Converted GX4000 .cpr - The topic (repository link in first post)

[dragon]

Like Hannibal said in the A-Team "I love it when a plan comes together"  . So here is a new version for :

TRIAXOS from Ariola Soft

So many thanks to @Urusergi for patching this and to @dragon for fixing a nasty bug that made the Amstrad version unplayable since 40 years 

Controls:
d-pad = Move the character in the desired direction
J1B1 = Fire / Open inventory (long press)
J1B2 = Pause game (J1B1 to resume game)
Pause button = Quit game

Any direction to quit the inventory, d-pad left/right and J1B1 to select an item (grenade or DHE-IV CHARGES)

Modifications for the GX4000 version:
    - Wait for fire on title screens,
    - Remapped keys,
    - Dynamite game corruption fixed (many thanks to dragon),
    - Pokes : Unlimited laser, shield, time.

Triaxos is one of these games on computer in which the publishers seemed to consider that discovering what to do was part of the game itself. Really frustrating, making these games completely cryptic without any reason. I offer you a new instruction manual in the archive to start more easily what is really a hidden gem, mature and more ambitious than many others 

You can discover an impressive longplay of the spectrum version here:

https://www.youtube.com/watch?v=hFV6Mrc_Ci0&t=2180s

The spectrum share code with the amstrad, they have the same Glitch

[iXien]

Like Hannibal said in the A-Team "I love it when a plan comes together"  . So here is a new version for :

TRIAXOS from Ariola Soft

So many thanks to @Urusergi for patching this and to @dragon for fixing a nasty bug that made the Amstrad version unplayable since 40 years 

Controls:
d-pad = Move the character in the desired direction
J1B1 = Fire / Open inventory (long press)
J1B2 = Pause game (J1B1 to resume game)
Pause button = Quit game

Any direction to quit the inventory, d-pad left/right and J1B1 to select an item (grenade or DHE-IV CHARGES)

Modifications for the GX4000 version:
    - Wait for fire on title screens,
    - Remapped keys,
    - Dynamite game corruption fixed (many thanks to dragon),
    - Pokes : Unlimited laser, shield, time.

Triaxos is one of these games on computer in which the publishers seemed to consider that discovering what to do was part of the game itself. Really frustrating, making these games completely cryptic without any reason. I offer you a new instruction manual in the archive to start more easily what is really a hidden gem, mature and more ambitious than many others 

You can discover an impressive longplay of the spectrum version here:

https://www.youtube.com/watch?v=hFV6Mrc_Ci0&t=2180s

The spectrum share code with the amstrad, they have the same Glitch

That was usually the problem with these "sister-systems" that shared the same CPU. What was true for one, was true for the other   But at least, we have nice mode 0 graphics and not only a "speccy-port". If only all the games had the same treatment.

[iXien]

And here is a new work from @SyX and RedAngel based on a GX4000 game:

FLIPULL+ from Ocean

Flipull is the original japanese title of the game we all know under the name Plotting.

What I can immediatly see in the new version : Name modification of course, lighter colours on the main screen, NES background, more arcadish GFX in general.

@SyX , maybe you can give us more informations about this new one?

You can download it here: https://www.mojontwins.com/cepeceros/descargas/

[SyX]

OK, IXien but you know that this is going to be another big story xDDDD.

Let me remember, because this patch was finished two weeks ago.

At the beginning, I didn't want to make this project, because I don't like the CPC+ version of Plotting. I always had the feeling that this game was finished in a hurry.

And after looking at the code my impression was even worse, the game credits has a person credited exclusively for compression, I can tell you that there is no compression at all in the game.

On the contrary, the game is a continuous waste of rom. The last 16 KBs cartridge page is the "loading screen", things like the upper and lower HUDs in the game are graphics storage in the cart, when they could have drawn the HUDs using tiles. In the same way, the 8 backgrounds graphics are uncompressed in the cart and they take around 40 KBs.

Every piece of code for handling the first player is duplicated with minimal changes for the second player. Even the funnier thing, the first player code is near around the same piece of memory, while the second player code appears in random places around the memory. Everything is signaling that the game was going to be originally for one player only, the same as in c64 or zx, but in the last moment it was decided that the game had to be for two players as Amiga and ST versions.

