Lua performance...

CasualEvolution

Hi, I came across with this video and I wonder why so much difference in performance between Corona and GS, if both finish in Lua...

quantumsheep

To be fair, the video's over a year old. I'm sure all the mentioned engines have improved since then.

And as one of my testers wrote to me about my latest game "This runs so smoothly – I’m assuming the GS engine has been improved quite a bit, although obviously some really good coding going on there too!"

I think it's a bit of both

QS

CasualEvolution

@quantumsheep said:
To be fair, the video's over a year old. I'm sure all the mentioned engines have improved since then.

And as one of my testers wrote to me about my latest game "This runs so smoothly – I’m assuming the GS engine has been improved quite a bit, although obviously some really good coding going on there too!"

I think it's a bit of both

QS

Hi @quantumsheep check this...

http://forums.toucharcade.com/showthread.php?t=238288&page=2

quantumsheep

I'm not sure what you're trying to prove with that, sorry.

Your initial premise was that according to an outdated video, there was a performance discrepancy between GS and Corona.

I suggested that the video was outdated.

How does what you link to corroborate that premise? All you've done, as far as I can see, is show a thread on TA where some people are having (minor) problems with just one GS made game.

Surely one game is not representative of all the games made with GS?

Cheers,

QS

BBEnk

I did the same Box test above with newest RC build 11.1.7 just now with 300 boxes at 40x40 and 600 boxes at 20x20 with retina checked on a mid 2010 mini mac.

Manto

@BBEnk said:
I did the same Box test above with newest RC build 11.1.7 just now with 300 boxes at 40x40 and 600 boxes at 20x20 with retina checked on a mid 2010 mini mac.

Sure that runs pretty smooth, however the test @casualevolution posted was done on Acer Iconia Tab A100 (dual-core 1GHz, pretty low end android tablet).

CasualEvolution

It is true that the video is old, but it show the difference in performance between GameSalad and Corona, and why doubt his veracity.
But my point is not the debate over which is better than GS (I love GS), is that if both are lua based, then an great increased performance seems possible in GS if we see the diference with Corona... the good thing is that GS can be more fast

Socks

@casualevolution said:
. . . . if both are lua based, then an great increased performance seems possible in GS if we see the diference with Corona... the good thing is that GS can be more fast

This is essentially the line that @CodeWizard‌ took when it was decided that working on a LUA-free version of GS should be abandoned, his reasoning was that LUA is capable of much greater performance than GS was (at the time) squeezing out of it - and that's the general direction we have going in since, performance is improving bit by bit.