I have been thinking more, lately, about the idea that DOM trees can just pop right out of local store in nested order. Load up your WebSql/indexedDB with the CSS and XSL files, then your reader is initialized. Keep the version count low.
Next, your reader can download the latests data., but the CSS files are already in DOM order, likely in local ram. They are all local, ready to go,. Namespace is simplified for the blogger.
Example: The boiler plate of your stock charts, for example, they pop right up, then there is a slight moment when the latest data is grabbed. Better than simple cache.
How does the web author get set up? Here is simle syntax:
Get 'http://some resource' as MyTable;
Don't mind the surrounding script tags, not a problem. That is the script, covered by script tags. Inn my document intro, make sure the local DOM stores are primed. I like as the alias predicate. I like:
set MyTable.sales as X, MyTable.volume as Y;
See, then I get a simple:
standard xy plot x,y;
The thing is, if it isn't a key word, like the word standard above, then it considered a variable, ready in local DOM store. You get a simple resource layout, based on local store, easy for the blogger. Local DOM store processed through stages to immediate display. Offer up a limited set of key words, and any other specialization should come from Dom local store, via variable local sore names.
It's all in my patent.
No comments:
Post a Comment