Hi,

> According to the FAQ, the editor of the HTML5 spec (of which canvas is
> a part) expects it to reach W3C candidate recommendation status around
> 2012, and will not be fully finalised until 2022 or later, so there's
> plenty of time to get your requests in :-)
> http://wiki.whatwg.org/wiki/FAQ#When_will_HTML_5_be_finished.3F
>
> (And anyhow, I'm sure there's going to be an HTML6 proposal by then...)

Thanks for the pointer.

> That seems like a rather surprising omission (if indeed it is an
> omission) - Canvas's drawing capabilities are usually in step with
> SVG, which does have that feature. From a quick scan through the spec,
> it looks like strokeStyle set to a CanvasPattern might possibly do the
> job, although it is a bit unclear.

Hmmm, I'll look at that - thanks.

--
Richard Heyes
HTML5 graphing: RGraph (www.rgraph.net - updated 23rd May)
PHP mail: RMail (www.phpguru.org/rmail)
PHP datagrid: RGrid (www.phpguru.org/rgrid)
PHP Template: RTemplate (www.phpguru.org/rtemplate)
PHP SMTP: http://www.phpguru.org/smtp

On 31 May 2009, at 20:30, Richard Heyes wrote:
> Hi,
>
> Is it too late to suggest additions to the canvas spec? And if not
> does any one know who the best person to suggest them to is?
>
That would be the WHATWG mailing list:
http://www.whatwg.org/mailing-list#specs

According to the FAQ, the editor of the HTML5 spec (of which canvas is
a part) expects it to reach W3C candidate recommendation status around
2012, and will not be fully finalised until 2022 or later, so there's
plenty of time to get your requests in :-)
http://wiki.whatwg.org/wiki/FAQ#When_will_HTML_5_be_finished.3F

(And anyhow, I'm sure there's going to be an HTML6 proposal by then...)

> Like
> others I think I would find being able to specify the line style
> rather useful. Eg:
>
> context.lineStyle = 'solid'; // Default
> context.lineStyle = 'dotted';
> context.lineStyle = 'dashed';
>

That seems like a rather surprising omission (if indeed it is an
omission) - Canvas's drawing capabilities are usually in step with
SVG, which does have that feature. From a quick scan through the spec,
it looks like strokeStyle set to a CanvasPattern might possibly do the
job, although it is a bit unclear.

- Matt

Hi,

Is it too late to suggest additions to the canvas spec? And if not
does any one know who the best person to suggest them to is? Like
others I think I would find being able to specify the line style
rather useful. Eg:

context.lineStyle = 'solid'; // Default
context.lineStyle = 'dotted';
context.lineStyle = 'dashed';

Cheers.

--
Richard Heyes

Hi,

http://www.phpguru.org/dev/CanvasWars/game.html

Dunno what the hell to do with it though, At the moment it's a "dodge
the clouds" type thing... :-)

--
Richard Heyes
HTML5 graphing: RGraph (www.rgraph.net)
PHP mail: RMail (www.phpguru.org/rmail)
PHP datagrid: RGrid (www.phpguru.org/rgrid)
PHP Template: RTemplate (www.phpguru.org/rtemplate)

Hi,

Haven't done any showing off recently, so here's a dollop:

http://www.phpguru.org/article/a-html5-canvas-clock

Very similar to the example I would imagine you've already seen,
albeit with numbers. Because it uses the text and shadow APIs, it
needs Firefox 3.1b2, Chrome 2b, or Safari 4b or higher.

--
Richard Heyes

HTML5 Canvas graphing for Firefox, Chrome, Opera and Safari:
http://www.rgraph.net (Updated April 11th)

Hi,

Does anyone know if there's an ftell()-like function in the canvas
API? Situation is that I want to draw an arc from the end point of a
previous arc, but can't because I don't know the end point coords of
the first arc to pass to the second.

Thanks.

--
Richard Heyes

HTML5 Canvas graphing for Firefox, Chrome, Opera and Safari:
http://www.rgraph.org (Updated January 31st)

it's a good idea, the task seems very simple now: get the ".jzip"(this ahven't anme yet, right?) file, upack it, read metaconfig and open the main document.

the rest seems natural: relative paths everywhere, define all possible config options, a sample runtime able to handle that zip file and - of course - gather more people arround the idea!

2009/1/6 Jerason Banes <jbanes@...>

I think we're on the same wavelength here. I was also thinking something along the lines of a JAR for Javascript. Here's the RFC I posted on the WiiCade forums on that idea:

http://www.wiinode.com/forums/index.php?showtopic=6194

However, it's probably best to ditch the extra MANIFEST.MF and/or WEB-INF data. Those are Java-specific forms of Meta Data. I was thinking of a similar idea, but in JSON format:

http://www.wiinode.com/forums/index.php?showtopic=6196

The nice thing about using JSON is that it leaves the option open for client-based integration. e.g. Rather than having a JSP or PHP page write out an IFRAME or include the script, I could have a bit of AJAX code read the extracted manifest from the server and generate the correct elements to display the application.

Thanks!

Jerason

(...)

2009/1/6 Jerason Banes <jbanes@...>

Hi all!

As many of you probably noticed from my recent message, I've managed to complete a very sophisticated Javascript game recently. In addition, I've been porting several older DHTML games to work on modern systems and game consoles. One of the issues I've noticed is that there is no standard method of deploying Javascript games. (Or applications for that matter.) 

(...)

> I just released a Canvas-based game through WiiCade.com. Thought everyone
> here might find it interesting:
> http://www.wiicade.com/gameDetail.aspx?gameID=1317

Impressive. Takes me back to my Speccy days... :-)

--
Richard Heyes

HTML5 Graphing for FF, Chrome, Opera and Safari:
http://www.rgraph.org (Updated December 20th)

> ...