The first thing the game does, it is to copy the first 32 KBs from the CART to RAM. This is the game engine that is going to be executed from RAM instead of ROM during the game. I have the suspicion that originally this game was designed as the other 8 bits version and it was going to be a CPC game, not a CPC+. But as we speculate about Pang, Ocean forced the developers to convert the game to CPC+ and fill the 128 KBs in the cart.
I said too that we were finishing an interview with an old ex-Ocean that could bring more light about how those CPC+ games were made. The interview with Mr. James Higgins was published last Monday and you can read here (english original, after the spanish translation).

But well, the rest of the game code was terrible, special prize for how they encoded the levels, twice, 32 levels per player and there was a few errors because the copy and paste process; or the animation system for the introduction or the game (when you launch a tile and how return to you); or the abuse of tables instead of using simply math.

You can imagine that after I saw this horror, I wanted to run far far far away xDDDD, but RedAngel didn't stop sending me pictures with their graphical improvements. And they looked great, but I told him that we could not change the playability, because every change had to be duplicated and tested.

Then because I can not look too much at the same thing over and over, I extracted the levels from the cart for being edited in our map editor. And after that I extracted the maps from the arcade version and I discovered that the arcade game has a random component for generating the levels. A feature that lacks the CPC version, although at the end of the day, the cpc levels were similar to the arcade ones (surely they used a longplay in an vhs).

RedAngel continued sending me more and more graphics, and I told him that we could not change too much. And I asked him why he was so interested in this game and he told me that he loved the Nes version, because it was very normal that appears in the nes clones sold in Spain.

At that moment, I took a look at the nes version, and extracted the map for the advanced mode. This mode is 50 levels, and every level is different to the arcade. Nes version has a random component but it is minor compared to the arcade. Then RedAngel adapted the maps from Nes to the style needed for the CPC.

The last detail were the backgrounds, because after I added compression, we got a LOT of free space in the cart for adding backgrounds. We discovered that the backgrounds for the cpc version, even if they are the same as the arcade ones, were converted from the Atari ST version.

Because I am an ex-amiga guy and this game is practically an st port; then in those cases I usually prefer to touch the Amiga version, instead of the Albino Speccy one

Then I disassembled the Amiga version, discovered the background compressor and patched the game in an amiga emulator to get all the backgrounds that RedAngel converted really fast to the cpc.

But now, we had another problem: there were only 32 backgrounds and we now have 50 levels. RedAngel converted the geometric backgrounds from the nes versions, and using 4 colour variants, we got 16 backgrounds more. And because I didn't know if I could patch the game for having an ending; because when you finish the original 32 levels, the original cpc+ game restarts from the first level and with one minute less than before. Then I asked RedAngel to design two backgrounds for the last two levels for being a price for reaching the game ending.

Then in short, the list of improved things are:

* All darker ST palettes (RGB 9 bits) now are using the full CPC+ palette (RGB 12 bits), that means, loading screen, menu, ...

* Change the name from Plotting to the original Japanese name, Flipull. And a + for saying that this is a CPC+ game.

* Loading screen has been crunched and improved graphically, not only the game logo has been changed to Flipull+

* Credits screen has been updated. And now we really have a compressor, in this case zx0

* Menu sequence animation: New big font that is more faithful to the nes version. all the animation sequence has been updated to the new font and new positions where the blob jumps (it has been a big pain in the ass and super boring). Palettes and tiles have been changed to be the same as the graphical changes that have been made for the game.

* FX IA Attract mode, where the player controlled by the computer used an advanced IA for trying to win the play xDDD xDDD xDDD

* Game screen borders and HUDs, everything has been redesigned and repalettized. I sincerely think that the game looks clearer and prettier than before. In the LCD white font used for the time, the digit 1 was wrong, the digit should be in the right, not in the left (a few more silly bugs like this were fixed).

* Every game sprite has been retouched and repaletized. And the colours for player 1 and 2 are different now, giving an extra personality to the blobs.

* 50 new levels taken from the nes version, but adapted to the CPC game motor and retouched by adding extra lives in specific levels, for improving the playability. I fixed the game code to use math instead of tables and now, the real limit is 128 levels, but 50 levels would be enough for everybody. And we started to count in 1, instead of 0, we are humans - no robots

* Instead of 8 awful backgrounds chosen randomly from the ST version, we have 50 backgrounds now. Yes, every level has its own background. First 16 levels are the geometric nes ones, they are perfect for focusing in the game mechanics while you are learning how to play. After that, you get the 32 backgrounds from the arcade/atari st/Amiga. And two bonus backgrounds in the last two levels as a prize for reaching the end.

