As you may or may not know, I've been a ML user since the days before MLServer was even a glimmer in Mario's eye (that is to say, a long time), when ML was only a client front-end. So I have been part of the (r)evolution since the beginning. That said, I think it is about time to ditch the Flash client (which is slow, not universally compatible with all OSes, etc) and migrate to a 100% browser based client. The real power of ML no longer lies in a "slick looking client", but in the backend power of MLServer and various modules. I know there have been some great strides with iPhone/iPad Safari compatible clients, so why not go the extra mile and convert ML to an all browser compatible client?
The advantages are endless, like:
- * the ability to use ANY device with a browser, regardless of OS. This would be huge! Imagine ML on your Android based touchpad, linux pico pc, iPad, Motorolla web device, iMac, Blackberry, etc, etc - this would catapult ML to the next level without question.
* the ability to integrate any other ip based component, and/or use browser add-ons and scripts (Greasemonkey, etc).
* the ability to remotely (off-site) mange and control ML from anywhere in the world...with ease.
* the ability to integrate flash, ajax, dhtml, html5, java scripting, etc all together.
* the ability to seamlessly integrate "internal" ML data with external data...without pain.
* much lower resource overhead required. As you know, Flash is a giant resource hog.
* develop scenes only ONCE, and control formatting, scale, aspect, etc with stylesheets.
You can probably think of a few more that I'm missing, but I think these are the "biggies".
So what are the disadvantages to a 100% browser based client? That's easy...none, nothing, nada, zip, ziltch...outside of course of the initial coding/development on Cinemar's side.
As a quick note- I do realize that the built-in webserver in MLServer allows from my understanding a fair amount of interaction already, but I'm speaking more towards the development of a browser client "app" - border-less, optimized, etc. Perhaps even a somewhat WYSIWYG type "editor/creator", or MLS converter?
I would settle for server side scripting ability (which might perhaps be a even better solution) and then "roll my own" version of Firefox.