> please consider to switch to a pure HTML5 applicattion

I agree it is necessary, and it is on my todo list...

btw Saturate has already started development of an app in Javascript/HTML5, it has the basic functionality, but is missing a few things that the current Java app already does:

• User is asked to enter a username (optional), age range, pronunciation dialect and microphone type. (if blank, user is name is “anonymous”)
• Users are given 10 random prompts from a selection of 200 or more prompts
• User sees a visual graph of the audio volume so they can confirm they are not speaking too loud, too soft, and have not started speaking too early or stopped recording too soon.
• If the audio is too loud or too low, the user is notified.
• On the 10th prompt, the upload button becomes usable
• user must be notified that upon upload that they are assigning their copyright to the FSF and licensing it under the GPLv3
• once the upload button is clicked, a progress bar is displayed (so user will not exit to another page before the upload is completed)
• clicking upload causes a zip file (or gzipped tar file) to be created containing the following:
• the audio files (in wav format 48kHz/16Bit - no compression),
• a file containing the prompts (called prompts.txt), with prompt ID as first 'word' followed by the prompts “sentence” (see attachments)
• a file containing the summary license (called license.txt), with a copyright for the Free Software Foundation and the current year (see attachments)
• a file containing the user's particulars (called readme.txt) (see attachments)
• name of the zip file is the username, or anonymous if none was entered.
• submission package is uploaded to VoxForge website using php server code from the postlet
applet site: http://postlet.com/scripts/
• app needs to be localizable in many different languages (see attachments for list of fields that are language specific)

You are more than welcome to take a crack at doing this!

thanks,

ken