Jump to content

Norman Sword

Member Since 08 Mar 2017
Offline Last Active May 19 2019 01:45 PM

#11038 [File] Manic Miner By S.P. And Bug Byte

Posted by Norman Sword on 17 May 2019 - 02:51 PM

File Name: Manic Miner By S.P. And Bug Byte

File Submitter: Norman Sword

File Submitted: 17 May 2019

File Category: Manic Miner [Remakes]

System: N/A
Third Party Author(s): Norman Sword

When Matthew wrote manic Miner he released the game originally with Bug Byte. When he re-released the game with Software projects some small changes were made to the graphics.


This version is not a new version of the game and its main novelty is the ability to swap between the versions of the game.


Whilst I made minimal change to the games playing style or data I changed some of the ways that the game plays.
This is not a simple edit of the game it is a rewrite but leaves the basic data intact.


I have added :-
item collection sounds.
sound effects for being hit by the solar ray
sprite collision sound effects
heavy landing sound effects.
death collision graphic and sound effect
an extensive cheat.
an explanation to why you have died
an animation of the high score being added
a redraw of the title screen (fixing the keyboard graphics)
a title screen animation of a spectrum analyser
a credits screen
a title screen animation showing the version about to be played (Bug Byte graphic or S.P. graphic)
a sprite mask on the main character
a redesign on how the screens update (to stop sprite glitching)
the list of changes is extensive (Some listed above) the overall feel will still be that you are playing the original Manic Miner.


This file is listed elsewhere on this site, Having been asked by several people where this file is listed, I realise it needs to be placed in the Manic Miner Download section so it can be found easier.

Click here to download this file

#11034 JSW Main Sprite possible improvement?

Posted by Norman Sword on 14 May 2019 - 07:28 PM

 "jagged finger fix by Norman Sword"

 The fix used to reduce the jagged finger look in the file mention in Post#7 is basically a copy of the code I placed on this website.
 So it comes as a surprise that someone else has fixed this problem. Have they really? It seems to me that they have used my code.

 This site mentions so often things like Goeff mode etc that this partial fix to this problem  should be call "jagged finger fix by Norman Sword".
 It was a problem that  prior to my posts, was not fully understood in it's nature or how to remove. The code used to partially cure this problem
 is basically the code I wrote.

 If you have a problem with what I say here then please show me any code prior to the code I listed that did anything of the same nature.
 Or any version of jsw(1) that fixed the problem prior to my posts.


#11026 JSW Main Sprite possible improvement?

Posted by Norman Sword on 13 May 2019 - 09:12 PM

The choppy description is basically what I have described as the jagged finger effect, and is nothing to do with television / monitor effects. Otherwise every version would exhibit the same effect on every game format. Since they don't it is evident the effect is game produced on the original spectrum version.

Play any of my versions jsw128 VK or jsw128 VL and the problem does not exist.

  • IRF likes this

#10826 Restart without recollection

Posted by Norman Sword on 12 April 2019 - 07:35 AM

Slight edit. Due to the original wording, which refers to five bytes. A figure I did not question and repeated ad verbatim.

However the figure should be six bytes. So an edit to correct and also change/delete some words which are misleading.


The code is monitoring 32884 (#8074) for the release of angry Eugene. This can be changed to monitor one of the objects directly,  and the best choice is the object closest to the portal. Since the order of collection is well known, this is probably not going to be noticed by the normal player. This suggested change is aimed more at a new room design and pushing towards your original goal of resetting the room in Manic Miner, but not the objects. 

You can change both the references to 32884(#8074) to one of the following, or even change either reference to one of the following. (mix and match)

Object 1 stored at 32885 (#8075)   position (1,30)
Object 2 stored at 32890                position (6,10)

Object 3 stored at 32895                position (7,29)

Object 4 stored at 32900                position (12,7)

Object 5 stored at 32905                position (12,9)

The logical change is to point at 32905 which is the object closest to the portal.

