| |
I agree with most of Luis statement.
I've worked for three years in the embedded software development (for cars),
and I think it's quite dangerous to have RCX functions with the same name
that Unix ones, because they don't do the same thing. Using Unix style
functions is interesting for standardization, but it generally leads to a
lack a performance. I prefer having an efficent program than a standard one
(in the field of embedded software, not in the field of workstation programs).
Concerning the network protocol, I agree that LNP may not be sufficient for
some usages of the RCX. We must nevertheless keep the possibility to address
different layers of communication protocols so that we can choose wether to
include them into the kernel depending on the kind of communication the
application is using.
As I did for LNP, I volunteer for designing an implementation of a new
protocol for the host computer in Java, using the javax.comm extension
(which drives the serial ports).
|
|
Message is in Reply To:
 | | Re: API changes
|
| I'm actually on my way to my baccalaureate ceremony right now, but I do want to say that I feel very strongly that API changes for the sake of API changes should not be checked in. The current API, while not always intuitive, is well documented and (...) (23 years ago, 12-May-01, to lugnet.robotics.rcx.legos)
|
6 Messages in This Thread:
  
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
