Prevue Guide Forums

AriX wrote:As further evidence to swest's finding that the EPG channel was started way before 1985, check out trademark serial number 73398017 for the Electronic Program Guide (and logo). It was filed in 1982.

And used ("FIRST USE: 19810923", "FIRST USE IN COMMERCE: 19811109") as early as September 1981, so that seems to sync perfectly with what the SEC report says.

@LH - Yes, that looks like exactly the same thing. I wonder if the failure of the video overlays to appear has anything to do with why they were pressing Ctrl-U? (Like, if they were pressing it in response to that, is what would be an interesting piece of information.)

About the two different Sneak Prevues, I did have it at one point on my old cable system, but don't know which it was and can't offer specific information on it (other than things learned on the net) because I had nil to do with it with UV.

And about EPG/PG's VBI data, all I could tell from looking at WGN's VBI way back when -- and not with a frequency monitor but with a proc amp that simply let me to roll it onto the screen for visual inspection -- was that it had lots of data in it. Because I knew their VBI "fed Prevue Guide," I assumed the bulk of it was UV's. Now that it appears UV used a dedicated data subcarrier for "text" (which I assume was that satellite list's author's word for "listings"), and mindful of those EPG Data vs. Control Data traces on the VBI decoder card, I'm thinking the VBI must have possibly only been for the "control" data (think separate data channel for anything that needs to be communicated instantly, so it won't end up being delayed by a traffic jam on the 2400 baud feed).

Re: "data transmitted as an audio feed too", technically this 7.237 MHz subcarrier I'm talking about was both. Data as audio... just like a dial-up modem on a phone line. So in total, it looks like the satellite decoder card may have watched two sources: WGN's VBI and 7.237 MHz (or for non-EPG boxes, PG's VBI and maybe 7.237 MHz there too).

@edit - It is also entirely possible that this 7.237 MHz subcarrier was only used with the pre-1985 units. Really, unless more information can be found, it's probably pointless for me to have speculated above as far as I did. I just know there was VBI data from 1985 onwards, and (thanks to this PDF) that there was an audio data feed before it.

LocalH wrote:Shift-D seems to read something from track 51.

Shift-D makes the program re-load curday.dat, OI_01.dat, nxtday.dat, 0I_00.dat and oinfo.dat

(slight edit sorry bout that if you replied before I edited )

tin wrote:Shift-D makes the program re-load curday.dat, OI_01.dat, nxtday.dat, 0I_00.dat and oinfo.dat

(slight edit sorry bout that if you replied before I edited )

Cool, how'd you find that exactly? :O

WOW! set your PC/Mac to 31st March 2004 and restart your emulator

Very nice. That means that all that is somewhere stored on the disc - I wonder if it's in the "raw" format that was sent across the RS-232.

One thing I noticed - on the copy of the disk image that I've been messing around with, this did not work. I had to revert back to the original one downloaded here.

tin wrote:WOW! set your PC/Mac to 31st March 2004 and restart your emulator

Aha it seems the year doesn't matter, and my guess at this data coming from 2004 was way off

I will now explain....

Having done quite some work on the Atari reverse engineer, it became clear that PG refers to any given date as number of days elapsed since jan 1st. It basically ignores what year it is, but it does take into account what year it is to calcluate how many days in February when doing that calculation. Notably the value is stored in an 8 bit register, so the value wraps round at the 255th day of the year.

Today is known as CURDAY, and Tomorrow is known as NXTDAY. They're also stored as values 00 and 01 a lot. All the routines on the Atari check incoming data is relating to CURDAY or NXTDAY and bins it if not. The 00 and 01 are often used as offsets to look up information depending on if the data is for/from CURDAY or NXTDAY. This lot I know, how it stores the full data, I don't yet know.

So far we're assuming that the Amiga versions use the same data format for the listings, and so far they appear similar if not exactly the same.

As we have seen from our .adf, there are two files, curday.dat and nxtday.dat. Once decompressed, they contain a lot of control chars, and the ASCII relating to programs during some random day of the year - evidently a cached version of the program information - so a freshly booted PG won't have no data like some of the videos we see on youtube.

There are also the files OI_00.dat OI_01.dat and oinfo.dat - the first two contain some strings that are shown on the PPV stuff, and the oinfo just has the digits 90. I noticed OI_00 and OI_01 have the digits 90 and 91 at the top...

Put that lot together and you wonder if the cached program information is from the 90th day of some year. I guessed at 2004 cos a lot of the file timestamps are jan 2004 (but why, I don't know, AriX rightly informs me Amiga PG was gone by 2000)

AriX wrote:
Cool, how'd you find that exactly? :O

I've installed PG onto a hard disk, and made it run snoopdos during the startup-sequence, so I can see what files are being accessed

Given the copyright dates at the end of curday.dat, I'm going to go ahead and say that the last time this disk was booted in an authentic AP2000 machine was March 31, 1999. It must have been decommissioned early, though - it's my understanding that the Amiga-based infrastructure wasn't replaced until December of that year.

I think you are right, the porn was filmed in 1999 so I guess it has to be

LocalH wrote:Given the copyright dates at the end of curday.dat, I'm going to go ahead and say that the last time this disk was booted in an authentic AP2000 machine was March 31, 1999. It must have been decommissioned early, though - it's my understanding that the Amiga-based infrastructure wasn't replaced until December of that year.

Yes, I was confused by that as well... I wonder why it was decommissioned so early? (Especially given that it was a working machine according to Jeff and co.)