The reason for the suggested change is that it would indicate which object triggers the anger of Eugene and collecting that object, would make Eugene change colour and descend. Collecting the trigger object before the other objects are collected would send a powerful message that that object must be left till last.

This suggested change was so a room reset, would only need to reset the last object and the room would be re playable. This suggested change unfortunately does not achieve that goal. The player could ignore Eugene and carry on collecting objects. Then the last object collected would no longer be the trigger object and we are back to square one.



Still might be interesting to change the logic so it monitors just one object and not wait till all objects collected. 

#10821 Restart without recollection

Posted by Norman Sword on 10 April 2019 - 02:00 PM

Modifying Angry Eugene

Minimum needed to modify both routines is 2 bytes


------------------ The first one byte modification ---------------------------------

The first six byte routine, only needs the last JR Z,XX  offset changing to 0  

S.P. Version: The six bytes in question are 36355 to 36360

36355 LD A , (32884)
36358 OR A
36359 JR Z , 36378 >>>> Poke 36360,0  ---- This changes the code to JR Z,36361

B.B. Version: The six bytes are at 36344 to 36349

36344 LD A , (32884)
36347 OR A
36348 JR Z , 36367 >>>> Poke 36349,0 ---- This changes the code to JR Z,36350


This above just changes how far the relative branch has to jump. 

-------------  The second one byte modification --------------------------

The second modification is as you stated (slightly edited)

Change the byte at 36449 from 32 to 24, this changes the instruction at 36449 from

      JR NZ,XXX              (Jump Relative Non Zero,offset)
to   JR  XXX                  (Jump Relative, offset)

Original:    36449    JR NZ , 36458
Changed:  36449    JR 36458

B.B. Version: Again the same thing just in a slightly different place change the byte at 36438 from 32 to 24 to effect a change from a JR NZ to a JR:

Original:   36438    JR NZ , 36447
Changed: 36438    JR 36447


Total modification 2 bytes


#10670 Free space and code optimisation in "JSW"

Posted by Norman Sword on 04 March 2019 - 12:41 AM

The way the logo is stored is poor design. It is possible (i have done so) to reduce the size of the title screen and its data by around 200+ bytes, it might be nearer 300 bytes.


#10615 Cheat Table in CPC amstrad version of JSW2

Posted by Norman Sword on 25 January 2019 - 07:48 AM

Amsrad CPC Jet Set Willy.

Cheat invocation. Any room and stood on the lower floor. This will be reset by any key that is not part of the cheat code.

I have invoked this several times recently. All done in the bathroom. The easiest way to invoke. (but it will freeze the game) is to make the assumption that it could be either of the cheat codes I have mentioned. 

press <space> this will clear the cheat buffer


then type in the longer of the two cheat codes-   <HIEMMRAIDNAPRRRTT>


From this point onward, game abort, will take you to the cartography room. This was designed as a cheat, To actually abort the game you need to loose all your lives. Break while playing the game will draw the cartography room.

A big if here. If the cheat mode worked, the cursor on the screen can be moved around the map. When over a room on the map that is the target/designated room that it is desired to go to, press <enter>


The cartography room will disappear and the selected room will be displayed.


The cursor will again be flashing and again it can be moved. The cursor this time is indicating the position Willy will be drawn in the room. Press <ENTER> and the game will resume with Willy at the designated position.

The above is what is supposed to happen.  (after invocation of the cartography room and its cheat)

The first problem concerns the inaction of the keyboard to input. This could be just configuration problems. But since I could not on my brief test get the cursor in the cartography room to move, the game was locked onto that screen. (so progress to the next part did not happen) 

The next problem, one I did not get as far as testing. Is that table of corrupt data. The code in the cartography room uses the data to find what room you are selecting as you move around the map. Very simple code. The little squares on the map are simply counted off (the squares indicating rooms) by starting in the top left hand corner and scanning each line to find a room square. If it finds a square it then steps one byte through the cheat table. If the cursor is flashing over the square picked and <enter> is pressed, it takes the value in the cheat table, and uses that as the new room number. Otherwise it just keeps scanning through the cartography room and stepping each time it finds a marked square indicating a mapped room. When it reaches the end of the screen, it knows you did not select a room, and loops through the movement code again.

