Ok fine, I had the title coined this way to bring in more visitors 
.
but seriously, I think the blog thingy needs a lot of ramping. I am new to this – so point me to any tools outside if I am missing anything. let me actuall step back, even before the voice feature. whats up with the comments stuff? I write a 200 line blog sectioned into 5 paras, 4 folks read it and write up comments on para3, para4 and para5.
why does the consolidated comments appears serially? Why cant blog sites give us tools that can inject the comments at the specific sections – like the comments section in Microsoft Office. Seriously, this makes sense. Now if someone gets this consolidated view, he can read all the comments in the correct context. I think this is super user experience. This creates an illusion for the end user to have been in a live conversation session and listening to various aspects at one shot.
Now let us get to voice – man, I need to use smiliies, “uh oh”, “duh”, “hmm”, etc to enable expressions. Text has no expressions, no tone , no emotions and thats why we need voice people. Take the traditional world for example, we see Jay Leno’s show not only for content but for his style of presentation. I think blogging has the potential to take it there but lacks it today. Now someone is going to say ‘text editing is easy but what if I want to make changes to my voice comment’. well I have 2 things – one, you cannot make changes to what you say in a live conversation and a voice blog feature is enabling it. two, fine agreed, can we allow folks to inject smaller audio expressions within a blog text. Now the browser can play it for you it starts displaying the text in a order and then you hear the audio expression wherever it is apt.
I agree folks have gotten used to adapting to the ‘only text’ aspects of the web – search, email, blog. but the future is voice and video. Things are changing rapidly. With TellMe integration, I am hoping that search will become audio search. the implementation does have interesting aspects – will the backend be simply text and the search query is translated to text or the backend text is enriched and augmeneted with voice tags and then searched or the backend is a audio database of wave files. The parser will simply strip open the backend wave files and search within them. Well ya the last option seems a stretch. but come one, who would have thought that you can have a page format, write it to disk, implement a buffer manager in memory, search for things in memory using bufsearch() (i see my SY buddies smiling) and searching through the page for a audio pattern data match. well if this was 2000, I would ahve some funding – well this is as good as WebVan for sure.
but again, if you are a media systems person, I will intereted in hearing from you. your thoughts on why this is fundamentally not possible. Can we invent a new audio format that allows this? metadata for patterns. then building a database engine with a buffer manager and access manager and a query engine on top – voila, hear comes the audio search engine.
Ok , let me stop now and give you folks to “text blog” your comments. as for me, i will enjoy my idli, sambhar and chutney.
