The latest thoughts on making things.
Don’t go to really understand something first, dig deep, and figure out the underlying principles. No matter how much you do, it’s just a low-level repetition. Go to the system to learn its knowledge first, gnaw through a book, and do not rush to do the project first. Go deep first.
No matter how much you do, it's just called there, and it's only at an amateur level.
It is better to report to a class first.
It's much better to see their flight control program after I got STM32 in the system.
The depth is not enough, no matter how much you do, it’s just a low-level repetition
Can be combined with this blog post
https://blog.csdn.net/sinat_16643223/article/details/108574301
I was right to emphasize the bottom three years ago, but at that time you might not know how to learn to gnaw through something. After going through the postgraduate entrance examination, I think you will.
You were right to hug the book three years ago.
You have just finished the postgraduate entrance exam, and it is right to chew on the Kalman filter book, just chew the above formula a little bit. This is only professional, just like my STM32 really understands the underlying principles and driver code. This is called professional. The implementation of functions under the previous call is really amateur, and I think it is good. After reading Huaqing's serial port class, I feel the blame is very deep. You just found out that you didn't even enter the embedded door, the STM32 door.
You have to gnaw professional information thoroughly. You can even apply for classes.
To get the algorithm, you have to gnaw the formula and push the formula by hand. To get the STM32, you have to go deep into the bottom layer, including the driver code, to figure out a little bit.
Instead of tensorflow running through a routine, it feels OK, and running through a routine written by a punctual atom will feel OK. You have to be able to write the low-level driver code in there yourself, which is called OK.
It’s okay if you make a drone SLAM. You can really gnaw through the Kalman filter book and gnaw through the SLAM algorithm. This is called OK, and you can communicate in front of people in the industry.
It is indeed faster to take things slowly. It seems to be quick to do things, but in fact there is no accumulation.
Don't rush to make one thing first, gnaw through one thing (a thing, a book, a project), bite it, it seems slow but it is fast.
You can feel confident in your heart only if you really have eaten something thoroughly. Otherwise, you have a guilty conscience, if you go to an interview.
。