With the help of PyGame and samples from PyAno, I've just made a mock sampler for Maemo. download from http://www.dj-geist.com/musicpad.tar.gz

I addressed the delay time in sample-playback by initializing the waves first and used PyGame for both sound and to draw the keyboards (its seemingly impossible to use GTK's button/grid to make a smooth keyboard.

I'm planning on developing some live production software for my N800 as soon as I get the SDK installed or scratch Gentoo and go for Debian.

My personal belief is that; since we all have our own ideas for what programs we envision, we leave all the others out. In a Windows world -( Monolithic: one application with all your features )- we need to translate to the Open Source Software world -( Centralized-Contributions: developing features towards a common framework for pure portability )-

In the world of sound, the Linux/ALSA studio developers lean toward the low-latency JACK audio service, which allows you to connect jackapp-to-jackapp and applications can be a simple any-ui LADSPA plugin, a graphical synthesizer suite (imagine zynaddsubfx PORTED to armel), and ONE SEQUENCER. As always, we are looking for a user friendly user interface.

What I everyone wants out of my N800 is to be able to create on the fly. This includes sequencing events with save/open functionality, even PD (Pure Data: A flexible and real-time GUI programming interface for audio & data-structures : doesn't even save its realtime songdata ). I have been searching for a programming framework similar to JACK and DSSI, but all I found in a language I understood was PyGame's Mixer.

But this raises the question;
Do we?

1: Start with an application framework that existing software can be ported. (requires SDK and cross-compilation)
[User Friendly UI: requires window-switching]

2: Create one strong native-armel studio application to accomplish all of these tasks (Monolithic structure)
[User Friendly UI: entirely controllable but requires lots of developers]

Or?

3: Create a plugin & transport framework(or sequencer since the less = the easier synchronization gets) in which each plugin may use the same output mixer and user-interface.

We know this is possible on the Nokia productline as the ACID.pd -sequencer-sampler-synth- example for PDa ( Pure Data Anywhere ) demonstrates. Yet PDa on Maemo lacks because it doesn't have the calls to hildon's textinput api, as well as its inability to save realtime song data (unless one precodes it in.)

This goes back to why I either need to Setup SDK or Backup & Scratch[box]
*End Transmission*