It's been the subject of debate. For a long time, and even on Sega Retro, there was a rumor that it was literally the SMS cart on a Sega Mega Drive PCB, with the lines physically switched so it'd boot into SMS mode, but a person on SMS Power with the cart debunked this. Turns out the cart lines are not physically jumped, so it doesn't boot directly into SMS mode, and when he dumped the cart, it was actually a different build from all known SMS builds (but actually matches the build on the Sega Saturn rerelease of phantasy star). It uses software mapping to force the Mega Drive into SMS mode, not hardware mapping. How it does this isn't understood well. Part of the mystery is that the JPN version of Phantasy Star uses the SMS Bios to set up the stack pointer, which the Mega Drive lacks. So it apparently has some bootstrapping code that is run before the game starts.
Two so far! AI T-152016-00 Math Blaster: Episode 1/Technical information AI T-87056 -00 The Miracle Piano Teaching System (software)/Technical information Tell your friends - there was "AI" software after the AI Computer. For the record I've done about ~450 Mega Drive games, aka ROMs I just happen to have lying around from previous experiments. I won't be covering all of the games on my own, so a great time for you guys to to learn all this stuff.
~Mystery time~ So the very last row in a Mega Drive-style header is meant to be "region", or more specifically J (for Japan), U (for the US), and E (for Europe), which realistically equates to NTSC-J, NTSC-U and PAL. This has been a thing since time immemorial - it's why all those old ROM sites tagged things with [JUE] - it's well understood, even though it neglects many key markets like Brazil, South Korea and Australia and translated versions for Western Europe. It's also very often wrong. In ROM header terms, JUE means "probably works on any system", but that doesn't mean the game was sold in every region. Or indeed that it actually works on any system. It wasn't policed properly so you can never go off headers alone... and also there are other symbols: 1 4 5 8 A B C F They're all coming up too often to be a mistake. My guess would be that they're multiple regions - I thought maybe "translations" but some aren't translated, and they're not meant to be software revisions. Obviously they might all be mistakes (there are plenty in these headers after all), but take a game like After Burner Complete - a proper Sega-published Sega game: region "5" or region "A".
They changed the date format in a technical bulletin published in November 1994. It's probably on Sega Retro somewhere in here, but I don't know where. Basically, instead of just J/U/E for Japan/US/Europe, you have a number with four bits, representing: Japan NTSC (Japan, Korea, Taiwan), Japan PAL (not used), Overseas NTSC (North America, Brazil), and Overseas PAL (Europe, Australia, Hong Kong). Then the number is encoded as an ASCII character. In effect, it's kind of like chmod in Unix, except using four bits instead of three (so hex instead of octal). You add 1 (0b0001) for NTSC Japan, 4 (0b0100) for NTSC America, or 8 (0b1000) for PAL Europe, with 2 (0b0010) being unused since it represents PAL Japan. Then you have one byte that represents all four possible regions (with the rest of the bytes for the region field being filled with space characters) instead of needing to have a separate byte for each region, and you can check for each region by checking for its particular bit. So US-only would be 4 (0b0100) and all regions would be D (0b1101). Or maybe they just used F (0b1111), I don't remember. In theory, there's a possible conflict with the old format with E (0x1110) representing PAL Europe, NTSC America, and PAL Japan. In practice, it never happens because there's no PAL Japan, so you should be able to reliably detect the new format by checking if it's one non-space character and not J/U/E. At least, that's what I remember, I might've gotten a detail wrong. I wrote a utility to do this in Visual Basic years ago, lol.
Getting better every day From the looks of things this new scheme was usually followed. The documentation for this new scheme dates back to 1994, but the earlier scheme still turns up in 1995. More opportunities to get the header wrong.
https://web.archive.org/web/20201221193902/http://mode5.net/32x-DDK/toc.html https://web.archive.org/web/20201201204058/https://mode5.net/32x-DDK/32Xtoc.html
I didn't think this would have been possible but then I suppose that the average flash cart must do something similar to support master system games, though I've heard the X'Eye, for whatever reason, doesn't actually support the flash carts that try this.
Just making a note that Sega Asia just changed it's website in the last seven days which added pages to their PlayStation and Nintendo Switch pages, adding a Steam tab, and adding an English Language version to the websites language select which now shows us which releases have Asian English versions. I'll also note that Sega Asia has had a YouTube channel for a long while for English subbed videos of the regular Sega Asia YouTube channel videos as well as its own original show in Chinese. (no subs for these though).
To bring us up to parity with TCRF I did a quick runthrough of World Cup Soccer (aka Soccer, Sega Soccer, World Championship Soccer, World Cup Italia 90). The very first version of the game had more real-world players - later versions changed some names, but not others: https://segaretro.org/World_Cup_Soccer/Comparisons#Mega_Drive_version It is widely reported that the players are based on those which competed at the 1986 World Cup in Mexico. And yeah... some of them are. But some aren't. I reckon 10/16 players on the England team appeared at that event, with the other six being retired - this is almost certainly Kevin Keegan, who played his last match in 1985. I think Bobby Charlton's in the lineup too.
