[NTLK] Learning Remote App?

[Matej Horvat]

> Is it possible to take the source from TV PAD and rewrite it
> to be a learning remote?

I don't think so. Quoting the developer Q&A documents:

Q: What are the hardware specifications for the Newton IR port?

A: In the Apple MessagePad 100, 110, and 120, the Sharp ExpertPad, and the Motorola Marco, the IR transmitter/receiver is a Sharp Infrared Data Communication Unit model RY5BD11 connected to channel B of a Zilog 85C30 SCC. Data is communicated along a 500 KHz carrier frequency at 9600 or 19200 baud, 8 data bits, 1 stop bit, odd parity. The IR hardware requires a minimum of 5 milliseconds settling time when transitioning between sending and receiving. Sharp's CE-IR2 wireless interface unit may be used to connect the Newton to MacOS or DOS machines, with the appropriate software.

The Newton supports four IR software data modes:
Sharp encoding, NewtIR protocol (specifications are NOT releaseable)
Sharp encoding, SharpIR protocol
Plain Serial
38 KHz encoding ("TV Remote Control")

Q: Data sheets indicate Newton IR communications occur at speeds up to 38.4 kbps.  Communications documentation implies that 19.2 kbps is the top speed. Which is correct?

A: IR communications occur at 9600 baud when using the Sharp IR protocol (the kCMSSlowIR).  When two Newtons are beaming to each other, a slightly faster protocol is used, which communicates at 19.2K baud.  PIE engineering reserves the right to modify this protocol (including transmission speed) in current and future products and versions of the Newton operating system.  This protocol is intended only for internal Newton communications.  Developers should use the kCMSSlowIR endpoint in their applications.

The 38.4 kbps speed (actually 38.4 kHz) refers to a mode of communication in which the Newton's serial communcations controller chip is not involved.  This system is used by "TV remote controls" and is a send-only mechanism.  Data is "bit-blasted" as a 38.4 kHz square-wave through the IR port by a software loop, bypassing the SCC entirely.  Besides being a very inefficient method of data transmission, this mechanism has no provision for error detection/correction.  The data to be transmitted must be described as a large binary object of square-wave state transitions and is therefore specific to each vendor's remote control protocol.

(end of quote)

However, this documentation was not updated for the MP2x00/eMate hardware, I've never done IR communication on Newton devices, and I only consider myself an expert for 1.x devices, so I can't say for sure.

Matej Horvat