Email: Password: Remember Me | Create Account (Free)

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread
???
01/30/06 04:34
Read: times


 
#108672 - My experinece
Responding to: ???'s previous message
Dear Stanley,
Well done to your work!!
I remember when I was at your age, I did write lots of programs in low level because there were only single board computers at that time such as the TK-80 from NEC, the MEK6800-D2 from Moto and I was using TI's TM990/189 board with TMS9980 CPU and also a uPC8080A CPU single board computer that only got 512bytes of memory! I was not using ASM at that time but pure mechine code!! That is eg. 3CH 03H 01H for JP 0103H. And then I needed to use the SBC's MON to enter the code digit by digit. And the built in monitor only got very little functions so every time if my code fail, the SBC just RESET, setting break points was not an easy thing in those old SBC indeed!! Therefore I needed days and weeks to finish a small program. Well this was not too bad anyway since the entire development time for my little project was lengthened in the implementation stage. And since the implementation, testing and debugging were difficult, then I became more series when I tryed to write the code after months of failure of getting the things work. I tried to get a clear understanding of what I want to do before I started to work. I wrote down the details in a project notebook. Say the purpose of the project, the methods that I use, the hardware details, and everytime I got any new idea, I would add to the notes. I also got a logbook that recorded down what I did and what problems found....etc. Well at that time the SBC could only store the mechine program in cassette tape so I need to record down where the code stored, and what is the file name..etc.
So when the time you get your code work, it is full of fun but belive me, when you review your project notes and your logg, it is also full of fun AND you WILL get many discoveries by doing so.

Today, implementation stage is not so difficult because you got the help of your PC and the development tools, the IDE, the higher level language...etc. However the problem of code managment still there. This time you find difficult to manage the large size code... Anyway I think my old method still can help,i.e. Good documatation and think clearly to avoid error!. Try to record down
1. project purpose- or a subset of the larger part of your prj
2. methods (algorithms) used and theory behind
3. any possible difficulties may encounter and the possible solns
4. the hardware details
5. any real problems encountered and the methods that you used to try to solve the prolbem, and the results
6. your code design- try using TOP DOWN DESIGN and MODULAR PROGRAMMING
6.1 record down everythings such as version number of the models, the name of the PRJ file if your IDE allows you to use that, ...etc. in your logbook..etc.

I suggest you to read some books on structure programming methods and software engg. I think you are not get the problem on software managment only but also need to solve the problem of managing your project. Those are key skill to your success in your projects and your learning.

As I read your last mail that you are going to enter to the University, and you are interested in robotic, I suggest you to read more books on Automatic Control, Digital control, Classical control. And you need to get familiar with some maths skill! - yes not interesting but you need it. The ODE, PDE is a must and Laplace transform, Z transform are also very very important. Try to get familiar with the P,PI and PID control algorithm...etc.

Hope every success in your projects and your learning!
Best 73 from Li.



List of 27 messages in thread
TopicAuthorDate
Servomotor head version III            01/01/70 00:00      
   Nice Work            01/01/70 00:00      
   Manageing code and debugging            01/01/70 00:00      
      Divide into modules            01/01/70 00:00      
         multi tasking            01/01/70 00:00      
            IRRC routine are much longer            01/01/70 00:00      
               No waiting for IR            01/01/70 00:00      
                  two mcu            01/01/70 00:00      
         Maybe this will help you            01/01/70 00:00      
            this is not IRQ save            01/01/70 00:00      
            RS232            01/01/70 00:00      
         Design First            01/01/70 00:00      
            Design            01/01/70 00:00      
               Pardon            01/01/70 00:00      
   Well done!            01/01/70 00:00      
      The "3.0"            01/01/70 00:00      
         Smart!            01/01/70 00:00      
         hardware            01/01/70 00:00      
            Mainland China            01/01/70 00:00      
      And Well done Kai!            01/01/70 00:00      
   My experinece            01/01/70 00:00      
      very impressive            01/01/70 00:00      
         Program and Code Structure            01/01/70 00:00      
            Spaghetti anyone ?            01/01/70 00:00      
      manually interpret machine code?            01/01/70 00:00      
         hand coded mechine language            01/01/70 00:00      
   Keep up the good work            01/01/70 00:00      

Back to Subject List