Thanks for making me feel old!
One of my first real jobs involved MIDP/J2ME development. I cannot say I loved the platform, but it somehow felt more magical back then to develop something and run it on a Nokia brick than (mobile) development of today.
The way I understand it, the shader overlay can only modify what is already being rendered below. It does not have access to the underlying application logic, 3D geometry or other internals of a game that you would need for this. You can increase the contrast etc of a game but you cannot see through walls or anything, so it probably won’t help you cheat.
* Acquiring Evernote, laying off most of its staff and raising prices.
* Acquiring WeTransfer, announcing 75% layoffs and pissing off their most loyal users by changing their T&Cs to grant themselves license to use their content for AI training purposes.
* Acquiring Filmic and laying off all their staff.
* Acquiring Komoot and laying off most of its staff.
Now would be a good time to poach some Vimeo engineers.
"Let's be honest: The current generation of robotic lawn mowers sucks. Basically all of these bots drive in a random direction until they hit the border of the lawn, rotate for a randomized duration and repeat. I think we can do better!"
The funny thing is: this actually works incredibly well. Perimeter wires are a PITA to install, but once that's done, they are a very practical and flawless method for making sure the robot does not escape into the neighbour's yard or worse. The random movement is really effective too. What exactly can a smart robot do better?
Removing the need for perimeter wires would be great, as long as it works 100% flawlessly. Obstacle detection would also be nice, so I can avoid my mower chewing up the toys my kid sometimes leaves lying around (though it is a great motivation to clean up!)
I have a Mammotion Yuba and trust me, the grass looks awesome as a grid or in lines. It can even do logos. So far nicer looking grass and much faster then random.
It looks like obstacle avoidance is the key thing remaining in the software todos. For positioning it seems you get your pick of RTK GPS sensors, so it'd be interesting to still support guide wires for "escape protection".
Regardless of how good the perimeter wire bots are, it's also not true that the more advanced generation "sucks." I have one and it works perfectly fine (Mammotion Luba 2). The hardware is great, positioning is great; there is always stuff to nitpick on the software side but at the end of the day that's great as well.
Amazing that they were able to figure this out. I have a PET that also sometimes boots up into a garbled screen similar to the one pictured in the article. I am usually able to get past it with a few hard resets. I don’t have access to (or much knowledge of) logic analysers, but this almost makes it look doable to figure out the problem.
If you want to go that route, you may find it easier to check your logic analyzer dump side-by-side with the debugger in an emulator (e.g. VICE). Break on the first instruction and then go step-by-step. (Just need to make sure you have the same ROM version, but that shouldn't be too hard.)
However, your problem kind of sounds like a power supply problem. So using a logic analyzer will maybe just produce a different result every time. So maybe check the 12V and 5V rails on an oscilloscope while turning on the computer. (Or maybe it's a problem with the reset circuit, etc.)
WARNING: nerd snipe material.
reply