For anyone that's interested I've now managed to get it working:
http://dev.rgraph.org I used [event].offsetX/Y for Safari, Opera, MSIE
and Chrome, and for Firefox it was indeed a case of walking up the DOM
and using the accumulated offsetLeft etc.

--
Richard Heyes

HTML5 Graphing for FF, Chrome, Opera and Safari:
http://www.rgraph.org (Updated December 20th)

> Happy to be of service. :-)

Hi,

Seems I spoke too soon. Sorry for the noise.

--
Richard Heyes

HTML5 Graphing for FF, Chrome, Opera and Safari:
http://www.rgraph.org (Updated December 5th)

> The Offset* APIs are based on the location relative to the parent element.
> You'll need to compute the offset by walking the DOM using "offsetParent".
> (Don't use the DOM parent or you'll get wacky results due to the way tables
> are constructed.) Add up the offsetLeft values for each offsetParent and
> you'll arrive at the true location of the Canvas.

Hi,

That's what I was afraid of. Still, I've manged to do it (!):

http://dev.rgraph.org/

Thank you!

--
Richard Heyes

HTML5 Graphing for FF, Chrome, Opera and Safari:
http://www.rgraph.org (Updated December 5th)

Hi,

I'm having severe trouble getting the mouse coords when a canvas is
colicked on AND it's nested within a table. I've setup an example to
illustrate the issue here:

http://dev.rgraph.org/

The top graph is within a table, and when you click on it, it shows
you the position of the mouse pointer (which seems fine), and the
offsetLeft of the canvas, which is coming up as 1. Clearly it
shouldn't be. I need the position of the canvas on the page, or the
position of the mouse relative to the canvas. The bottom graph is not
in a table, using a DIV for positioning, and this is fine.

Thanks for any help.

--
Richard Heyes

HTML5 Graphing for FF, Chrome, Opera and Safari:
http://www.rgraph.org (Updated December 5th)

> Seems to work just fine in Opera 10.

Hi,

Really? I only have Opera 9.6 installed. I think I'll wait for this
then instead of hacking (or fumbling...) around. Thanks Eric and
Philip.

--
Richard Heyes

HTML5 Graphing for FF, Chrome, Opera and Safari:
http://www.rgraph.org (Updated December 5th)

On Fri, 05 Dec 2008 12:45:34 +0100, Richard Heyes <richard.heyes@...>
wrote:

> Hi,
>
> Has anyone had any problem with using the rotate() method in
> conjunction with Opera?
>
> Eg: http://dev.rgraph.org/trash/tradar.html
>
> Works fine in Firefox, Chrome and Safari, but not Opera. Cheers.

Seems to work just fine in Opera 10.

Cheers
/Erik

--
Erik Dahlstrom, Core Technology Developer, Opera Software
Co-Chair, W3C SVG Working Group
Personal blog: http://my.opera.com/macdev_ed

On Fri, Dec 5, 2008 at 11:45 AM, Richard Heyes <richard.heyes@...> wrote:
> Hi,
>
> Has anyone had any problem with using the rotate() method in
> conjunction with Opera?
>
> Eg: http://dev.rgraph.org/trash/tradar.html
>
> Works fine in Firefox, Chrome and Safari, but not Opera. Cheers.

It looks like the problem you're experiencing is that Opera applies
transforms to paths when it's stroking/filling them, not when you're
constructing the path. So "rotate(); lineTo(); stroke()" is equivalent
(in Opera, not other browsers) to "lineTo(...); rotate(...);
stroke()". That means you can't do "lineTo(); rotate(); lineTo(); ...;
stroke()", because both lines get rotated by exactly the same amount.

(Opera's behaviour used to match the HTML5 canvas specification, but
then the spec changed to match what other browsers did, and Opera
hasn't been fixed yet.)

To make it work in all browsers, you'll probably have to not use
rotate() and instead do the trigonometry manually. I think it should
work if you change the "Thi bit draws the graph" loop to something
like:

     for (i in this.data) {
         /* ... same as before, then: */

         var angle = RGraph.degrees2Radians(360 / this.data.length) * i;
         if (i == 0) {
             this.context.moveTo(xPos*Math.cos(angle), xPos*Math.sin(angle));
         } else {
             this.context.lineTo(xPos*Math.cos(angle), xPos*Math.sin(angle));
         }
     }

--
Philip Taylor
excors@...

Hi,

Has anyone had any problem with using the rotate() method in
conjunction with Opera?

Eg: http://dev.rgraph.org/trash/tradar.html

Works fine in Firefox, Chrome and Safari, but not Opera. Cheers.

--
Richard Heyes

HTML5 Graphing for FF, Chrome, Opera and Safari:
http://www.rgraph.org (Updated November 29th)

> If I have ui elements in a canvas that I want to be clickable - what
> is the best way to go about doing that? Do I need to create some kind
> of collision detection to see if the mouse x,y is in a region? Or is
> there some better way?

You can:

1. Attach a function to the canvas onclick event
2. Get the mouse coords from the resultant event object.
3. Those, combined with some magic, will help you get the position of
the mouse cursor.

Below is a function which I use. If nested within a TABLE though, it
doesn't seem to work. :-( Any amount of DIV nesting does though. You
can see it in action here: http://dev.rgraph.org The top graph has
clickable bars.

http://www.phpguru.org/downloads/mousecoords.txt

--
Richard Heyes

HTML5 Graphing for FF, Chrome, Opera and Safari:
http://www.rgraph.org (Updated November 1st)

If I have ui elements in a canvas that I want to be clickable - what
is the best way to go about doing that? Do I need to create some kind
of collision detection to see if the mouse x,y is in a region? Or is
there some better way?