Going great guns on a Prince of Persia port...

[daveejhitchins]

I am SO excited for this. Can we support this in some way? A Patreon, or some such thing? The amount of effort you all are putting in is amazing!

Seconded, this Game is going to set-some-standards! . . . With Kieran working on this and Sarah working on her "tube type thingy" and White Light under her belt already . . . Wow

Dave H

[oss003]

Congratulations guys, the gameshots are looking fantastic.
I know it's a big and long proces to check all game assembler and game data but checking 128KB is a huge, huge job........ respect ......
Keep up the good work by posting some screen shots now and then, makes waiting easier ....

Greetings
Kees

[Lardo Boffin]

I am SO excited for this. Can we support this in some way? A Patreon, or some such thing? The amount of effort you all are putting in is amazing!

Seconded, this Game is going to set-some-standards! . . . With Kieran working on this and Sarah working on her "tube type thingy" and White Light under her belt already . . . Wow

Dave H

Thirded.

This was my main motivation in getting a Master...

[kieranhj]

I am SO excited for this. Can we support this in some way? A Patreon, or some such thing? The amount of effort you all are putting in is amazing!

Thanks for the feedback & encouragement, that is all the support I need for the project (plus using jsbeeb to distribute test builds of course. )

The usage rights for the POP source code were made quite clear by Jordan when he posted them to GitHub: https://github.com/jmechner/Prince-of-Persia-Apple-II

As the author and copyright holder of this source code, I personally have no problem with anyone studying it, modifying it, attempting to run it, etc. Please understand that this does NOT constitute a grant of rights of any kind in Prince of Persia, which is an ongoing Ubisoft game franchise. Ubisoft alone has the right to make and distribute Prince of Persia games.

As we've said, this is a labour of love for those involved and we do it because it's a challenge, it's fun and for the love of the BBC! No Patreon required.

That said, whilst three of the team are gainfully employed elsewhere, our musician Inverse Phase does make a living creating chiptune music (a niche gig by any measure) so any time he spends working on POP is time not spent on paid work. I would therefore encourage everyone to visit his Patreon page and consider supporting him so we may continue to have new, awesome BBC tunes in the future https://www.patreon.com/inversephase/overview. (I know you're already a supporter Matt!)

[Matt Godbolt]

Cool! Thanks Kieran. We'll keep up our side of the agreement by continuing to give our enthusiastic support!

[RobC]

Wow - that looks amazing. Congrats to all involved

I'm sure it'll be well worth the wait

[trixster]

It looks absolutely amazing! More screenshots please!

[richardtoohey]

We'll keep up our side of the agreement by continuing to give our enthusiastic support!

Hear, hear! It's looking great, but don't kill yourselves with delivery deadlines/stress; it will be ready when it's ready!

[Dave Footitt]

Aye, top work, looking really lovely!

I've just had my Master Compact refurbed and this will be the perfect excuse to get it going

[Dethmunk]

FYI It's close my Acorn loving friends....
Note: This is just a bit of extra artwork I put together for Kieran. (Subject to changes). Up to him what he does with it mind you. He seemed to like it. I was just doing a test print of it.

[richardtoohey]

No, no, you've got no time for that, get on with the game!



Just kidding, it looks great.

Just read Jordan's journal from that time (when he created Prince of Persia); interesting read (but very very light on any technical detail.)

[FourthStone]

This PDF is a good read of some of the technical bits.

http://www.jordanmechner.com/downloads/popsource.pdf

[kieranhj]

Almost a year after first looking at the code here we go: http://www.stardot.org.uk/forums/viewto ... 57&t=14858.

