Hacker News Reader

Comment by melling | Hacker News Reader

This is great. Now maybe if people can get as upset about Microsoft not better supporting HTML5, in particular the canvas tag.

http://www.webmonkey.com/2010/03/internet-explorer-9-shows-u...

IE9 is a big improvement for Microsoft. However, if they included more support for HTML5, developers could get "all their wood behind one arrow."

View on HN

Why do you need <canvas>, if you have a high-performance SVG implementation?

Different use cases. It's like why there is a need for both Illustrator and Photoshop.

SVG is great for vector graphics but for fast bitmap-based graphics you need <canvas>.

You do not want to handle bitmaps where each pixel is an object.

Besides some kind of interactive bitmap editor, where would someone need dynamic bitmap graphics that would be too much for a very fast SVG implementation but not too much for <canvas>?

Whenever "vector look" is not what you aim for.

One huge example is games - vast majority of games use bitmap graphics.

Or if you want to do some pretty data visualizations like heatmaps.

----

Edit: Here are some examples showing what fast canvas bitmap graphics can do (and what would be very very hard to do with SVG):

http://29a.ch/2010/3/24/normal-mapping-with-javascript-and-c...

http://mrdoob.com/lab/javascript/effects/fire/01/

http://mrdoob.com/lab/javascript/effects/water/00/

http://www.nihilogic.dk/labs/canvas_music_visualization/

http://www.nihilogic.dk/labs/strange_attractors/

http://www.p01.org/releases/512b_jspongy/jspongy.htm

http://www.p01.org/releases/20_lines_hypno_trip_down_the_fra...

http://www.p01.org/releases/20_lines_dynamic_hypnoglow/

http://www.ehkä.fi/

Flash isn't bitmap-based though. SVG is more like Flash in so many more ways that would make the implementation easier and better (scalability, for example).

Flash can do both vector and bitmap graphics (which is BTW much faster than current canvas implementations, even in fast browsers like Chrome [1]).

Adobe used canvas probably because you can redo vector graphics with a decent performance in canvas, while it doesn't work the other way around.

Flash exporter to SVG could be marginally better for vector parts of what Flash can do, but it would lose all bitmap capabilities, while Flash exporter to canvas can potentially replicate everything what Flash can do.

----

[1] Flash bitmap graphics is fast enough to do pretty decent full-screen 3d graphics with textures and shading:

http://away3d.com/

http://dailypv3d.wordpress.com/

Flash is vector based for the most part, but a few versions back they added objects that deal with bitmaps the traditional way, it works pretty well. But on the other hand, why bother with the vector stage at all? The image will have to be rasterised before being displayed anyway, so what difference does it make whether you have a bitmap surface you can draw bezier curves on, or a vector environment that ends up as a bitmap?

An aside, bizzarely I find it's often faster to redraw some graphics "vector style" (calling the lineTo/bezierCurveTo etc methods again) on canvas than it is to use drawImage and use a cached version of the same graphics.

Reply on news.ycombinator.com