* Ending: If you finish the last level (50), the game will end instead of going back to the first level.

* A lot of silly bugs were fixed, I don't remember everyone, but there was a lot.

And now the sad things, we were not able to make more changes to the code, it is a terrible big spaguetti plate for two (everything duplicate by no reason than finish fast).

Don't ask about the animation system, seriously, I need a strong drink for forgetting what I saw there; even if we wanted to change the music and I was working in converting the original Plotting ST music to my Sakis music player (Yes @Gryzor, I am so crazy that my music player routine took its name from Sakis Rouvas xDDD); at the end, the silly animation system depends in the music player code and my tries for fixing that hell, made the game break for everyplace. 
I think that is all.

[Gryzor]

Very interesting read until that last paragraph -oh Lord. 😂

[iXien]

OK, IXien but you know that this is going to be another big story xDDDD.

(...)

I think that is all.

Yes, and I love your big stories 

It was very interesting, as usual. I think all the games exclusive to GX4000 have this "spaghetti-code"  You're a warrior of a man diving in this ugly lines of code to upgrade what would have been so nice CPC games if once again developers didn't have to work in a hurry.

Personally, I hate the NES graphics and colours in general, I think it remembers me the C64 too much. So I won't tell you I'm very fond of the 4 coloured backgrounds I will have to play on before retrieving the good ones, arcade based. And to be honest, I loved the original HUD and game screen border. Once again, too much NES-style for me here. No matter, you changed a lot of other things for the best, so many thanks for your work .

PS : Really, no chance of starting to play on the 32 arcade backgrounds and have the NES-style ones AFTER ? 

[SyX]

PS : Really, no chance of starting to play on the 32 arcade backgrounds and have the NES-style ones AFTER ? 

Sure! You only need to play first to the original version and then put Flipull+ xDDDD

You don't need to think about so much gray is to c64, you need to think that CPC+ open a new palette with no limits, hehehe.

Thanks for the good words!

The problem is that you have explained perfectly. And if you have read the Higgins interview, there is a moment where he says that he though that Ocean and Amstrad had a contract for only doing things for CPC+ and left the CPC behind. We don't know how real is that, but the feelings inside Ocean were that: CPC is OUT! Convert everything to CPC+! And Fast!

And the original idea was not bad, using the 8 bits code + 16 bits design and graphics... if only Amstrad and Ocean would have given the chance for a second wave of GX4000 games, maybe we would have been able to see the real potential of the system.

[iXien]

PS : Really, no chance of starting to play on the 32 arcade backgrounds and have the NES-style ones AFTER ? 

Sure! You only need to play first to the original version and then put Flipull+ xDDDD
You don't need to think about so much gray is to c64, you need to think that CPC+ open a new palette with no limits, hehehe.
Thanks for the good words!
The problem is that you have explained perfectly. And if you have read the Higgins interview, there is a moment where he says that he though that Ocean and Amstrad had a contract for only doing things for CPC+ and left the CPC behind. We don't know how real is that, but the feelings inside Ocean were that: CPC is OUT! Convert everything to CPC+! And Fast!
And the original idea was not bad, using the 8 bits code + 16 bits design and graphics... if only Amstrad and Ocean would have given the chance for a second wave of GX4000 games, maybe we would have been able to see the real potential of the system.

But we are able to see the true potential of the machine. It's just that we're discovering it a little late... I mean... more than 30 years late 

[SyX]

But we are able to see the true potential of the machine. It's just that we're discovering it a little late... I mean... more than 30 years late 

I am not so sure, I am waiting until @norecess464 launches Sonic for seeing a real GX4000 game 

Funny thing about the cold reception of Flipull+, specially in general retro news sites and youtube channels, it looks that the people thinks that it is only a minor hack and nobody understand the graphical look that RedAngel brings to the game, as I said I prefer this clean look to the original one (for me all those tiles are too much noise for my eyes).

But the detail that nobody is getting, and it's logical because they are not playing, they are only looking pictures or videos; it is that now the game is less arcade and more puzzle. You can create strategies for finishing every screen, everything is less random. And that, plus the minimalist look, makes possible than now I can enjoy the game (even if the music is annoying and I worked hard to change it).

