Just figured I'd paste this here. This is what planetbeing said about giving ETA's a few days ago:
"The problem with giving any definite ETAs is that there can always be problems that crop up that blindside us that will take an unknown amount of time to solve (if they are even solvable). Let me give you a little timeline about this sandbox issue, for example (though I have to avoid giving currently classified technical details
).
Generally, constructing jailbreaks takes a lot of time researching issues: how, conceptually, are we going to break Apple's protections and mitigations against it. Then, to figure out what precise techniques will we use to implement our ideas. Those take an unknown amount time and may actually be impossible. It is theoretically possible for Apple, if they don't make any mistakes, to create a system that prevents jailbreaking. Then, after those are figured out, time can be spent coding the jailbreak, which takes a more predictable amount of engineering time. Unfortunately, sometimes while coding, you happen to discover the method you thought would theoretically work, even a method you tested before, might have unforeseen difficulties because you made some incorrect assumptions.
When we're at the research stage, we can't give an ETA because we don't even know if it's going to be possible. When we're at the engineering stage, we can't give an ETA because heaven forbid we find that we've made a mistaken assumption and need to go back to the research stage.
I started to actively participate on January 6th, and I thought I could contribute by helping to figure out the sandbox issues since that's the one thing that needed to be researched at that point. The following five days has been one of the more irritating weeks of my life. On day 1, pod2g and I both independently came up with a way to circumvent the sandbox that would've been nice and simple. Unfortunately, later that day we discovered we misread the sandbox profile and it would not be possible. The next day was spent trying to see if any clever variation of the first idea could get around the sandbox: no. Then we were messing around and found a small vulnerability in the sandbox, a one millimeter hole in a huge wall if you will, and it seemed like it could be weaponized to get around the sandbox so we can break out. Eventually, I came up with the plan that formed the basis of what we have right now, but it needed three different pieces to make work. I managed to furnish the first piece myself pretty quickly, but the other two were not forthcoming. The next few days were filled with brilliant ideas by brilliant people that would work if only such-and-such were true. Every day, it was two steps forward and one step back. Then pod2g made a suggestion on how the second piece could be obtained and saurik managed to find it fairly quickly going off his suggestion. The day before, we thought we had finally gotten it: The idea had gotten past basic testing, so we made a few tweets on progress. Then later we found out, crushingly, that there was a weird behavior that prevented our method from working in practice. The next day was us scrambling to figure out a variation of the idea that would work, which required finding other candidates for the idea's prerequisites. saurik managed to figure some stuff out that gave us a little more invaluable wiggle room in finding the prerequisites and I managed to find something that would work. I created a proof of concept and it worked!
The result is basically building a program that aims a tiny bullet to shoot out of the one millimeter hole in our sandbox, having it bounce off of a few different surfaces (that we were lucky to find) to adjust our aim and have it go down the exhaust port of the Death Star, and instead of blowing it up, bounce off a few more surfaces inside the Death Star to get to the control room, and have the bullet bounce off buttons and levers to aim the Death Star at the sandbox wall to blow it up.
We don't think there's any more issues and we are starting to engineer it, but we don't really want to say we don't think there's any more issues because one or more might crop up. So basically, we've had really significant progress, but we can't give an ETA."
Sign In »
Register Now!
Help
This topic is locked
Back to top
Helped you out? Press 
