Hey, check this out, JIT~

[LilGame]

yay just did it... 3.549

how da hell that guy got 5.215.

damn

[Cer[v]3r]

thats what im wondering

[hangover]

with this the battery of the cellphone will be less time or is more or what? Thanks.

[Danation]

with this the battery of the cellphone will be less time or is more or what? Thanks.

I doubt it will have a large effect on battery life. Maybe a little better, I guess, but probably not noticeable.

[HandlerExploit]

My thought on jit is it should slow your phone down over time like compcache.

[fogozito]

can anyone make and APK for this? i just tried and bricked my phone, i am restoring it now

lol i did something wrong im geetin .47

[MillyMoo]

It was definately faster booting up...after the random reboots i have had anyway. Have had 2 now and never really had any before on BrownMod v3. Will probably leave it for a while, see how it goes

[fogozito]

i am getting 3.7s

it keeps crashing motoblur and some other weird issues

[Danation]

My thought on jit is it should slow your phone down over time like compcache.

How so? Running native code is always going to be faster than interpreting.

[fastludeh22]

the problem with this JIT here is that it is the eclair version. it was really buggy and very unstable. also the reason for a 50% boost versus 550% boost in the froyo version

in general the eclair JIT was very unstable. that's why cyanogen left it out of his eclair rom

jrodd hit it on the head. the jit compiler was incomplete when eclair was released and why it was left out of the offical 2.1... however some people still swore by it.

...

Why?
JIT is a memory hog, and some people complained it made their HTC Dream (G1) slower than before contradicting their higher Linpack scores.

So yeah, their's a little more room for comfort (64MB) with the CLIQ/DEXT than the HTC Dream due to 256MB vs 192MB RAM. In my opinion, they could've gone with about 330MB RAM (anymore wouldn't be worth the performance gain) with these processors.

YouTube - Android 2.2 (Froyo) on the T-mobile G1 (CyanogenMod 6.0.0 RC1) Review!

i can confirm this as i have a g1. on 2.1 with jit i would get random weird stuff. sometimes u would have to turn it off to install a certain app then turn it back on. my experience on a g1 was benchmarks were higher, emulators seemed to run better once opened, but overall use of the phone had slowed down. this could be due to low mem of the g1 or unstable 2.1 jit. on 2.2 the unstable issues went away, but still didnt run good on the g1, this is def due to its low memory. maybe the cliqs have just enough. good luck testing!

[thrice43]

Heh, I've been using this for awhile now actually. Since Eclair2Cliq alpha4 actually. From my experience, with the new Eclair2cliq Beta1, you may get FC's on gapps occasionally.

When you check your dalvikvm settings using the terminal command

dalvikvm -h

you WILL see "segmentation fault", which I am not too sure if it is an error or not.
From what I know about the Jit compiler is that apps are precompiled on first run and then most of the compiled data is saved and changed in small portions instead of interpreting the app in entirety on the first run. As for performance gains, I've only seen gains in Eclair2Cliq, and have seen small, if any, gains in motoBlur roms.
JIT seems to be capable of being implemented in roms that are Android 1.6+. And IMO (emphasis on the IMO) I believe that due the slow pace of development on the Eclair and Froyo roms on the cliq, the amount of time needed to have a fully functional Eclair/Froyo rom will far exceed the lifetime of the Cliq. Too bad the development on the port of Cyanogenmod 4 is dead, I think we could've gained a whole lot more from it in a shorter amount of time.

[futado]

nice find man. i think my phone is faster. but i notice more fcs. usually only motoblur services though.

[toxigenicpoem]

Can't wait to try this out on my Cliq! Its @ the repair shop.

[VontaelP]

Nice. 3.701

[qilc]

Got up to 3.8, but also got a lot of FC and random reboots.
After a day, went back to 2.6 and stability.