non-lawyers takes a
> back seat to that.
>
> But this part is not hard to understand.
>
> (1) You must provide sources that are reasonably convenient to edit.
> (2) Which do you think is more editable?  Ship it.

Must confess that point of view is rather new for me :)

Preferences might change and be close to tastes. Should really all the 
ephemeric helper-functions used in the edit-process need to be included?

Always understood GPL legal term "source" in sense of 
required-to-build-the-binary.
I.e. in sense of "Corresponding Source", where it's said:

  The "Corresponding Source" for a work in object code form means all
the source code needed to generate, install, and (for an executable
work) run the object code and to modify the work, including scripts to
control those activities.

;;;;;;;;;;

I.e. a script needed for checking the dependecies would be a required part.
Some Emacs-script for editing-convenience not - otherswise any Emacs-edited 
GPL'ed must come with all the Emacs-stuff alongside, etc. That would point into 
eternity.

Cheers,

Andreas