OLPC Flash discussion

Thu Jan 18 10:13:41 EST 2007

Hi SJ and list,

I've recently tried to introduce a suite of open technologies I'd like
to see at the forefront of e-learning, particularly in OLPC context.
Not only is the language (haXe) powerful and a pleasure to work with,
but I can attest of the good direction the owners of these
technologies provide. Here is my previous post again:

from		Zjnue Brzavi <zjnue.brzavi at googlemail.com>
date		Jan 13, 2007 6:27 PM	
subject		Re: Flash support at current OLPC browser?	

Dear all,

I've previously posted to introduce a new language called haXe
(haxe.org), that compiles to flash (6-9) bytecode, js and a server
language neko (nekovm.org). More recently I've been getting involved
with a project called xinf (xinf is not flash - xinf.org) that
promises exciting options in terms of exposing a consistent gui api to
haXe for targetting both js, flash 9 (potentially other versions also)
and a neko-based format that will run on a xinfinity player, currently
in development. This means your gui will be able to 'fall-back' onto
the flash player when xinfinity is not installed or while it still
lacks in features. I'm very interested to look at combining these
technologies for the purposes of e-leanring and because they all have
a very small footprint, I believe it to be very well suited to the
OLPC project. If anyone else share these interests please get in
touch. I do feel there is great potential.

