Update on pixel header

[aodonnell-ca]

Not an issue, but I have some information for you.

Firstly, thanks for your very informative page https://alusch.github.io/flipdot-protocol/
Your doc was a great jump start for me.

I've been working on a flipdot module, testing against the SCP (sign control program) by ecbarnum. I'm not sure how I got that program, it came via a circuitous chain of technicians in the industry.

Using SCP I think I can fill insome of the blanks re the pixel map preamble in https://github.com/alusch/flipdot/blob/master/libs/core/src/page.rs

///                   ┌─┬ ┄ ┬─┐
///              Bits │7│...│0│
///                   └─┴ ┄ ┴─┘
///                    \     /
/// ┌────┬─────┬────┬────┬────┬────┬────┬────┬────┬────┬ ┄ ┬────┬ ┄ ┬────┐
/// │ ID │ DLY │   SFX   │  0 │  1 │  2 │  3 │  4 │  5 │...│0xFF│...│0xFF│
/// └────┴─────┴────┴────┴────┴────┴────┴────┴────┴────┴ ┄ ┴────┴ ┄ ┴────┘
/// ┆   4-byte header   ┆            Data bytes           ┆   Padding   ┆

The header is

• ID 8bits unsigned int page number in sequence 1's based. I still don't know an upper limit, but I suspect 64 in some Luminators, just based on size of on-board RAM.
• Persistence (deciseconds). 8 bits unsigned int Fixed point representation of message visible timing for self timed displays. eg 10 (0x0a) = 1.0 second. 5 = 0.5 seconds.
• SFX. 16 bits flags? I can confirm
  • 0x0400 = "from left" where pixels transition one column at a time from current display to incoming page
  • 0x0800 = "from top", though the exact animation I am not sure of. I believe its a scroll, not a reveal
  • 0x1000 = "from bottom".

Right now, I can't confirm any other animation patterns, though I have seen SCP samples with alternate "from left" codes, the sign we have for reference, Luminator 7x75 display, show all "from left" patterns the same.

I figured this is the least I can do to repay you for what you taught me. I'm coding on the sign side, not the host side, and in a tiny 8 bit MCU, so I've done it all in C, as a consequence I didn't find your repo in earlier GitHub searches. It looks like most the work was done a long time ago, but since you were updating the code 9 months ago I figured it was worth letting you in on these "secrets".

[alusch]

Thanks @aodonnell-ca, I appreciate it! Unfortunately, when I gave the flip direction flags a try, my MAX 3000 just did everything from the left 😞 I did want to confirm that the flags are big-endian, i.e. for "from bottom" the last two bytes of the header would be 10 00? Maybe Luminator decided nobody was using the extra modes and didn't bother including support in my sign.

[aodonnell-ca]

Yes, the flags SCP used are big endian, and the SCP emulates a sign, but I don't know which.

I see I also put DLY in the protocol, but described in terms of persistence.

I never had the system I was emulating, that was always in our Southern California office, but its a retired model and about 15-20 years old, so it may not be relevant to any modern display. That's the trouble with reverse engineering remotely.