gasp

0 views
Skip to first unread message

Jeffrey Elkner

unread,
Feb 27, 2007, 1:03:49 PM2/27/07
to Vern Ceder, tom.h...@gmail.com, kjc...@ubuntu.com, gasp...@googlegroups.com, edup...@googlegroups.com
OK, this looks like the beginning of an exciting collaboration opportunity / free software project management school!

I've created a google group:

gasp...@googlegroups.com

and made you an owner.

We already have a project on launchpad:

https://launchpad.net/gasp

We can use launchpad as the software repository. Steve Alexander just told me that he and Barry Warsaw are here at pycon sprinting on a feature that will automatically create a mailing list for each team inside launchpad and add team members to that list. As soon as this is ready we can move the mailing list to launchpad.

In the mean time, the next task seems to me to get the software repository and revision control (bzr) setup properly. Kevin Cole offered to help us with this.

Does this make sense as a directory structure:

gasp (the common api stuff lives here)
tk (the livewires fork)
gtk (the pygtk/cario backend)
sdl (the pygame backend)

I'll wait 2 days to give folks to approve/object to this arrangement, and then work with Kevin to set it (or whatever else we decide) up.

Thanks!

jeff

On Tue, 27 Feb 2007 08:16:21 -0500, Vern Ceder <vce...@canterburyschool.org> wrote:
>Hey guys,
>
>I sent a reply to Jeff last night, but let me re-send and amplify to all of
>you.
>
>My idea, which I sketched out on my plane ride home is that we should
>encapsulate the graphics functions in a base class and then subclass that
>for each display system we want to support. I'm really uncomfortable with
>the idea of having different versions in the scheme you outline.
>
>So in my thinking the main library's call to, say, box() would be actually
>just wrap a call to g.box(), where g is an instance of the appropriate
>graphics class, created by begin_graphics(), if that makes sense. The
>individual graphics subclasses would be responsible for things like mapping
>to the appropriate coordinate system, etc. and the main library could be
>responsible for some of the normalizing of various parameter options that
>the library offers (you know 4 separate points vs 2 tuples of points, things
>like that).
>
>That way, the library works as soon as we get one version working, and we
>just add graphics implementations.
>
>Finally, I took Tom's idea about porting to the XO a different direction. I
>have a senior girl (who has a good chance of being accepted to MIT) who will
>need a 40 hour internship this spring. I would love for her to have the
>opportunity to implement the cairo port of this, letting her work with
>various aspects of the open source development process. If Tom were willing
>she could even tap into some of his experience, etc.
>
>Let me know what you think,
>
>Vern
>
>Jeffrey Elkner wrote:
>>Hi Vern,
>>
>>Tom Hoffman had an idea this morning: Let's get
>>livewires/scale/gilliam/gasp (we need to get this name business resolved)
>>on the OLPC. So he started a version of the API written in pygtk and cario
>>(which is what is available on the OLPC). So we will soon have three
>>versions of this wonderful little API, and unless we nip this in the bud
>>immediately, a naming nightmare.
>>
>>I would like to propose that we call them all gasp. We can differentiate
>>the packages with gasp_tk, gasp_pygame, and gasp_gtk. The last two will be
>>GPL, while the first one will have the original livewires license.
>>
>>What do you think?
>>
>>jeff
>>
>>On Mon, 26 Feb 2007 00:09:04 -0500, Vern Ceder
>><vce...@canterburyschool.org> wrote:
>>>I agree on pretty much every count. Grail the browser may be gone, but
>>>it's not forgotten....
>>>
>>>So yes, I'm happy with the current version of SCALE. I spent my time in
>>>airports today working on some implementation details... so lets just fork
>>>this turkey... or something like that... ;-)
>>>
>>>Cheers,
>>>Vern
>
>--
>This time for sure!
> -Bullwinkle J. Moose
>-----------------------------
>Vern Ceder, Director of Technology
>Canterbury School, 3210 Smith Road, Ft Wayne, IN 46804
>vce...@canterburyschool.org; 260-436-0746; FAX: 260-436-5137
>

Douglas S. Blank

unread,
Feb 27, 2007, 2:44:20 PM2/27/07
to edup...@googlegroups.com, Vern Ceder, tom.h...@gmail.com, kjc...@ubuntu.com, gasp...@googlegroups.com
Hi all,

I must have missed a post or thread; can someone explain what is (or
point me to) GASP and how it is related to tk, gtk, and sdl?

I looked at the website, but it doesn't seem to jibe with having
"different backends".

Thanks for any pointers!

-Doug

Jeffrey Elkner

unread,
Feb 27, 2007, 3:18:30 PM2/27/07
to edup...@googlegroups.com, matt...@gmail.com, jlha...@gmail.com, sunam...@gmail.com, gasp...@googlegroups.com, geo...@rightinternet.com
Hi Doug,

You didn't miss anything from the mailing list. I was responding to discussions several of us have been having at Pycon, so the confusion is completely my fault.

A little history... a module named livewires implements a simple graphics API that Vern and I have been using with our students and really like a lot. Two years ago I asked a student, David Muffley, to re implement the livewires API using pygame instead of tkinter (which livewires uses). "gasp" was born. Yesterday Tom Hoffman and Vern Ceder independently began thinking about getting the API on the OLPC, which means wrapping pygtk + cairo. A bunch of discussion on all this yesterday led us to where we now are, and motivated my posting.

So, if all this interests you please join the mailing list at:

gasp...@googlegroups.com

Thanks!

jeff elkner

Vern Ceder

unread,
Feb 27, 2007, 4:03:35 PM2/27/07
to Dethe Elza, edup...@googlegroups.com, tom.h...@gmail.com, kjc...@ubuntu.com, gasp...@googlegroups.com
Dethe,

That sounds great to me.

I sketched out something that I will try to get out there in the next
day or two.

While the GASP launchpad already exists it still needs to be remodeled
to reflect the new direction we want to take it. We'll try to get things
worked out this week.

As Jeff says, join gasp...@googlegroups.com for more discussion.

Cheers,
Vern

Dethe Elza wrote:


> On 27-Feb-07, at 10:03 AM, Jeffrey Elkner wrote:
>
>> Does this make sense as a directory structure:
>>
>> gasp (the common api stuff lives here)
>> tk (the livewires fork)
>> gtk (the pygtk/cario backend)
>> sdl (the pygame backend)
>

> This looks cool. Where is the API defined? I might be able to find
> time to work on a Cocoa backend.
>
> --Dethe
>
> Simple things should be declarative. Complex things should be
> procedural. --Adam Bosworth

Andrew Harrington

unread,
Feb 27, 2007, 8:05:02 PM2/27/07
to edup...@googlegroups.com, Vern Ceder, tom.h...@gmail.com, kjc...@ubuntu.com, gasp...@googlegroups.com
Dear Jeff,
Having the graphics engine be dynamically set at runtime is certainly
the neatest thing, and not very complicated. I am excited about the
idea of the OLPC version!

I am not sure what mailing lists I am on. I would like to be included
in the gasp list.

What does Gasp stand for?

Andy


--
Andrew N. Harrington
Computer Science Department Director of Academic Programs
Loyola University Chicago http://www.cs.luc.edu/~anh
512B Lewis Towers (office) Office Phone: 312-915-7982
Snail mail to Lewis Towers 416 Dept. Fax: 312-915-7998
820 North Michigan Avenue aha...@luc.edu
Chicago, Illinois 60611

Reply all
Reply to author
Forward
0 new messages