proof that saurik found ctl.restart from init.rc

cielavenir >
is saurik available? or sleeping?
so... as I said several weeks ago, I needed to decompile substrate installer to see why I couldn't install substrate on my HTCJ (and I was able reproduce the failure using adb shell)
very sorry about that
now, the question is how you selected the rebooting method:
setprop ctl.restart zygote
I and homar found that this method is the most compatible around many devices
some HTC device crashes in start;stop or killall zygote
We have discussed about this on XDA and have found setprop is working
so, you selected setprop by careful analysis?
well, I think this finding should be credited, right?

saurik >
so, let me get this straight: you are coming here to challenge me that I've somehow come up with something you came up with, and that I should be crediting you due to something you said on XDA? is that the idea here?
here's why I use ctl.restart: on January 9th pulser was testing a beta version of Substrate, and complained that it failed on his Galaxy Note
after a bunch of debugging with him, I determined that it was due to "onrestart write /proc/sysrq-trigger c"
  6138 15:20:40 < saurik> ooooo
  6139 15:21:29 < saurik> 'c'  - Will perform a system crash by a NULL pointer dereference.
  6140 15:21:29 < saurik>           A crashdump will be taken if configured.
  6141 15:21:33 < saurik> why the fuck would they do that
  6142 15:23:28 < pulser_> who knows?
  6143 15:32:50 < saurik> apparently android 2.2 added a ctl.restart control message, so I can use that instead
  6144 15:37:27 < saurik> (maybe even only Android 2.3, but that just barely manages to cover my deployment target, so it works)
  6145 15:38:14 < saurik> pulser_: so, I guess sanity check: `setprop ctl.restart zygote` does not reboot, right?
  6147 15:38:32 < saurik> (I feel like it /should/, but I have reason to believe that it doesn't)
  6148 15:38:41 < saurik> (like, I consider it a bug that it doesn't, but it is a bug that really is there)
  6149 15:39:00 < saurik> (in essence, ctl.restart is considered stop/start, not "restart", so "onrestart" doesn't fire)
  6150 15:49:51 < pulser_> saurik, let me test
  6151 15:50:23 < pulser_> that gave me a "soft reboot", but without going back to bootloader
  6152 15:50:25 < pulser_> basically a "stop start"
  6153 15:51:23 < saurik> ok, great. next version of Substrate will do that
  6154 15:51:34 < pulser_> :)
so that's the path: pulser reported a bug, I tracked it down to a specific init.rc fragment and a corresponding codepath in init. I then looked for things in init that surrounded the restart code for one that wouldn't trigger that path, and came across ctl.restart. ctl.restart also just sounds better than doing a killall myself of zygote, and stop;start was off the table long before that due to race conditions. I later noticed that I was wrong about it being in 2.2: it was only added in 3.0, so I do the fallback based on API level

cielavenir >
oh it was init.rc stuff...
anyway we couldn't think of init.rc, so our works are really derivative lol
ok thank you
ah sorry now I understood the first the sentence
well, it's opposite; our ExtraPowerMenu is substrates's derivative.
sorry to have worried you

saurik >
ah, ok