[Kde-games-devel] GSoC Proposal: mancala games

[Josef Spillner]

Hello,

I'm going to comment on your proposal by mail. Don't take my notes as personal 
criticism, in fact a lot of it applies to other proposals I've seen as well so 
everybody should invest some time into improving the proposals and submitting 
them.

Am Dienstag, 31. März 2009 10:15:29 schrieb mudit malpani:
> i am submitting my proposal for mancala games. please give me some
> suggestions for its improvement.

First hint: Submit it to the GSoC web app already. The earlier the proposals 
appear there, the better.

> I am very much interested in making programmes in C++. I have been
> programming in C++ since 5 years and have made many interesting games like
> snake (human vs. comp), Othello, gobblet, Chess and many other projects.
> This experience of mine in making AI of different games will help me to
> easily complete the project.

Second hint: When you've done something before, point the reviewers to the 
websites of the projects so they can verify the quality of earlier work.

> Also I have a significant knowledge of functional programming (have been
> using scheme for about a year) due to which I can make fast and efficient
> programmes. This provides me an edge over other programmers.

Again it would be good to point to an example, especially when you can show 
that a certain task was accomplished in shorter time than with imperative 
programming languages (e.g. a refactoring for direct comparison). It looks 
like AI development in kdegames is something where functional programming and 
higher-order logics often appears to be a suitable choice, but then again the 
majority of developers don't use such languages and can hardly judge the 
merits based on such phrases alone.

> *I will start my project by first analysing the codes of existing strategy
> kde games so that I can get the style in which codes of kde are written.*

Yet again, being more specific and already more into the project helps a lot 
conveying a certain feasibility of the proposal. For example, it would be wise 
to already invest some time to dig through current games and come up with a 
summary of how strategy is implemented (often not by specific rules I'd think) 
and why and where to improve this.

> *Then I will start building proper algorithm for each game and making its
> code which will include:*

Hm, I guess you don't mean each existing KDE game but rather each Mancala 
game?

> 1.      Making a 2-player game

How in detail? (I guess you mean two players on one PC or "hot-seat" here).

> 2.      Making 1-player game (i.e. incorporating AI in 2-p game)
>
> 3.      Adding special effects to the game like adding additional graphics
> and sounds.

More details please: Are you aware of free arts already for Mancala or do you 
think that artist skills need to be found for that task? Starting a game 
project and completing all of its aspects like art, documentation, translation 
etc. is a huge task.

> 4.      Providing extra features which will make the game more attractive
> like changeable interface, allowing user to set time for each move etc.

More details please. This sounds like an innovative part which really improves 
gameplay, so it should be explained better, maybe with mockups or comparison 
to existing UI and gameplay concepts.

>  Till 15 may: getting familiar with kde coding environment and learning the
> basic rules and strategies of different mancala games (most of which I
> know).
>
> 15-30 May: Completing Kalah and Owari games (which I know very well)
>
> 1-30 June: Completing all other games (5 days each). All the games have
> many features in common so it will not take much time in making the codes
> for these games after completing Kalah and Owari.

A link to a website with game rules would be nice. Nowadays a lot of rules can 
be found on Wikipedia, but then again there are often enthusiast websites with 
better explanations.

> 1-15 July: adding additional features to the games.
>
> 15-30 July: Reviewing the codes for any bugs and consulting my mentor for
> further improvements.

Mentor consultation should be a regular activity.

> If I could complete my project early I would like to make other nice
> strategy games after taking advice from my mentor.

This particular sentence is not very convincing to me. I'd rather focus on one 
thing and leave additional time for quality and feature improvements.

Josef