It's unfortunate Sega Retro didn't exist in the mid-90s when Mega Drive ROMs were being dumped and organised. Columns/Technical information I mean to be fair it's partly Sega's fault but ehh. There are two versions of Columns on the Mega Drive: a ROM marked April 1990 in the header, and another marked June 1990. They mislead: GM 00054035-00, JUE GM 00054035-01, JUE If you ever wanted to know why we refer to Mega Drive games as "Revision 00" and "Revision 01", it's because of Sega's naming scheme - the last two digits of the product number is meant to symbolise "version": you start at version 1.00, subsequent releases are meant to be 1.01, 1.02, 1.03 etc. There's never a "version 2" so you just get the minor numbers, i.e. 00, 01, 02 etc. (the highest any game ever went was 05 or 06, aka The Revenge of Shinobi) From a code perspective, there are very much two revisions of Columns. The 01 build changes things - it's a revision. However, I'm not sure this is strictly accurate: 00 was only released in Japan, while 01 was only released overseas... so they're actually regional variants, and so are technically separate products (which probably should have different codes?). And as two separate products, they both ought to be marked as revision 00: Japan revision 00, overseas revision 00. Unfortunately they're also not separate products, as despite being released in different places, "JUE" suggests they're region-free. This means even though the later build wasn't intended for Japan, it'll run on a Japanese system, and could theoretically be sold in the region, thus making it an 01 revision. In reality that didn't happen, but the ROM header suggests that it could. But then again, they might not have known that at the time isn't this system fun. How do we know 00 was for Japan and 01 for overseas? Because the Sega logos are hard-coded, and you weren't meant to release products with the registered trademark symbol (®) outside of Japan (presumably because the trademark wasn't registered). Sega's later solution was to drop the ® entirely - 01 would still be valid in Japan (at least after they made that decision), but 00 would never be valid overseas. Anyway as a result of this stupidity ROM databases often list these two dumps as Columns [W] [00] Columns [W] [01] when actually it should be Columns [J] Columns [UE] But you would probably only know this if you ran a Sega wiki and had run into these situations before. And even then I'm not totally sure I've got this right. Guess what else does this? Sonic the Hedgehog! GM 00001009-00, JUE GM 00004049-01, JUE except this time the 00 revision was for overseas, and 01 revision was Japanese only... and they changed the product code this time and areshuikfsdjghnaughbreiksdfvnj I'm currently thinking Sega Retro is "half-right" and the rest of internet is "mostly wrong". Let's see what I think in an hour.
Still wouldn't work in a Japanese system without an adaptor or hardware mod due to the cartridge lock (hence PAL games in Japanese shells for Asia).
Yeah, but still meant they couldn't sell European carts in Asia due to existing MD1 owners, along with MD2 releasing nearly 3 years after Columns.
Are Cinematic Developer and Web Developer good generic terms to use for disambiguating? I need to split Takeshi Itou into Takeshi Itou (cinematic developer) and Takeshi Ito (web developer). The former with a lot more credits (there'll be more credits from his uncovered MARZA future) uses 伊藤 武志, while I randomly found a thread on BBS.CHINAEMU.ORG that lists the Dream Passport developers and includes 伊東 武 (武 can be read a couple ways, including Takeshi, whereas 武志 is usually only read as Takeshi). EDIT: Forgot to mention that 伊東 武 showed up on a couple Dreamcast games before that, woah a triple!
It's as good as any - I'm not sure "cinematic developer" is a real job title but when you're dealing with pre-rendered 3D cinematics I guess you have programmers, artists and animators, which would be confused with those who work directly in games.
It's especially tough with Itou because he's gone through being a director and a producer, and sometimes his credit is as an artist for opening/ending cinematics, and then on Cyber Troopers Virtual-On: Marz his role was just OP & ED Movie, so he has his hands in a lot of aspects of the cinematics.
If you're watching me tackle Mega Drive ROM headers and are wondering how much is left to do, here's some code: Code (Text): {|class="prettytable sortable" !name !something |- {{#dpl: | category=Mega Drive games | category=Old technical information | notcategory=Unlicensed Mega Drive games | notcategory=Mega Drive unlicensed multi-carts | notcategory=Mega Play games | notcategory=Mega-Tech games | notcategory=Mega-CD games | notcategory=Saturn games | notcategory=Mega Drive aftermarket games | include={releases}:md_rrp_uk | ordermethod=title | mode=userformat | secseparators=\n¦''[[%PAGE%]]''¦¦,\n¦- }} |} (obviously take out some of the "notcategory"s if you want the more complicated examples. p.s. you can use DPL to query pretty much anything on the wiki) Basically everything I've tackled so far... I already had lying around already from previous Sega Retro adventures (turns out I've accumulated literally hundreds of ROMs over the years... which is terrifying). These ones I haven't sourced (yet?). But I'm starting to run into issues - some dumps are missing or mislabled on Sega Retro, while we're not sure if others are good or bad. It's almost certainly because we're a few years out of date - very few of the Mega Drive Mini (2) ROMs are accounted for, and assumptions are probably being made about regions which aren't true. There's also some games that haven't been dumped yet. There's not much we can do about those, unless a copy of Osaka Ginkou no Home Banking Service My Line has turned up recently. I'm not on top of these things. If you want to see what we've got, you can check the database.