Re: [Chicken-users] Eggology

[Ivan Shmakov]

>>>>> "SR" == Shawn Rutledge <address@hidden> writes:

[...]

 SR> I think code reuse is generally a good thing.  As long as there is
 SR> no circular dependency, what's wrong with depending on a few eggs?
 SR> It's better than rewriting the parts you need, right?  (and if your
 SR> few dependencies each depend on few more, pretty soon it's a lot of
 SR> eggs.  Oh well.)

        It seems not too hard to implement dependency resolving in
        software.  Surely, it shouldn't be necessary to download all the
        dependencies manually.

 SR> But it's important for popular eggs to be kept in a working state,
 SR> and not change existing APIs in them when avoidable.

        Could it be the reason to include version numbers into the .egg
        names, so that, when the interface changes, both LIBRARY-M.N.egg
        and LIBRARY-(+ 1 M).0.egg be available?

[...]

 SR> Maybe staged releases will become important... make a clear
 SR> distinction between the stable version of an egg and the
 SR> new/improved/under-development version.

        Including version numbers into the .egg names is the one way to
        achieve this.  (Version -> ``stability'' mapping is to be made
        available separately.)

        It's the way the Debian archive is organized.

 SR> When one egg depends on an older version of another (and will not
 SR> work with a newer version) maybe there should be a way to specify
 SR> that in the metadata, so that installing the one egg can
 SR> automatically install the older version of the other.  (Or can they
 SR> even co-exist?)

        It would be too infeasible in my opinion (apart from the ``major
        interface changes'' case above.)  ``Fixing'' the another egg to
        work with the new version of the changed one seems to be a much
        better solution.

 SR> Gentoo's portage system can do stuff like that.