chocolate-doom split

[Simon McVittie]

On 04/05/15 13:46, Fabian Greffrath wrote:
> Am Montag, den 04.05.2015, 13:10 +0200 schrieb Alexandre Detiste:
>> I think you lost the doom-wad-shareware dependency in the
>> chocolate-doom package split.
> 
> It's a provider of "doom-wad" and a non-free package that I do not want
> to (alternatively) explicitely depend on anymore.

We've been trying to have games' data package names as produced by g-d-p
represent a playable game that makes sense to users, rather than an
abstract concept that makes sense to developers. The full version of
Doom is a game, which you might reasonably want to play; the shareware
version of Doom is a (short) game, which, again, you might reasonably
want to play; but the abstract concept of "a WAD compatible with Doom
engines" makes very little sense to me. If I want to play Doom, I don't
consider (for instance) Chex Quest to be an acceptable substitute, and
vice versa :-)

If you want to insist on doom-wad being a virtual package name
designating "any Doom-compatible wad", then we could rename the real
package that represents "the gameplay, plot, graphics etc. of id
Software's 1993 FPS game" to doom-data or really-doom-wad or some such,
or we could have the virtual package be called doom-compatible-wad...
but I really think those two concepts deserve to be distinct, and I
don't think having a name for (the thing that we could call)
doom-compatible-wad necessarily makes sense.

>> - heretic-wad | game-data-packager,
>> + heretic-wad | heretic-shareware-data | game-data-packager,
> 
> Why doesn't heretic-shareware-data already provide heretic-data?

Because it is neither the same game (it's a subset) nor an
interchangeable "ABI" (one is heretic.wad and the other is heretic1.wad).

    S