Are applets slow?

Question

Hi all,
Quick question. I am developing a Java Applet and an Application. And, I just want some feedback from the pros on the difference between exectution speed between an applet and an application..
[1] Are Java Applets slower or faster - compared to an equivalent Java Application?
[2] If there is a difference in speed between and Applet and an Application, can you explain why there is a speed difference? Is there any extra overhead?
[3] If you do not want to re-invent the wheel, can send me link to anything that might answer the above questions.
thanks,
-Adam

Answer 1

cotton.m wrote:
JustSomeGuy wrote:
It doesn't necessarily have to be a browser. But an applet does require a host of some sort and therefore all the resources of that host. So yes, technically the applet requires more resources.
I'm sorry but "technically" it does no such thing.
The applet runs in the VM and that process, as Kaj noted may well be outside of the browser (or whatever) process and so no the applet does not require more resources. Hmm, I think you and I have different definitions of "require". When I say require, I mean that you can not under any circumstances ever open an applet without a host. Since the applet requires the host, it requires the hosts resources as well (funny thing about host programs, they require resources to exist). Therefore, an applet technically requires more resources.
cotton.m wrote:
I understand your point but it's important to be clear else you leave future readers a faulty impression. Your point is akin to my saying that a Java program is slow because I have a USB printer. I mean just because someone had a bad experience with an applet because their spyware infested bloat-browser chews up all the system resources doesn't mean the applet is accountable for that or that this is an inherent fault in the concept of applets. I'm sorry but your comparison does not match. Unless you specifically went out of your way and hardcoded your printer in, then the java program does not care what kind of printer you have. That is to say, your usb printer is not required to run the program so it is not the same as an applet which does require it's host.
Consider this if you will, an applet and application are written to do exactly the same thing. They have more or less pretty much identical code. Now, on the same machine, run the applet after a cold start and benchmark it. Then cold start the machine again and benchmark the apllication. Which wins?
If you concern yourself with just the applet vrs the application they should be identical (actually not true, the applet should have a slight advantage) but because of the extra resources consumed by the host which must be started for the applet to run but is not started for the application, the applet is going to loose every time.
My point in mentioning this was exactly to prevent faulty impressions. If you look at the big picture an applet is actually more resource dependant and slightly slower (though admittedly most users will never notice). It is not a fault in the concept of applets but rather a tradeoff for the benefits of applets. The whole purpose of an applet is to provide functionality that the host (usually a browser) does not have, and to do it inside of that host. So, if going by the pure purpose of applets, there should never be an application that matches it, if there is then the applet should have been an application in the first place.
cotton.m wrote:
I think a better way of phrasing it would be that on some systems an applet's operating environment may be performance restricted due to a lack of system resources. That is more or less what I said, the only difference is I actually provided an example of just such a system.