Re: [pygame] Google Summer of Code participation

[Iuri <iurisilvio@xxxxxxxxx>]

Can anyone explain or comment this topic of the gsoc2009ideas?

Math for pygame (vectors, matrix, etc.)

Using numpy is always an option, but there's a few reasons to include some types useful in graphics and games:

• there's always the problem of people having to have numpy installed.
• game specific optimized routines. Some stuff is quite a bit faster than what numpy can do.
• Some people find numpy a bit difficult to get theirs heads around, and find stuff like http://pygame.org/wiki/2DVectorClass simpler.
• lots of algorithms for games are written in a way that is hard to do with numpy.
• exposing the routines from something like tinyGL would be pretty easy to do.
• There's a few existing libraries too: http://cgkit.sourceforge.net/doc2/module-cgkit.cgtypes.html
• GLSL vector types would be good to look at. To support things like swizzles, v.xy += 1 where it only increments the x and y members.

Be aware that you should be familiar with the basics of vector and linear algebra already. If you just started with the OO parts of Python, but have some good knowledge about what is written above, this might be the project for you.

In the wiki we have 2DVectorClass and 3DVectorClass classes. Why these classes are not so good? Or what I can improve on that?

What other classes are good to add to pygame? Matrix, quaternion, what else? Probably I'll apply on that. I'm thinking in add useful algorithms to my application (like the AI module listed on the gsoc2009ideas).



I'm a computer engineer student, and I started with pygame last year, doing a project with wiimote and led tracking to create some interactive games/examples.

If someone have tips about this, it'll be great.

Thanks,

Iuri de Silvio

Computer engineer, Technological Institute of Aeronautics (ITA) - Brazil

On Wed, Mar 11, 2009 at 6:49 PM, Devon Scott-Tunkin <djvonfunkin@xxxxxxxxx> wrote:

>Please do not mix font and image formats up. Focus on one of them, as
>implementing a bit of this and a bit of that is unlikely to be
>accepted. The main reason for this is that both are completely different
>areas and as such need several special considerations and probably API
>design changes to get all - to be planned - features into them.

I'll focus on one before I apply then, probably image formats, depending on which I understand best and find more interesting after some research.

>SVG would be some cool and huge task, I guess. As there is no
>cross-platform SVG library (to my knowledge and no, cairo-svg's not the
>best option here) to rely on, this would be a nice project and require
>you to show us some good work of you (because SVG is a complex topic).

ImageMagick does SVG, although incomplete. I don't know how useful that would be. Inkscape's libnr might be something to look at, although it might be saying something that they use cairo for faster rendering in outline mode.  If I had to roll my own from scratch, yeah that would very likely be beyond my current capabilities for the summer.

Devon

--- On Wed, 3/11/09, Marcus von Appen <mva@xxxxxxxxxxxx> wrote:

> From: Marcus von Appen <mva@xxxxxxxxxxxx>

> Subject: Re: [pygame] Google Summer of Code participation
> To: pygame-users@xxxxxxxx

> Date: Wednesday, March 11, 2009, 2:11 PM

> On, Wed Mar 11, 2009, Devon Scott-Tunkin wrote:
>
> >
> > I'd like to apply for gsoc 2009 in pygame when the
> time comes. I'm
>
> Great!
>
> > doing an MS in Game Development (programming), but
> have spent my first
> > year taking undergraduate prerequisites in computer
> science classes as
> > my background is in graphic design and not
> programming. That leaves me
> > probably at a beginning undergrad junior level in
> computer science.
>
> That's no problem at all.
>
> > Coming from graphic design I'm interested in
> adding more font and
> > graphic file supports to pygame. I know C reasonably
> well, but I
> > haven't really looked at the C parts of pygame or
> even fiddled with
> > the pygame python source, only used the api.
>
> Ideally those tasks could be done on a clean pgreloaded
> branch (yes
> ma'am, I'm strongly advocating here). It's API
> is a lot cleaner than
> pygame's at the moment, so a clean backport could be
> done once
> anything's settled.
>
> >
> > To get ready to apply, obviously it would be good for
> me to review
> > some of the pygame source, build pygame, and maybe
> even write a few
> > tests. As well as researching the technical side of
> implementing new
> > fonts and image format conversions as I am very
> familiar with using
> > many types of both, but conversion routines for images
> I only have a
> > general idea that different formats store the channel
> bytes and pixel
> > arrays in different ways.
> >
>
> Please do not mix font and image formats up. Focus on one
> of them, as
> implementing a bit of this and a bit of that is unlikely to
> be
> accepted. The main reason for this is that both are
> completely different
> areas and as such need several special considerations and
> probably API
> design changes to get all - to be planned - features into
> them.
>
> >
> > Implementing SVG or some sort of basic vector image
> support is a
> > related project I'd like to do, but may be out of
> scope if I end up
> > working on font and raster image format support.
> >
>
> SVG would be some cool and huge task, I guess. As there is
> no
> cross-platform SVG library (to my knowledge and no,
> cairo-svg's not the
> best option here) to rely on, this would be a nice project
> and require
> you to show us some good work of you (because SVG is a
> complex topic).
>
> >
> > Do you have any other suggestions for me? Ways to
> improve my chances?
> > Code examples I should write or focus on?
>
> Depending on what you want to do in the end, show us what
> you did
> already (projects, etc.) when it comes to applying (and we
> got some
> slots from Google). The easier the project, the less you
> would have to
> show ;-).
>
> However, first we will have to be accepted as
> organisation...
>
> Regards
> Marcus