[kieranhj]

I did promise everyone a memory map of the game. RobC has prompted me to do this so here's a first attempt:

Code: Select all

-------- 
POP-BEEB MEMORY MAP
-------- 

-------- 
ZERO PAGE
-------- 
&0000 - &00FF: ZERO PAGE
-------- 
Zero page high watermark =  &A0 
Zero page free =  &0 
--------
Global variables &00 - &9F
Local variables &D0 - &DF
-------- 

-------- 
LOWER Workspace 
-------- 
&0300 - &0DFF: "Lower" RAM aka "Lang" (legacy name)
-------- 
Lower workspace high watermark =  &CF8 
Lower workspace RAM free =  &8 
-------- 
General run-time buffers
E.g. image lists, collision buffers, moveable object tables, global variables (not ZP)
--------
&0C00 - &0CFF: Lower data loaded from file "$.Lower" at boot
Contains filenames, sfx lookup table, crash message text, version number at &CF0
-------- 

-------- 
CORE Modules 
-------- 
&0E00 - &2FFF: "Core" RAM
-------- 
BEEB BOOT size =  &1EC              ; System and game bootstrap
AUX JUMP TABLES size =  &81E        ; Jump tables for AUX (gameplay) functions
DISKSYS size =  &187                ; Disk system (file + sector load)
SWR size =  &6                      ; Sideways RAM routines
BEEB TEST size =  &0 
BEEB PLATFORM size =  &129          ; ULA palette, sprite palette routines
MASTER size =  &446                 ; attract mode, load data for levels or cutscenes
TOPCTRL size =  &77E                ; main game loop
AUDIO (LEGACY) size =  &1 
PUCRUNCH size =  &15B               ; decompress static screens etc. 
BEEB AUDIO size =  &106             ; Beeb audio module handling
HIRES size =  &1DC                  ; original Apple II screen plotting routines
BEEB PLOT FONT size =  &F6          ; font plot routines
BEEB PLOT FASTLAY size =  &13C      ; fast sprite plot routines
BEEB PLOT LAYRSAVE size =  &167     ; unrolled sprite "layrsave" routines
-------- 
Core code size =  &20E1 
Core data size =  &40 
Core high watermark =  &2F21 
Core RAM free =  &DF 
-------- 
Code entry is at &E00
BEEB BOOT section is trashed after boot completes and used as the Exomizer decompression buffer (for music player.)
Aux jump table transparently redirects function calls to appropriate Aux (gameplay) module in sideways RAM bank, preserving registers and ROM slot on entry and exit.
Fastlay sprite routines are byte aligned and have no clipping - used for background objects
Layrsave routines save the screen buffer beneath the character sprites into the "peel" buffer memory area
-------- 

-------- 
MAIN Modules 
-------- 
&3000 - &7FFF: "Main" RAM (also shadowed x2)
-------- 
BEEB PLOT size =  &360              ; character sprite plot routines
BEEB PLOT WIPE size =  &64          ; unrolled sprite wipe routines
BEEB PLOT LAY size =  &0 
BEEB PLOT PEEL size =  &125         ; unrolled sprite "peel" routines
-------- 
Main code size =  &4E9 
Main high watermark =  &34E9 
Peel buffer address = &3500
Peel buffer size = &C00
Screen buffer address =  &4100 
Screen buffer size =  &3F00 
Main RAM free =  &17 
--------
Character sprite plotting routines support per-pixel plotting with mirroring and per-byte clipping on all four sides
Sprite peel routines "peel" sprites pff the screen by writing previously saved ("layrsaved") data back
Screen is MODE 2 but at 160x200 resolution (24 character rows for game and 1 character row for the status line)
Entire Main RAM area is double-buffered using SHADOW RAM
--------

--------
ANDY RAM
-------- 
&8000 - &8FFF: "Andy" RAM
-------- 
AUDIO BANK 0 size =  &D6B           ; title screen music
AUDIO BANK 0 free =  &295 
AUDIO BANK 3 size =  &F14           ; in-game musical ditties
AUDIO BANK 3 free =  &EC 
AUDIO BANK 4 size =  &B09           ; in-game cutscene music
AUDIO BANK 4 free =  &4F7 
-------- 
Audio banks are overlaid as required by master.asm module
-------- 

-------- 
SIDEWAYS RAM BANK 0 
-------- 
&8000 - &BFFF: "Bank 0" (ROM slot 4)
-------- 
BGTABLE 1A size =  &171D            ; background sprites set 1A (Palace or Dungeon)
BGTABLE 2 size =  &1223             ; background sprites set 2
CHTABLE 4 size =  &14C1             ; enemy character sprites (Guards)
-------- 
BANK 0 size =  &3FC1 
BANK 0 free =  &3F 
-------- 

