[Scummvm-devel] state of the Mohawk engine

[A. Milburn]

Hi,

I think it might be a good time to write a little to the list about the
state of the Mohawk engine. As mentioned in the recent thread about
what's ready to be released, Living Books v1/2 games should be ready
(and clone2727 enabled Mohawk in trunk for them, although we need to
disable all the other Mohawk games by default..). I hope that sounds
reasonable, although I have no idea what the usual procedure is for
this kind of thing and will happily leave it all to other people.

Some notes about specific subengines:

* Riven: Support is coming along really well. I have successfully
completed the whole game in ScummVM, and at this point it's mostly just
missing some video code, although clone2727 tells me it's quite tricky
to implement those last pieces and so it's indeed not likely to be ready
for the next release.

* Myst: I know nothing about the progress of Myst, but judging by the TODO,
it's not quite as far as Riven yet. :)

* Living Books: I've been working on support for this, and as clone2727
mentioned in the release status thread, the v1/v2 games ought to be ready
for the release. My 'goal' for these games is to implement enough of the
engine to support the LB v4 engine used in the Rugrats Adventure Game, but
that is still quite a distant goal - the scripting of the engine became
quite complex in v3/v4, hopefully the just-announced Living Books testing
should make sure that the basics of the engine are stable before we move
on with that.

* CSTime: I've also implemented most of an engine for Where in Time is
Carmen Sandiego, one of the point-and-click adventure games using Mohawk
code. Some of the scene logic is hard-coded, and so far I've only added
the code necessary for the first 'case', so this is also unlikely to be
completable beyond that for a while, since I haven't had much time to
work on ScummVM beyond bugfixing recently.

As mentioned when discussing adding the Mohawk engine, it was used for
various other non-adventure-game children's educational software. While
a lot of the code (resources, sound/music, scene management) was shared
among the games, the game logic for these tended to be hardcoded..

I implemented a lot of the logic for one of those (The Logical Journey
of the Zoombinis) out-of-tree, and while it's nowhere near finished nor
likely to be any time soon, I'd like to hear opinions on eventually
merging this if ever complete -- it is not simply sharing resource code,
so it would be quite a pain to maintain as an external subengine (although
git makes it considerably easier), but it ends up being many thousands
of lines of hard-coded logic simply because that's how the game was
written.

- fuzzie