Flash and Backwards Compatibility

[pixelwit]

First let me start by saying I'm not a Computer Science guy or even really an Internet guy but I do seem to get along with logical concepts which "kind of" makes me a "logic" guy. So being the logical and inquisitive type that I am, I just have to ask this question...

Why does Flash maintain backwards compatibility?

The usual answer is that if they didn't then all the old code would break whenever Macromedia released a new player. But every time they release a new player they don't "update" the old player with only the new features, they completely rip out the old player and install a brand spanking new one. Why do they rip out the old one? Why not keep the old player for old content and install a new player for new content then just specify which plugin is needed in your HTML page? Wouldn't this allow them to make some speed optimizations or maybe get rid of some old legacy code that's no longer needed?

The curiously stupid mint,

-PiXELWiT
http://www.pixelwit.com

[Cota]

I look at it this way, ours is not to question, but to use and profit from. But I do often ask myself the same question. I've come up with this answer, because they just do. I'm sure they have a reason, but I dont think we'll ever know, but if someone has the answer, do share.

[stealthelephant]

bootstrapping i imagine

[pixelwit]

Well my best guess is that it has something to do with tiny little devices and their tiny little hard drives but I can't really say for sure. Wouldn't mind knowing the whole story though.

-PiXELWiT
http://www.pixelwit.com

[farafiro]

then just specify which plugin is needed in your HTML page?do u mean the plugin automatically adjust itself (or the browser agjust it), or the users do this??
I think in both ways, it's outta their hands.
you got a cool point in here, but if u r MM (or u own it) what would you do??
would you keep loading all these players plugins to all kinda browsers and also try to satisfy all kinda users
donno, but it's hard to chose

[pixelwit]

What I meant was that you wouldn't just have a SWF plugin type, but more of a SWF1 plugin, SWF2 etc. and each file type would prompt you to download the necessary plugin if you didn't already have it.

It's not that I expect Macromedia to do something like this for every player update but wouldn't it make sense to do something like this at one point in time or another?

-PiXELWiT
http://www.pixelwit.com

[CyanBlue]

Yeah... What pixelwit mentioned makes more sense... The only problem that I can think of with that is the instance where two plugins are needed at the same time... For example, I have a game that is aimed to the public who has F5 plugin installed, and the banner on top of the page happened to be published for F7, then somehow the user's computer need to run two instances of ActivX controls for them... I am not sure if that is good or bad thing though... (In terms of the processing speed and whatever else...)

[boyzdynasty]

Why do they rip out the old one? Why not keep the old player for old content and install a new player for new content then just specify which plugin is needed in your HTML page? Wouldn't this allow them to make some speed optimizations or maybe get rid of some old legacy code that's no longer needed?

The curiously stupid mint,

-PiXELWiT
http://www.pixelwit.com


Should come by the meeting and ask that question :D

[pixelwit]

If I can get there I'll be sure to ask.

-PiXELWiT
http://www.pixelwit.com

[Korona]

i dont actually see any advantages in what your proposing. for one, legacy bytecode runs faster on FP7 than it does on say, FP6. they dont "update" the player sequentially because its not worth thier while implementing new code which is fully compatable with an OLDER and SLOWER framework - the activeX is ~1MB anyway, so you might as well replace it. During development, you can specify which minimum version to target, so it will run on an older version of the player.

i dont see the logic, or have i missed something?

[pixelwit]

Flash 4 syntax "currently" runs faster than Flash 5 syntax. Would performance for F5 syntax increase if the player was optimized for F5 and no longer supported F4 syntax?

-PiXELWiT
http://www.pixelwit.com

[Korona]

"Flash 4 syntax "currently" runs faster than Flash 5 syntax." i just googled about this, and was shocked reading this (www.fatorcaos.com.br/flash5tests/speed_1.html ), just like its author was.

this is more to do with how the compiler works with the two different syntax types than anything else. the bytecode produced by the compiler could be the same, because the two syntax types do the exact same thing, but the newer syntax just isnt as optimized.

However, i think you would find that even old syntax would still run faster on newer players than it does on older ones. i remember reading that some of the core classes (such as Array and String) are hundreds of % faster in FP6 than they are in FP5. (this isnt anything to do with syntax, but how the player itself works).

on a side note, i did the tests in that link again, but with FP7. it turns out that variable reference with the new syntax is about 5-10% faster than the old syntax. This is partly because of some of the optimizations the new compiler performs (and apparently it only does "mild" optimizations, so there is still room for some improvement).

[pixelwit]

I just finished reading this article (http://ohwhen.typepad.com/mx_traveller/2004/05/the_mystery_of_.html) from MX_traveller concerning the ever-unpopular "pixel-shift". It appears that the problem could be solved but...

"It would take several hotshot graphics engineers many months to do it, and it would have to be a coordinated effort for both the Flash Authoring Tool and Flash Player. There would be mammoth considerations for compatibility with existing FLA and SWF files to not break anything. A lot of up-front design work would have to happen, followed by a long and hard implementation and testing phase."

To me, it sounds like the quality of the SWF player suffers in part due to the sheer complexity of maintaining backwards compatibility.

Any thoughts?

-PiXELWiT
http://www.pixelwit.com