W3C home > Mailing lists > Public > public-webapi@w3.org > May 2008

Re: File IO...

From: Timur Mehrvarz <timur.mehrvarz@web.de>
Date: Wed, 7 May 2008 22:47:21 +0200
Message-Id: <04A4457E-6A52-45EB-BA81-793E24317762@web.de>
Cc: Maciej Stachowiak <mjs@apple.com>, Charles McCathieNevile <chaals@opera.com>, "Web API WG (public)" <public-webapi@w3.org>
To: Scott Shattuck <idearat@mindspring.com>

On 7. May 2008, at 20:23, Scott Shattuck wrote:
> Clearly if all i want is scratch area I don't need file system  
> access. But there are use cases for integrating a browser-based  
> application with the host operating system that can't be addressed  
> in that fashion. It's naive to imagine that these use cases won't  
> continue to create pressure to expand the capability of the browser  
> as a "VM" until it is on a par with, say, the JVM, in terms of what  
> it can do when properly authenticated.

Exactly. But how long will it take to get there? You know, there is  
this really cool new Java bridge (6u10/plugin2), that allows seamless  
integration of scriptable Java code in any Web application. And there  
is this almost endless list of available extensions.. File IO, Socket  
API, XHR with event-based parsing and no-same-host, mpeg, MIDI,  
Address-book, PIM, SMTP/POP3/IMAP, Bluetooth, Zeroconf, Sensor API's  
and what-have-you other JSR's, all of it available today. Does it  
really make sense to basically replicate this into browsers? Why not  
use the browser for what it is good at (portable presentation, UI)  
and embrace that GPLed JVM for all device and security related needs?


Received on Wednesday, 7 May 2008 20:47:10 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:16:26 UTC