[iXien]

But we are able to see the true potential of the machine. It's just that we're discovering it a little late... I mean... more than 30 years late 

I am not so sure, I am waiting until @norecess464 launches Sonic for seeing a real GX4000 game 

Funny thing about the cold reception of Flipull+, specially in general retro news sites and youtube channels, it looks that the people thinks that it is only a minor hack and nobody understand the graphical look that RedAngel brings to the game, as I said I prefer this clean look to the original one (for me all those tiles are too much noise for my eyes).

But the detail that nobody is getting, and it's logical because they are not playing, they are only looking pictures or videos; it is that now the game is less arcade and more puzzle. You can create strategies for finishing every screen, everything is less random. And that, plus the minimalist look, makes possible than now I can enjoy the game (even if the music is annoying and I worked hard to change it).

Honestly, I think the NES backgrounds can be blamed a little. Nowadays, when an old arcade game conversion is upgraded, it's usually to make it nearer to the arcade experience, at first visually. Here, you prefered to take the NES look of the game. You start to play and you discover that the arcade backgrounds disappeared, replaced by monochrome ones (and less detailed, due to the mode 0 constraincts). Strange for people that didn't really played the game and can consider these simplier graphics as a regression.
That's why I'm so puzzling why you didn't decide to prefer starting the game with nice backgrounds based on the Amiga version and closer to the arcade, making the NES backgrounds as a bonus remembering a nice conversion. You can be sure people would have told "whaooo, nice improvement to the original". Your upgrades are very nice but people judge a game to their first contact with him, the reason why first levels of lot of arcade games are more nice looking than their second half.

More seriously, you already got it. People looking at this new version doesn't really play it, while the biggest improvements you made are in the game mechanics. It's also something difficult to show in videos.

And simply, don't forget that Plotting/Flipull is a games far less famous than Pang. You can't expect as interest in this nice but very little puzzle game 

[overange]

@SyX that was such an interesting read on how you achieved the redesign and as it was mentioned the GX4000 was such a good system and only now, what it can do is being seen.
I really do hope someday someone can reproduce the ASIC and somehow either integrate a variant of this into an original CPC, or even make a modern GX4000 system Clone.

With the work that was done with plotting/Flipull+ could it be possible to re-engineer Gazza into the never released Plus Version based on the limited screen shoots?
Or even port ChaseHQ 2 from the Speccy and recreate a new version of what could of been?

[GUNHED]

Syx, what a nightmare! You be proud that you finished it, because pretty much anybody else would have given up on that thing. However the result is very charming!!! 

[SyX]

Feel free of making any comment @iXien, I respect your opinion and I learn with every feedback.

The reason for our choice, it was because is not an arcade game, it is a console game. And the nes backgrounds are only in the first levels for putting your focus in learning the gameplay. Then after that, every new screen that you reach will be a prize, hehehe.

With the work that was done with plotting/Flipull+ could it be possible to re-engineer Gazza into the never released Plus Version based on the limited screen shoots?

That is not possible, at least until I can finish the ArnoldGPT. Until then you need to develop a full game, hehehe.

Or even port ChaseHQ 2 from the Speccy and recreate a new version of what could of been?

Well, the game is awful in speccy and the things that we know about the gx4000 version is that the game is terrible in every possible way... the developers are the same of Out Run, Turbo Out Run, Cisco Heat, ... then nothing fun can come from there.

I can show a picture about how this game would look before getting the order of being converted to plus (no IA used here):

[iXien]

@SyX , a little bit of topic, but the more I play this FLIPULL+ version, the more I'm asking myself if it would be possible to convert the game on CPC? With these new GFX, you give a CPC touch to the game, more than a "Plus" touch. But I suppose the "spaghetti-code" would be a real pain in this exercise?

[SyX]

@SyX , a little bit of topic, but the more I play this FLIPULL+ version, the more I'm asking myself if it would be possible to convert the game on CPC? With these new GFX, you give a CPC touch to the game, more than a "Plus" touch. But I suppose the "spaghetti-code" would be a real pain in this exercise?

It looks like I have lost the ability to explain something in a few words, sorry xDDDD

After 150 pages, a little off-topic is going to disappear, then we can indulge in it for a few seconds.

