Any thoughts on speed limitations of Revlet?

James Hurley jhurley0305 at sbcglobal.net
Sun Jul 26 13:55:07 EDT 2009


>
> Message: 5
> Date: Sat, 25 Jul 2009 18:05:53 -0700
> From: Richard Gaskin <ambassador at fourthworld.com>
> Subject: Re: Any thoughts on speed limitations of Revlet?
> To: How to use Revolution <use-revolution at lists.runrev.com>
> Message-ID: <4A6BABF1.2030204 at fourthworld.com>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> James Hurley wrote:
>> Before I begin to dream about the potential of Revlet, I would like  
>> to
>> have some idea about the potential, particularly the speed in running
>> graphics.
>
> I haven't tested graphics operations, and would be interested to  
> hear of
> any differences from those.  But FWIW, I ran RevBench in both the
> browser and the IDE and could discern no significant differences in
> performance.
>
> --
>  Richard Gaskin
>  Fourth World
>  Revolution training and consulting: http://www.fourthworld.com
>  Webzine for Rev developers: http://www.revjournal.com
>
>
> ------------------------------
>
> Message: 6
> Date: Sun, 26 Jul 2009 11:21:06 +1000
> From: Sarah Reichelt <sarah.reichelt at gmail.com>
> Subject: Re: Any thoughts on speed limitations of Revlet?
> To: How to use Revolution <use-revolution at lists.runrev.com>
> Message-ID:
> 	<f99b52860907251821y7524738co9c14e5615af9900d at mail.gmail.com>
> Content-Type: text/plain; charset=ISO-8859-1

(Snip)


>>
>
> I built your image moving stack and uploaded it to
> <http://troz.on-rev.com/rainbow/move.html>.
> It seems to work at full speed, so it isn't the mouseMove itself, but
> the amount of processing that has to happen on every mouseMove.
> I tried several options: locking the screen, only drawing the lines if
> the torch has moved more than a certain number of pixels, limiting the
> number of redraws per second, flushing events. I even tried replacing
> the torch image with a button in case redrawing the image was too
> much.
> I think I made it a bit better, but it's still not great, but you can
> try it again at the same address.
>
> As regards the odd URL, in the standalone settings, I recommend
> entering a name for the revlet with no spaces or other odd characters.
> When you build the revlet, you get 2 files: the revlet file and a
> test.html file. Rename the test.html file if you want and then upload
> them both into the same folder of your on-rev site. Then the web
> address will be: http://yourname.on-rev.com/foldername/test.html
> (filling in the blanks as appropriate).
>
> Cheers,
> Sarah

Sarah and Richard,

I have uploaded an old stack of mine which is both calculational and  
graphic-display intensive, my old Nine Ball simulation:

go url "http://jamesphurley.com/NineBallWithSpin.rev"

As I suspected from both Richard's and Sarah's findings that the  
calculational parts (calculating the positions and velocities of all  
the balls and determining the new positions and velocities after    
collisions) run very well in Safari. But the motion of the balls is  
very stuttered. It runs at full speed, that is it takes the same time  
to execute one "shot" but the motion of the balls is herky-jerky.


I am puzzled, Sarah, when you say that the mouseMove stack works at  
full speed and then you say that it is not great. (And you speak of  
the torch. Did you mean the winter scene of my barn?) Maybe you mean  
it runs at full speed but is bumpy? Did you find the torch image  
breaking up into two pieces as you dragged it?

I also tried "grab me" but with the  same bumpy results.

This is something that I hope can be remedied. After all, this is just  
alpha. We have the whole greek alphabet to go.

Sarah, thanks for the info on getting an app on the web. I'll work on  
it.

Jim






More information about the use-livecode mailing list