The cheat table is a list of the rooms as defined in the cartography room. Without that table, the data is random. What happens is the room is selected and the selected room should be drawn. The code looks up the room number and (in most cases) draws a pile of junk on the screen. Which indicates to me that the room number extracted from the cheat table is higher than the room count. So in most cases the cheat table must have values over 140ish , which are room numbers outside of the game data.

At other times if it finds data that is within the room number total it draws the room that corresponds to that number. Since this is basically a random room number. We get shown a random room in the game. This mismatch between the map and the room shown, makes the cheat worthless. 

The problems I am encountering now, seem to be greater than the problems encountered five of six years ago. Which is why I assume that the inability to move in the cartography room is a configuration problem only.



Its not Escape that quits to the title screen! :D , will retest with various other keys although the fact he says "another key or two" is going to make it more difficult. Praise for snapshots for quick reloading!


As an aside there is talk of memory constraints, although it would potentially only save a 'few bytes' both the AMS and ZX versions with keypad intact have a blank page with a 'Press enter to continue' line, rather than like say other games with keypad protection (ZX at least) that immediately take you to the entry page. All I'm saying here is a few precious bytes *could* of been saved by not having this slightly un-needed page/pause before code entry ? :)


The only sensible next test I can think of is to try it on my real CPC464 although I'd need to sort out the tape image (tzx/cdt to mp3/wav or something) and use that rather than use my original but re-release tape, mind you that would be quicker. This way I can rule out key mapping concerns, as its not ESC it may be CLR but that does not  work either.


 As stated above it is indeed the break key, that invokes the cheat jump into the cartography room.

One of the many flaws of Matthew Smith's game  (spectrum) is the way it asks for a colour code. If you run the game in an emulator you will find that it will ask for the same code on every single game that is played. It will even ask for the same code on the second attempt. Every time the game is played. This is due to usage of one of the spectrum's variables as its random number seed. Since this is fixed on every load of the game when using an emulator, that is why you get asked the same input code. This problem is not a new problem, it existed back in the 80's. You can save a snapshot of the game. If the snapshot saves the system variable, then the game will ask for the same code every time it is loaded via a snapshot.

The simplest method of stopping that simple problem is to force the user to have some interaction with the game before asking for the input code. So to update the variable it would have been better if the simple question/statement had been presented to the game player. Which is <Press enter to continue> or something similar. To stop this question being negated by pressing enter as the game is loading,. Which would immediately upon completion of loading  bypass the question, because the <enter> key is being pressed. You set up the loop to wait for its release as well. The wait for key and wait for key release will generate a better random number, and a snapshot will still ask for a new colour code on every run. Even the modern versions run on an emulator would give a new number on each run of the game...

The above is why the Amstrad version asks for the user to press a key. All the time the question is on the screen the games code is changing the value of the code to ask. If the end user has enough skill to bypass this code  (asking for enter to be pressed), you can assume that he is or she is already delving into the code. 


Searching for the cheat code activation text. Derrick thinks that the code might be written backwards, and even have a simple addition of setting bit 7 on the letters. Any decent debug/monitor allows the toggling of ASCII in its output display. The setting of bit 7 on ASCII is a common practice, and the debug/monitor should  permit you to see the ASCII from values over 128 by toggling the display output.  



The hacking of the game for infinite lives will also make it impossible to exit the game once cheat is invoked.
(if it worked)


#10606 Cheat Table in CPC amstrad version of JSW2

Posted by Norman Sword on 23 January 2019 - 11:29 PM

The whole scope of this topic is the missing Cheat code table on the CPC Amstrad.


Information direct from Derrick.