As I said before, I didn't want to touch this game at all, because I felt that the GX4000 game is not a good game. And I only saw minimal graphical improvements and that is all. 

But RedAngel saw something that I didn't, then I took a better look at it and I felt that this game could be converted to the classic CPCs without too many problems. And it is perfect, for everything that I have been talking about the first generation of gx4000 games: they were developed in a hurry and they started as classic cpc games.

Sure, after looking at that awful code, I thought of a few alternatives:
1.- I could port the zx version, all the logic is there and because it is not an advanced game, all the graphics could be improved to use mode 0. I even could use overscan or egx if I was crazy enough.

2.- Because the game logic is not complex and when I was a child and I saw the arcade, I made a few tests in basic using the cpc font as graphics. Well, 30 years later, sure that I would do that in assembly or even c in a few evenings. There are already two CPC games inspired in plotting; Inferno, a 1994 public domain game, it looks basic but it is in overscan; and Hot Plot, a type-in from a German magazine.

But after the first scary look, at the end of the day I don't need to touch that spaguetti code too much. In fact, Flipull+ is only redirecting a few of the bugged routines to my patched ones or they are patched directly.

The problem or blessing is that RedAngel had already revamped the graphics, adapted the nes levels and converted a lot of the new backgrounds. Starting a CPC version would mean remade everything again. And if I wanted to use all the backgrounds, I should load new backgrounds after a few levels or up the specification to 128 KBs machines or disk drives (neither of those things are bad, in fact in my mouse patches if I can, I use 128 KBs and floppy drive for hiscores, because if you have a mouse expansion, then you should have more expansions surely).

Then a CPC version is more than possible, the original cart organization is:
* Page 0: Game engine that it will go in $0000 and a few fonts (uncrunched)

* Page 1: Game engine that it will go in $8000 (includes bytes at 0 that are used as ram variables or working space when is copied to ram), game tiles (uncrunched)

* Page 2: Map, Fonts, logos, huds (everything uncrunched) and the menu code.

* Page 3: Sprites, fonts, other graphics (everything uncrunched)

* Page 4: Backgrounds 1-3

* Page 5: Backgrounds 4-6

* Page 6: Music player, song, ... and the last two backgrounds

* Page 7: Uncompressed loading screen 

With that information, we can consider that for making a CPC 464 version, we should start by converting the loading screen into a real loading screen (we load and forget!), cart page 7 is not necessary anymore. Then we can put our video memory in $C000.

We are going to skip the backgrounds for now, that means cart pages 4, 5 and half of page 6 are not necessary anymore.

I am lazy, the menu sequence is going to be shown once; the first time, after the game is loaded, after that it will go out of the RAM. 

Of course, I will not store the HUDs uncrunched, they would be drawn using tiles. And we don't need two copies of every tile; one when they are tiles and a second time when they are cpc+ hardware sprites. Meaning that I can put everything graphic related and the levels map, everything that lives in cart pages 2 and 3, now they will live in $4000. Surely, I can put the arkos player and the new songs in $4000 too.

Cart page 0 goes to $0000 and page 1 goes to $8000. The code changes that I need to do are: replace the cart pagination code, we can NOP those bytes, and CALL directly to our routines that are already in memory. And replace the routine that copies sprites to the sprites in the asic ram, for our sprite routines.

Now for extra credits, we can look for empty space in RAM for putting all the compressed backgrounds that we can get, and that is, you have Flipull for CPC.

... although if we have been making all this work, why should we be using the spaghetti code? The code necessary for handling this game is not hard. In fact it is a wonderful project for somebody that wants a funny project for a few months.
.
.
.

@iXien, I hope that you are starting to enjoy the game, seriously, the more you play it, the better the game is, hehehe.

And you can return to the topic now!

[iXien]

To start the weekend, we thought this one was missing. Many thanks to @Urusergi , as usual  :

MACH 3 from Loriciels

Controls:
J1B1 = Fire
J1B2+UP = Normal controls
J1B2+DOWN = Reverse controls
Pause button = Pause game (any button to resume)

Modifications for the GX4000 version:
    - Wait for fire on title screen
    - Patch keys
    - Bugfix, the game no more crash crossing an arch
    - Poke: Unlimited lives

This conversion is based on the international release of the game, with music on a main screen absent from the French version