[Distutils] Seeking the Essence of the Concept of an Egg

[Phillip J. Eby]

At 04:46 AM 2/28/2008 -0600, Jeff Rush wrote:
>In preparing for a tutorial on Python eggs, I'm wrestling with the 
>*precise* definition of a "Python egg".  It appears to be a slippery 
>term from the various documentation:
>
>   "There are several binary formats that embody eggs, but the most
>    common is '.egg' zipfile format, because it's a convenient one
>    for distributing projects. All of the formats support including
>    package-specific data, project-wide metadata, C extensions, and
>    Python code."
>
>Ok, so what binary formats -other than- .egg files are actually 
>eggs?  I figure RPMs, .debs, etc. are eggs, as long as they have an 
>EGG-INFO subdirectory and conform to certain basic structural rules.

No - they're not.  They *contain* eggs, they aren't eggs themselves.


>   "Eggs are pluggable distributions in one of the three formats
>    currently supported by ``pkg_resources``.  There are built
>    eggs, development eggs, and egg links."
>
>Hmmm, so it's not just binary distributions that can be eggs, but 
>also egg-links and development eggs.  The concept of a "built egg" 
>egg still includes an RPM or .deb though.

Development eggs are .egg-info eggs.  This definition *is* a bit 
slippery though, in that an egg link isn't "really" an egg, except in 
the sense that a symlink to a directory is a directory.  :)

If you want a solid definition, I'd say that built eggs (.egg file or 
directory) and development eggs (.egg-info eggs) are the only "real" 
eggs, as they are directly importable (as described below).


>   "Python Eggs are the preferred binary distribution format for
>    EasyInstall, because they are cross-platform (for "pure"
>    packages), directly importable, and contain project metadata
>    including scripts and information about the project's
>    dependencies. They can be simply downloaded and added to
>    sys.path directly, or they can be placed in a directory on
>    sys.path and then automatically discovered by the egg
>    runtime system."
>
>Well drat, so RPMs and .debs are -not- eggs, because you don't just 
>place them on sys.path directly to use them.  Egg-links don't quite 
>fit that definition either, unless you look at them funny.  So what 
>-other- binary format could that first paragraph be talking about 
>that can be dropped onto sys.path and be picked up by Python?

There isn't one now - it was phrased that way to leave open future 
possibilities.