The conversion from CPC Amstrad back to the spectrum, involved a game redesign. The CPC has an interrupt set for every 300th of a second. These are arranged to occur on frame flyback and five others per frame. This enables the screen to be recoloured on every screen frame. The top part of the screen has one set of four colours, and the lower part of the frame has another set of four colours. The interrupt is supposed to occur as the raster passes over the screen in the area of the room name. So the room name area has two colours that are common to both the upper part of the screen and the lower part of the screen. This enables the game to have six colours in some rooms. (in 4 colour mode)

The game uses xor on and xor off. One of the advantages of this method is that it is fast, another advantage is that the Amstrad game does not then need to have any copy screens.  This means that the sprite is erased on the visible screen and for a moment it is possible to see the difference between the new and old sprite. Think about the spectrum and the way It was pointed out the jagged finger effect. The visual affect of xor ing off a sprite and xor ing on a new sprite makes that effect look tame. You end up with a very visible jerking or tearing on the large sprites. The worst offender is the three sprites joined together to make the large skull. Drawing that looked a mess. What would normally happen in this type of update, would be the game waits for a specific interrupt and then tries to beat the raster before it overtakes the drawing routine. Instead of possibly waiting nearly a whole frame for a specific interrupt, Derrick decided to uses the interrupts as a screen position indicator. Then only wait then, if the raster would overtake the sprite while it was being xor'd on and xor'd off. A very slight change but a dramatic change in the visuals.

The point of part explaining the XOR on,  XOR off problem. Is to partially explain why that could not happen on a rewrite from the Amstrad back to the spectrum. The spectrum needed to have a copy screen to stop the problems of screen updates. A copy screen uses a big chunk of Ram, and this game was being converted from a 64k amstrad to a 48k spectrum. All un-needed code was immediately deleted. The first deletion was all the cheat code.

The 2nd problem of moving from the CPC to the spectrum was collision detection. The CPC has pixels with colour. This is changed to the spectrum having pixels that are coloured via an attribute overlay. The CPC used the colour of each pixel for its collision detection. In essence a white pixel under a sprite indicates a collision. The spectrum has to use a whole new logic to detect a collision. And that new logic made a need for a copy screen to be implemented

Collision detection moved from sprite drawing on the cpc, to sprite collision based only on Willy on the spectrum. This was a major change in collision detection. 


So any time I mention cheat code on an original from the CPC or the cheat code in JSW2 . You can immediately assume I am not including any spectrum version of JSW2. 

So only two tape versions written for the CPC. And only one tape version written for the spectrum.

Any spectrum version that does not include the colour code input question is also a hack of the original, in the same manor as the hacks generated for the Amstrad versions. But in the spectrums case, there was no mysterious data to worry about.



#10605 Free space and code optimisation in Manic Miner

Posted by Norman Sword on 23 January 2019 - 01:29 PM

STEP2_INDEX: is/was used by another routine as you suggested.

The unused bits. The routine was written to encompass change in the data. So the length of each data table was selected to enable editing. 3 bits for time and five bits for pitch change. The subroutine contains an "AND 31" which is handy for the pitch change data. No point in adding another instruction somewhere just to reclaim the bit that is/was not used in the data. 

The notes length still needed to have an "AND ?" added somewhere, and here I felt the editing of the data was unlikely to add longer notes. So "AND 3" was selected and then the additional bit was free. The point of including the statement that this is unused. Was to emphasize that the bit is available if wanted. and the "AND 3" can be changed to "AND 7" and additional note lengths can be added.

I can think of no use for the spare bits, apart from the what they were originally designated to be used for.





#10602 Cheat Table in CPC amstrad version of JSW2

Posted by Norman Sword on 23 January 2019 - 11:24 AM

I delved briefly into the Amstrad world again. Just not welcoming enough to pursue further.

As shown I managed to get as far as invoking the cheat. To progress further I needed to move the cursor etc. This did not happen. Every time I ran the program on this particular emulator, it managed to place the interrupts in the wrong place (hence the yellow band across the screen)  

Five emulators later and several versions of JSW tried, I lost interest. The emulators and their presentation/interface does not inspire any interest to me. I will stick with an emulated spectrum.