Best,

Zjnue

On 1/18/07, SJ Klein <sj at laptop.org> wrote:
>
> Cc:ing the devel list.
>
> Yes, there is a lot of Python work in and around OLPC; some people are
> also trying to use pygame as a model/kernel for a general intro-to-Python
> tutorial...  Bridge builder is pretty fly; I don't see why something with
> a similar physics engine couldnt' be built without high-end requirements.
>
> SJ
>
>
> On Thu, 18 Jan 2007, Ben Sawyer wrote:
>
> > I don't know the author per-se but InXile Entertainment just licensed the
> > rights to make it into a game.  While I think the educational aspects of it
> > are not as explicit as other things people cite we shouldn't forget the sheer
> > creativity it represents and the ideas it offers in terms iterative
> > development and other interesting aspects.
> >
> > There are many etoys (bridge builder anyone?
> > http://downloads.theregister.co.uk/Windows/Games/Building/bridge-builder.html
> > yes high-end requirements but humor me) and games that could be identified
> > for the laptops and of course many are built in Flash but I agree in general
> > that part of the core goal really being discussed here is the idea that there
> > be some etoy/game like programming system that is ideally suited to the
> > laptops.
> >
> > My gut tells me the script flash within flash is probably a dead end for any
> > number of reasons but I think it's a good conversation to force on adobe
> > (even if it were a special version they did and then ported everywhere) on.
> > I don't know enough about Squeak to know if it really fits the vision of
> > things like Click-n-Play and other "gamemaker/etoy" maker products we've had
> > in gaming but my sense is that despite it being used to make games by kids
> > that it probably doesn't do it to the degree and focus those products have.
> > It's complimentary to some extent perhaps even a step above as it moves you
> > further up the programming scale then a gamemaker but again I don't see it
> > 1:1.
> >
> > The critical issue will not be the issues of scripting or programming but of
> > the entire IDE and graphics process which is where from an etoy/game making
> > perspective is where the hard work lies.    Perhaps something that is written
> > to combine SDL with PyGame? (www.pygame.org)  Again the problem here is that
> > there isn't much of an IDE and graphics environment here but we might be able
> > to construct something with this that then works easily as I believe Python
> > support is quite heavy with OLPC.  The students at CMU's Entertainment
> > Technology Center are working with PyGame and Jesse Schell from CMU
> > recommended it highly.
> >
> > Sorry for lurking but my .02
> >
> > - Ben
> >
> > On Jan 17, 2007, at 12:23 AM, Samuel Klein wrote:
> >
> >>
> >> I love this discussion.  I also REALLY want linerider on the laptops...
> >> Does anyone know *fsk , the author?
> >> http://lineridervideos.blogspot.com/
> >>
> >> We're reaching thread-length and cc: length for a public list.
> >> Can we take this discussion [and the notes below] to devel at laptop ?
> >>
> >> SJ
> >>
> >>
> >> On Tue, 16 Jan 2007, GameMogul wrote:
> >>
> >>> As far as networking would go, *that* is where we have our proprietary
> >>> coding, right?  Because there's no way to *just* use AS for multiplayer
> >>> flash games, you have to have server-side code (whatever language) to
> >>> control this, and interact with each client.  Multiplayer usage won't be
> >>> supported in non-OLPC machines, right?
> >>>
> >>> And to answer Joshua, loading local files is easy if the files are
> >>> XML-based (shortened with URL-shortener):
> >>> http://geobay.com/f12341
> >>>
> >>> You are right in that you can't save files... like you think.  Look at
> >>> what line-rider does.  How does it remember the 'levels' you make?
> >>> SharedObjects.
> >>>
> >>> Here's an example of LineRider's cache path for M$ (honestly, I don't know
> >>> Gnash's Flash cache for Linux... likely
> >>> /home/<user>/.gnash/etc./etc./etc.):
> >>>
> >>> For the exe:
> >>> C:\Documents and Settings\<user name>\Application Data\
> >>> Macromedia\Flash Player\#SharedObjects\
> >>> <some stupid cache name: LVSZR946>\localhost\
> >>> Documents and Settings\<user name>\Desktop\
> >>> LineRider_beta.exe\undefined.sol
> >>>
> >>> For the web player:
> >>> C:\Documents and Settings\<user name>\Application Data\
> >>> Macromedia\Flash Player\#SharedObjects\
> >>> <some stupid cache name: LVSZR946>\ic1.deviantart.com\
> >>> fs12\f\2006\266\6\f\engin2_16devart.swf\undefined.sol
> >>>
> >>> In this final folder is a Shared Object Library (SOL) file, where *all*
> >>> the level information is saved.  There is a (albeit... not nice way)
> >>> possible way to save user input, and just have it available next time they
> >>> launch the application.
> >>>
> >>> I know it's no the best solution, but I'm just putting it out there.
> >>> -Zak
> >>>
> >>> Alan Kay wrote:
> >>>> The plan is go further than this. The thinking has been that each OLPC
> >>>> machine will be a server, and can serve itself if there is no network
> >>>> connectivity.
> >>>> Again, there is no problem making any of this happen on the OLPC machine.
> >>>> The entire environment and permissions are controlled. It's helping
> >>>> children who don't have these machines in the rest of the world to be
> >>>> part of the same conversation and activities.
> >>>> Cheers,
> >>>> Alan
> >>>> At 03:25 PM 1/16/2007, Joshua Ellis wrote:
> >>>>> Okay, I understand now. The key is to have the simple IDE easily
> >>>>> deployable across platforms, not just the SWFs it creates. I'm sorry
> >>>>> -- I lost that somewhere.
> >>>>> One of the main problems is that Flash *cannot* interact with the host
> >>>>> file system at all. It can't write files. I don't even think it can
> >>>>> *load* local files. (I've never tried, but it's a pretty good
> >>>>> assumption.) And this is unlikely to change -- it's a security thing.
> >>>>> Adobe's Flex Server can compile apps over the network on the fly:
> >>>>> check out http://try.flex.org/index.cfm to see an example. And this
> >>>>> would be possible to replicate as an open-source tool.
> >>>>> SJ mentioned to me the idea of a server OLPC as well as a client one
> >>>>> -- from what he'd said, I'd envisioned a sort of LAN where all of the
> >>>>> laptop OLPCs could connect to one server OLPC, which could function as
> >>>>> a router, or a machine for a teacher. If this is the case, couldn't
> >>>>> one run a compiler on *that* machine?
> >>>>> jze
> >>>>> On 1/16/07, GameMogul <zak at arkitek.com> wrote:
> >>>>>> Well, if we had the ability to use outside applications, then, like
> >>>>>> Alan
> >>>>>> said, why are we stuck to Flash to begin with?  If we want this to play
> >>>>>> on ALL flash-capable computers... it has to built-in.  If there are
> >>>>>> external applications or external linkages to the Internet required
> >>>>>> (i.e. server-side scripts), then it is very limited; such as, as Alan
> >>>>>> said, school-districts not allowing the installation of executables.
> >>>>>> Which also brings up the point of Internet-dependency, *requiring*
> >>>>>> server scripts would make it so that people had to have a connection.
> >>>>>> What if a laptop is out of range of others?  Unfortunately, you
> >>>>>> couldn't
> >>>>>> say we'd just use a PHP engine on the laptop, because 1) large
> >>>>>> size-bloat there and 2) it wouldn't work for just your average Joe on
> >>>>>> an
> >>>>>> M$ box who won't have a PHP engine on his computer.
> >>>>>> -Zak
> >>>>>> Joshua Ellis wrote:
> >>>>>>> Also, has anybody looked at Apollo yet?
> >>>>>>>
> >>>>>>> http://labs.adobe.com/wiki/index.php/Apollo:developerfaq
> >>>>>>>
> >>>>>>> If you could get Adobe to release/write a version of the Apollo
> >>>>>>> runtime for OLPC, that would kick serious ass -- you could develop
> >>>>>>> apps for the OLPC in HTML/Flash/Flex/CSS/Javascript!
> >>>>>>>
> >>>>>>> jze
> >>>>>>>
> >>>>>>>
> >>>>>>>
> >>>>>>> On 1/16/07, Joshua Ellis <jzellis at gmail.com> wrote:
> >>>>>>>> Also, here's a consideration: Flex generates Flash apps based on the
> >>>>>>>> MXML language. So wouldn't it be possible to create a simple Flash
> >>>>>>>> IDE
> >>>>>>>> that generates MXML files, which could then be passed to a compiler,
> >>>>>>>> rather than trying to make Flash evaluate user-written code from
> >>>>>>>> within the player itself?
> >>>>>>>>
> >>>>>>>> Oh, my poor head is aching. :-)
> >>>>>>>>
> >>>>>>>> jze
> _______________________________________________
> Devel mailing list
> Devel at laptop.org
> http://mailman.laptop.org/mailman/listinfo/devel
>