-------- 
SIDEWAYS RAM BANK 1
-------- 
&8000 - &BFFF: "Bank 1" (ROM slot 5)
-------- 
BGTABLE 1B size =  &7D8             ; background sprites set 1B
CHTABLE 1 size =  &129E             ; player sprites set 1
CHTABLE 2 size =  &1283             ; player sprites set 2
CHTABLE 3 size =  &E1A              ; player sprites set 3
AUDIO SFX size =  &3A2              ; sound effects
-------- 
BANK 1 size =  &3FA2 
BANK 1 free =  &5E 
-------- 

-------- 
SIDEWAYS RAM BANK 2
-------- 
&8000 - &BFFF: "Bank 2" (ROM slot 6) aka "AUX B" gameplay modules
-------- 
SEQTABLE size =  &9F1               ; animation sequence data
FRAMEDEFS size =  &7B7              ; animation frame data
CTRLSUBS size =  &86F               ; sub-routines for controling characters
COLL size =  &736                   ; collision detection routines
AUTO size =  &A4E                   ; AI / enemy character auto control
INVERTY size =  &DF                 ; code to invert all plot routines in Y axis
BEEB MASTER size =  &4B             ; Beeb specific functions for attract sequence
-------- 
Aux B code+data size =  &2CC5 
Aux B high watermark =  &BFC5 
-------- 
BANK 2 size =  &3FC5 
BANK 2 free =  &3B 
-------- 

-------- 
SIDEWAYS RAM BANK 3
-------- 
&8000 - &BFFF: "Bank 3" (ROM slot 7) aka "AUX High" gameplay modules
-------- 
CTRL size =  &B4C                   ; character control routines
FRAMEADV size =  &B41               ; render a single gameplay frame into image lists
GAMEBG size =  &507                 ; functions to plot specific game objects
GRAFIX size =  &40E                 ; plot sprites from image lists onto screen
SUBS size =  &868                   ; general sub-routines including cutscene
MOVER size =  &A60                  ; moveable object routines
MISC size =  &3EE                   ; misc routines including energy meters
SPECIALK size =  &5E2               ; keyboard handling
BEEB SCREEN size =  &89             ; Beeb specific screen functions
ATTRACT size =  &20F                ; entire attract sequence
-------- 
Aux High code size =  &3ED2 
Aux High high watermark =  &BED2 
-------- 
BANK 3 size =  &3ED2 
BANK 3 free =  &12E 
-------- 

-------- 
HAZEL Modules 
-------- 
&C300 - &DEFF: "Hazel" RAM
-------- 
BLUEPRINT size = &900               ; level data loaded from disc
TABLES size =  &57C                 ; multiplication tables
BGDATA size =  &1EB                 ; data tables related to background sprite pieces
HRTABLES size =  &190               ; lookup tables for sprite plot
BEEB PALETTE DATA size =  &4C1      ; unrolled palette tables for fast sprite plot
EXO size =  &134                    ; Exomizer decompressor
VGMPLAYER size =  &132              ; VGM music player module
FONT size =  &1CA                   ; font data
-------- 
HAZEL data size =  &F00 
HAZEL code size =  &2FC 
HAZEL BSS (blueprint) size =  &900 
HAZEL high watermark =  &DEFC 
HAZEL RAM free =  &4 
-------- 

-------- 
CUTSCENE OVERLAYS
--------
CHTABLE 6 size =  &3E0A             ; cutscene sprites (4bpp) level 2+
CHTABLE 7 size =  &8EB              ; cutscene sprites (4bpp) attract only
CHTABLE 8 size =  &1B9E             ; cutscene sprites (4bpp) attract + level 1-2
CHTABLE 9 size =  &2B0F             ; cutscene sprites (4bpp) attract only
AUDIO BANK 1 size =  &128C          ; story cutscene music
AUDIO BANK 2 size =  &2B48          ; epilog cutscene music
--------
CHTABLE 6 or 8 are loaded into BANK 0 (ROM slot 4) between gameplay levels, displacing background sprite sets
CHTABLE 9 is loaded into BANK 1 (ROM slot 5) during attract cutscene only, diplacing background and player sprite sets
CHTABLE 7 and AUDIO BANK 1 are loaded into BANK 0 (ROM slot 4) during attract sequence only
AUDIO BANK 2 is loaded into BANK 0 (ROM slot 4) before the final epilog cutscene
--------

Also attached as a text file for convenience. I will check this in to the GitHub repo when I get chance.