The cartography room on the small 60 room version- Does not need to have the room name in memory, only the data for the rooms.  I would expect the data to be just tagged onto the end of the 60 rooms. On examining that area on the emulated Amstrad , all I found was a message saying "What's this then".....

Typing in the cheat on the 60 room version has no effect.

Derrick has stated that only two tape versions of JSW exist (master taped by him) and any other version was not mastered by him and must be an edited version- (scope of editing being how the game loads)

And I myself will not be looking any further into this.



Attached Files

#10598 Cheat Table in CPC amstrad version of JSW2

Posted by Norman Sword on 21 January 2019 - 10:51 AM

The versions of Jet set Willy available. (CPC Amstrad) 

S.Wetherill left Software Projects just after completion of writing the game, D.P.Rowson stayed on at Software Projects long enough to reduce the game to 60 rooms for the "they sold a million" compilations.

The MASTER GAME tapes used at software projects were not just copies of the game, they were versions of the game that saved the game. For both Manic Miner and Jet Set Willy on the Amstrad, a version of the game was provided that saved itself, or more specifically the game from within its data. The MASTER GAME version was used by the duplicators to generate a clean copy of the game to duplicate.

The above MASTER GAME tape version  is used every time a new batch of tapes is duplicated.

Since D.P.Rowson and S.Wetherill had left Software projects the creation of any new MASTER GAME tapes (the one that saves itself) is a problem. The programmers that created the originals have left the firm.

So any decisions from that point onward, that forces a need to change how the game works or its environment, creates a need for the creation of a new MASTER GAME tape, or if the media is changed the creation of a MASTER GAME disc

Looking at the versions that have been released by Software Projects we are presented with slight changes in how the game loads. These changes can only be achieved by creating a new MASTER GAME tape or disc. 

Problems created by the master tape - I assume software projects decided at some stage to release the game via the budget software labels, or it was decided to put the game on disc or perhaps just remove the colour code entry part.  What ever the reason this decision involves the creation of a new MASTER GAME tape or disc. The problems presented to anyone who has tried to copy or break into the original game are now presented to the actual software house who want to do a re-release, any change needs to have a new MASTER GAME available. So someone is tasked with hacking into the game and changing it. I would assume that since the colour code entry part and the part that scrambles the data are the same. It was necessary to invoke that code and unscramble the game. Thus removing the protection and the code entry part. The person who managed to remove the protection (possibly) also casually removed the cheat code table. The in house hacker is perhaps not aware of its existence and so its removal seems to do no harm. (assumption) Without reloading all the various versions I believe that any commercial version that does not have the code input question are derived from the in-house hacked version. The programmer who created the new MASTER GAME version had (possibly) destroyed the cheat table. From that point onwards it becomes easy to save the game to disc minus the code input. While I had an emulator for the CPC running, I could not find a version that allowed the full usage of the cheat code. And that is the basis of this assumption. 

The cheat code has two variations. The first as mentioned EMMRAIDNAPRRRTT and the second is the same code with the addition of the letters HI before it. This by logic must be the cheat code for the 60 room version of JSW on the Amstrad, and this means the 60 room version has the cartography room in it. The cheat code was extended for the other version of JSW, and I only know of two versions on the Amstrad CPC 

Is the cheat table intact on the 60 room version ? Far less memory used and thus far easier to put on disc or re duplicate with the data intact. This version was not checked. 



The above is concerned with versions that were commercially sold, on top of that we have the versions that are hacked versions of the commercial versions. These could have the cheat code only if hacked from the original tape (I did not find any when I was looking)


The above is fragmented, just like a badly fragmented hard disc. So very difficult to discus a topic, when it is possible a sentence contains snippet of information that is open to misinterpretation or its details are unknown to the reader. 


Edited to be hopefully clearer.





