a couple misc updates, including different foreground color

[agentbooth]

Hey @philpem and @arnoldrobbins -- I updated the foreground color to try to be more like the original monitor color after looking at various 3b1 monitor photos -- I hope you don't hate it. If not, let me know and we can go with whatever default you want to. I also updated the print statement on boot to reflect the window size (so you can validate what window size is being used when using the -s scaling option).

I was looking into the expansion slots for fun the other day so put a few notes in on those.

Also realized that sending 0xFF when queried about the modem will be interpreted as "no modem" which I was hoping might help with the .phinit hang on installation but alas it did not.

And final fix was to the logging of the baud rate setting for rs232 -- not material, but I did start getting more activity on that log entry when messing with the serial port so figured I would correct it. e.g. if you do 'stty 1200', 'stty 300', 'stty 9600', etc when logged in through the serial port, you'll see calls with the respective speeds to that baud gen write.

[agentbooth]

also just pushed a commit to use localtime, ahhhh, my 3b1 clock is now correct :)

[philpem]

I like the modem and timezone work - nice work there. I hadn't realised the hardware clock was set to local time.

Regarding phosphor colours, I'd be a little wary of using photos to derive CRT phosphor colours (or at least their RGB equivalents). Most of the photos I've seen tended to be overexposed, leading to the phosphor colour being brighter than it looks with the naked eye.

If we want to tweak this further, there's a thread on StackExchange about phosphor colour equivalents: https://superuser.com/questions/361297/what-colour-is-the-dark-green-on-old-fashioned-green-screen-computer-displays

Looking at the phosphor table on Wikipedia (https://en.wikipedia.org/wiki/Phosphor#Standard_phosphor_types), the P1 green-screen phosphor is very close in wavelength to the green phosphor used in colour CRTs. Peak green is probably very close to the real colour. The difference is, on a colour CRT, the shadow mask will reduce the effective sharpness and resolution of the beam.

I like the idea of configurable CRT colour, but I think it might work better if we had it as a command-line parameter, e.g. "-c green1", "-c amber", "-c minty", or a HTML hex code "-c #00FF00".

[agentbooth]

I'm honestly not sure about the hardware clock time -- but at least it gives me the correct time now. (My /etc/TZ is set to EST5). This may be why during the install process it asks to reboot after you set the time and timezone. It probably records somewhere during the install what your "h/w TZ" is, because if I change /etc/TZ to PST8, logout and log back in, it will give me the time offset by 3 hours correctly. (But the title bar time still remains EST.)

Haha, yeah I was reading those same articles too trying to find something that made sense to me about the phosphor -> hexcode conversion. I took a sampling a colors from different 3b1 photos and the blue component definitely seemed significantly higher than the red component. I err'd towards some lower values as I didn't want to lighten it too much. I just tried 33FF33 that's referenced in the article for P1 phosphor and it just looks a little more subdued green. We could go with 33FF33 if you want? Or just leave it at 00FF00. Let me know what you prefer. This would be another great option to put in that config file :)

[arnoldrobbins]

I honestly don't remember what exactly the screen looked like. "Minty" is a good description for your change though.

[agentbooth]

Hey @philpem -- I reverted back to the pure green, and left my "minty" color in a comment.

[philpem]

Hey @philpem -- I reverted back to the pure green, and left my "minty" color in a comment.

Okay, fair enough -- is this ready for review now?

[agentbooth]

Yes- ready for review.