Hi, I'm a bit confused by your idea. Do you want a standard way to deal with formatted data being displayed in an HTML page from Gears? I'm not quite sure what facilities would do this for you; JavaScript has first-class regular expression support which can already help. Firefox and Safari can also apply an XPath expression to a web page that is being displayed, while Internet Explorer can get the same functionality when using the Sarissa open source library; this can help you isolate particular portions of a page.
On Sun, Sep 7, 2008 at 8:11 AM, stylized.fact <[EMAIL PROTECTED]>wrote: > > One of the constant nuisances associated with mashup/extension/ > greasemonkey development is having to develop a regular expression to > retrieve data from some DOM node of an HTTP response text string. > Google Gears/Chrome seems to offer a good juncture to try to develop a > simpler way to gather useful data on the web. This could come in the > form of an object model, or a query function (some kind of query > generator linked to a GUI tool like Firebug would be nice). > > Remember that most of the important things that people check on the > web have strict defenses around them to prevent the information from > being commoditized. No operator of a travel web site wants their data > to be easily readable by anyone other than a person browsing their web > site. No retail merchandiser wants to provide people with a > straightforward way to do price comparison across multiple vendors. > They will resist any effort to develop an open standard for data > interchange. This is something that was overlooked when momentum > gathered behind the XML web services idea. > > There is long standing value in the ability to reverse engineer > formatted data presented in HTML. Adding generic functionality that > addresses this in Gears would be a major step forward. > -- Best, Brad [EMAIL PROTECTED]