#10594 Cheat Table in CPC amstrad version of JSW2

Posted by Norman Sword on 19 January 2019 - 02:51 PM

The original version of Jet set Willy written for the Amstrad, and was subtitled "the final frontier". It was only at a later date that this was reduced back into Jet Set Willy. (e.g. removal of around 77 rooms from the 137 rooms. And the reduction to the mere 60 of the original) The reason for the re-write was that is was going to be included in a compilation tape, and the original 60 room (short version) was deemed to be more in line with the original spectrum version.

The original Amstrad version was undertaken on a 464 and at that time the 664 and all the subsequent versions were not known about.

Weeks (perhaps days)  before the release of the Amstrad version of JET SET WILLY the 664 was released. A version of the 664 was obtained and the game was checked to see if it would run. It was found that Amstrad had moved the location of data concerning the keyboard. This change made the keyboard inoperative from a games perspective. This was easily solved by modifications to the games code made from basic. This is the reason that the question is asked before the game is run.    

The above is the time frame of the game game in relation to its release onto the market. The point of interest is the statement that the 664 was not available at the time of writing the game. This shows that discs and disc drives were not considered in the games writing.

*** The above (two paragraphs) was edited to be crossed out. The reason it was crossed out was the assumption I was incorrectly thinking about Manic Miner. Since then I have loaded the game and it does indeed ask for the machine type. 

The game, loads from tape and the tape version (original) has a full cheat code and cheat routine built into the game. When a disc drive is added to the operating system, The disc drive is allocated a chunk of memory just for disc operation. The allocation of the disc drives and the memory allocation of the game clash. The memory allocation that is overwritten by the disc drives and the disc routines (especially the loading from disc) is crucial to the cheat code and the way it works.

Unfortunately the memory overwritten seems to have no effect on the game. So every copy since the original tape version seems to work fine with the data missing. However the very extensive cheat code relies on the missing data, and it seems that for many the cheat code was a mystery, and therefore it's omission was not noticed.

The cheat data is a small table of values (around 150), whose purpose is to index into the data from the cartography room. Since this data is ommited from all disc and subsequent versions, when the cheat is invoked the translation from a position on the cartography room map into a real room number does not happen. This makes invoking of the cheat pointless. ( another reason most people do not know of its existence- once cheat is invoked, it does not seem to do anything worth having)

Anyone familiar with the Tandy and its operating system will be aware that a similar problem was occurring on that machine. Through its evolution and need to be backward compatible, the machine also needed to allow programs to load via tape, and run. But be able to load the same programs from Disc. The conflict in memory was overcome by the simple addition of another routine to the games loading routine. This routine would load the game into memory that was NOT occupied by the disc system. Once the game had loaded it would, shut the disc system down and then move the game into the area used by the disc system and then run the game.

The game Jet Set Willy could be fixed in a similar way. In this case it does not need the whole game moving, It needs just the data from the area that is missing from all the games since the original. 

If the game is loaded (as normal) from disc, and at the point of normally entering the machine code from basic, we instead load the missing data to the screen. The missing data around 150 bytes can have a very simple LDIR routine appended. The procedure then is to run the code and move the data over the area used by the disc operating routines and directly enter the game. The area overwritten by the LDIR copy will be corrupted by the game anyway, and doing this will have no impact on the machine. But it will fix the cheat. Loading such a small amount of data should be so rapid that it would hardly noticeable. 

In JSW Amstrad the cheat code is Invoked by typing   hiemmraidnaprrrtt. This will bring up the cartography room and from there you can select a room. Then you can select a position to start the room.

The finer details of the cheat invocation and usage are not listed here. It is such a long time since I have looked at the CPC Amstrad version, I can not even remember what exactly is done to invoke cheat, (it might be pressing the break key)   

It is a shame no has looked into doing this. At present I do not have the CPC set up and Emulated. (deleted many years ago) and the set up. emulation , modification, save etc needed to do this modification is not something I plan on doing.  


