The Disappearing Desktop - It's Real This Time

David Bovill david at openpartnership.net
Fri Nov 18 07:01:26 EST 2005


On 17 Nov 2005, at 04:46, Dan Shafer wrote:

> Pricing models will change. We'll see pay-per-use, pay-per-month,  
> pay-per-file, pay-per-K and other similar models. When it's not  
> necessary for the manufacturer to package, distribute, sell, track,  
> upgrade and otherwise deal with thousands and thousands of copies  
> of the software out there -- and when piracy becomes all but  
> extinct thanks to the new models -- software prices will dive while  
> profit margins remain high.

Can't help myself - irresistibly prodded.

This may get a little philosophical - so I keep it short.  
Centralisation of data is not only dangerous socially, but  
economically. We have game companies to thank for ubiquitous,  
powerful desktop machines. Natural business means diversity of  
processing and always processing data at the lowest possible level -  
higher levels of abstraction are necessary, but always slower and  
more expensive. Diversity of processing is not only more efficient,  
it drives down prices and provides a firm base for evolutionary  
growth of new technologies.

We now have dirt cheap ubiquitous processing and high capacity local  
storage - these things will keep being used, experimented with and  
many new application developed to make use of their capacity. The  
question is only how do the data and GUI processes that run on this  
cheap end user hardware fit more productively into a connected world.

Dan - your focus on thin clients over the other much more important  
issues regarding AJAX technology and web services is in my opinion a  
distraction. Make use of the local processor. Make use of the hard  
disk. Figure out how to author and make use of web services. Sort  
security. Sort data synchronisation to online storage.

> On Nov 16, 2005, at 9:43 PM, Alex Tweedly wrote:
>
>> I don't believe that a "Web-Photoshop" would need to satisfy the  
>> digital photography professional (mapping professionals aren't  
>> using Google Maps !). I think to get a commercially successful web- 
>> based photography editing app you need to satisfy 75% of the  
>> population - who start out with 3-6M-pixel photos compressed down  
>> to 1/2Mb JPEGs, not the pros using 32Mb RAW images,.
>
>
> Is one implication that, in the brave new web-app world,  
> professional-grade applications -- because nobody but professionals  
> will be using them -- will get really, really expensive? Yes, many  
> are now; but many aren't.

There was a great project - French company that used server side  
image files which you could edit client side only sending the  
compressed changes back and forth - effectively allowing the  
manipulation of screen resolution images on your client machine to  
apply the right changes to the larger online image - swallowed by the  
same sort of merger process that killed mTroplois I believe.

A number of people are interested in the same for Video - edit your  
local low resolution copy on your laptop, have these changes render  
your videos on the server without taking up your own CPU cycles. Also  
some very mature open source projects now in the area of network  
processing - particularly 3D rendering. Summary - expect to slowly  
see a number of quality media related web services allowing you to  
process your heavy media files without requiring a G5.

But do not expect any of these interfaces to be in Internet Explorer  
- or anything resembling it. They will be made in something  
resembling Revolution. They almost certainly would be if Rev was open  
source - they will probably be made in something else if we and  
RunRev don't get where this is going.

Question: if you wanted to knock together a quick cross platform  
photo-editing application which used web services to upload the  
images and metadata to Flicr for instance - what language would you use?

Question: if you wanted to knock together a quick cross platform  
video-editting application for laptop journalists which used web  
services to upload the video clips and metadata to a mobile phone  
service for instance - what language would you use?

I for one would not use a browser - Java or no Java. If this changes  
- then it will be because the browser has become like Revolution and  
we will all be switching platform.



More information about the use-livecode mailing list