[RobC]

Thanks Kieran - that's awesome.

[polluks]

Hi!
Is a Electron with 128Kb SWR also supported?

Greetings,
Stefan

[kieranhj]

Hi!
Is a Electron with 128Kb SWR also supported?

Greetings,
Stefan

Hi Stefan - unfortunately not. This is for the BBC Master only as it relies on the additional memory and features available on that machine outside of just the SWRAM space. I hope you can enjoy it on an emulator if you don't have a real machine.

[tricky]

Has anyone got this working on a compact?
I failed with dfs and a gotek as it stops after the initial intro.

[kieranhj]

Has anyone got this working on a compact?
I failed with dfs and a gotek as it stops after the initial intro.

Sorry, it's a known issue or perhaps one of several with MOS 3.5 / Compact. I never got round to finding / fixing it last year as was a bit burnt out on POP by the time it was released. I'll take a look when I get chance.

[BobuS]

I am wondering: Since BBC B+128 also has 4x16Kb of Sideways RAM and 20Kb of Shadow RAM (which makes him "almost Master") is it possible to port PoP on BBC B+128? From my point of view, that is almost the same amount and same organisation of Memory.
Or am I wrong?

[danielj]

Someone with more detailed knowledge will probably jump in, but I think PoP relies on the fact that in the Master you can write to the shadow screen memory even if it's not being displayed, whereas on the B+ you can only write to it if it's selected for display. That's the main difference between them with the shadow memory implementation.

[cmorley]

You'd have to bin the double buffering on the screen which would make it flicker. You'd also have to remove all the 65C12 instructions Kieran will have used on the Master like PHX,PLX,STZ. As this would bloat code it might be a tricky job as you'd likely get branch targets exceeding the range and a whole host of knock ons as code spills. No Hazel on a B+ so you'd have to magic up something there

Not impossible but it would need a B+ port rather than a simple tweek.

[kieranhj]

Yes, for the reasons mentioned above it wouldn't be possible to make POP run as-is on a B+ model. The game absolutely requires double-buffering for the graphics (this is the way that the original Apple II drawing code is written) so the ability to display SHADOW RAM whilst displaying main RAM in MODE 2 is essential, and this is only possible on the Master.

On reflection, it might be theoretically possible to create a Model B(+) version (requiring at least 64K SWRAM, maybe one bank more) but this would have to be in MODE 5 or 4, either of which would be disappointingly less impressive that the original Apple II version (which runs at approximately MODE 1 resolution, albeit with artifact colour.) As most of the graphics were already stored 2bpp compressed, in theory this would "just" be a rewrite of the sprite and tile plot routines for MODE 5, but often with such endeavours, once you start pulling on a thread there's bound to be a lot of gnarly things attached to the other end of it. It would undoubtedly have an extremely high frame rate mind!

The source code is all on GitHub should anyone feel sufficiently motivated to give it a crack. I warn you there is lots of it though (40,000+ lines...)

[tricky]

Has anyone got this working on a compact?
I failed with dfs and a gotek as it stops after the initial intro.

Sorry, it's a known issue or perhaps one of several with MOS 3.5 / Compact. I never got round to finding / fixing it last year as was a bit burnt out on POP by the time it was released. I'll take a look when I get chance.

The Compact compatible version is in this post viewtopic.php?p=278962#p278962 as a .SSD

[tricky]

Did a nula version ever happen, or even a replacement palette?

[bob147]

Wasn't there something funky about the engine that made nula non trivial? I'm not aware of any existing nula version at least.