Re: [glob2-devel] Revive Glob2 with a kickstarter

[Stéphane Magnenat]

Hi Zenfur and all,

I am very late with this answer, but the end of year 2014 was super busy 
for me, now it is holiday and I finally have the time to read all your 
nice emails, think about them and give a feedback.

First, as Leo I feel very heart-warming that you are interested in 
reviving Glob! I will help the best I can, considering that I have 
limited time. Maybe at some point, I might have some student work force 
to contribute, but it is uncertain so for now let's not count on it.

Reviving Globulation through a Glob2+ or a Glob3 is also a dream of 
mine. With Emmanuel Eckard, the person who drew the buildings in Glob2, 
we started to do some concept drawing for a Glob3 game [1]. The idea 
illustrated there is to have globules living in the clouds of a dense 
atmosphere planet (for instance Venus), allowing interesting 
construction activities (platforms and bridges) and interactions with 
the environment (wind, etc.). Think about "dwarf fortress in the 
clouds". This is a long term dream, but maybe these drawings give "food 
for thought" ;-)

At a shorter term, Emmanuel and I thought about improving the economy of 
Glob2 to make it less linear. I have put the preliminary result of that 
brainstorming in attachment. The idea is to improve the dynamics of the 
game with minimal changes in the engine. One thing I would change is to 
script the unit behaviour, as the current system based on C++-coded 
hierarchical state machines is unmaintainable. As discussed by others, 
Python would be a language of choice (but Lua and Javascript should be 
considered too). The idea would be to use continuations (i.e. generators 
in Python, or a hack in the VM embedding code) to allow to code the 
behaviour of units very easily and implement that behind the hood using 
co-routines. In the context of robotics, I made a prototype for C Python 
demonstrating the concept [2]. A challenge with Python is to ensure 
proper sand-boxing and deterministic execution. There are tiny Python 
implementations [3,4] which might help. TinyPY seems the most promising 
approach after one hour investment in looking around. More thoughtful 
search might change that.

At an even shorter term, I agree with Leo that as a community we could 
"prove" our interest by fixing the bugs in the existing game. By the 
way, what is the main repository, is it [5]? What is the latest state, 
is the tip of the master branch in hg 0.9.4.4? People also started to 
post issues both there and on savannah, savannah is the official 
tracker, isn't it? I recently re-read parts of the code base, and I 
think it would be worth trying to port it to mobile devices. That would 
mostly imply UI changes and maybe using hi-res graphics (twice the size 
of the current ones).

At the level of monetization, why not having a fully free version for 
single and local game, and a paying version for YOG games on mobile? As 
a first step, it will help to pay for the server.

So, maybe the following first tasks are reasonable for a revive of the 
existing version:

1. fix bugs
2. compile current version for android
3. try to embed a proper scripting language for the behaviour of units 
and maps
4. explore improved economy system, thanks to scripting

What do you think?

Have great holidays!

Stéph

[1] http://h.magnenat.net/~steph/glob3-concept/
[2] http://wiki.ros.org/executive_teer
[3] http://www.tinypy.org/downloads.html
[4] https://code.google.com/p/python-on-a-chip/
[5] https://bitbucket.org/giszmo/glob2

--
http://stephane.magnenat.net

new-economy.txt
Description: Text document