#10593 Free space and code optimisation in Manic Miner

Posted by Norman Sword on 19 January 2019 - 10:38 AM


; The only reason for this updated routine is the change in lighting the keyboard on silent notes
; if two notes are defined, and with the same values the keyboard will still light up. For silence use 0,0



;the complete title music routine for Manic Miner
      LD DE,L33902                ; music data
; is the music finished
      LD A,(DE)                       ; length and pitch shift
      OR A                              ; terminate on zero
      RET Z
; else find out the note length
      AND 3                            ; a third bit is available if needed
      LD C,A                           ; NOTE LENGTH
; and extract the pitch shift
    LD A,(DE)                        ; THE note shift INDEX
    LD B,A                            ; THE NOTE SHIFT
    INC DE
; extract the first note and light keyboard
    LD A,(DE)                      ; (NOTE 1)
    INC DE
    PUSH DE                      ; the data pointer saved - we want to use the registers
    LD D,A                          ; (NOTE 1)
;This must calculate at least one value for HL
    CALL NOTE_PLACE    ;  we want HL the pointer
;calculate 2nd  NOTE
    LD A,D                          ; NOTE 1
    ADD A,B                       ; NOTE 1+NOTE SHIFT = NOTE 2
    LD E,A                          ; NOTE 2
; here we light the keyboard only if this NOTE HAS A VALUE
    INC D
    DEC D

    JR Z,same_notes1
    LD (HL),PAPER2+BRIGHT ;#50 ; press the key  #50=(PAPER RED +BRIGHT)
    CALL NOTE_PLACE          ; returns HL
    PUSH    HL
; same as above, only light if this has a value
    DEC E
    INC E
    JR  Z,same_notes2
    LD (HL),PAPER5             ; PAPER5=40=$28   PAPER5=(PAPER CYAN)
    LD B,0                              ; initialise b
    LD A,L                              ; the border is set from L
;   XOR A                              ; an alternative to the line above, force no border flash 
    LD L,E                              ; NOTE 2  (LOWS)
    LD H,D                             ; NOTE 1  (HIGHS)

;what's what now
;E   =  L - NOTE 2  (low registers)
;D   =  H - NOTE 1  (high registers)
;BC =  - overall duration ; note B=0 length is in "C"
;A    =  last piano keyboard position and this dictates the border colour

        OUT  (#FE), A             ; note 1 (HIGH REGS)
        DEC  D
        JR  NZ,tl_1
        LD  D,H
        XOR  #18
        DEC  E                        ; note 2 (LOW REGS)
        JR  NZ,tl_2
        LD  E,L
        XOR  #18
        DJNZ  tune_loop       
        DEC  C
       JR  NZ, tune_loop
; noise completed, erase the pressed keys
    POP HL
    LD (HL),PWHITE                ; PWHITE=56=$38   ; the default keyboard colour 56=$38=(paper white + ink black)
    POP HL
    LD (HL),PWHITE                ; PWHITE=56=$38      ;
; restore the data  pointer
    POP DE                             ; restore the note pointer
;check for exit and then loop
    CALL test_enter                  ;#9337    ; Check whether ENTER or the fire button is being pressed
    RET NZ
    JR manic_loop


ATT15  equ $59e0                 ; The attribute position of the piano, which is defined as line 15


; calculate the screen position to press a key based on "a" the note
    LD HL,ATT15                   ; this address is free to move up or down the screen
    SUB 8
STEP2_INDEX:                  ;used by the delay routine
    AND 31
    ADD A,L
    LD L,A
    ADC A,H
    SUB L
    LD H,A

    LD A,(HL)

;: Title screen tune data (The Blue Danube)
;  The tune data is organised into 95 groups of two bytes
;      two bytes of data for each dual note played.
; The first byte is split into two values.
; byte 1 ------xxb the xxb value is an index into the TIME_SHIFT array. This supplies the note duration.
; byte 1 xxxxx---b the xxxxxxb value is an index into the NOTE_SHIFT array. this supplies the shift from the first note to the second note
; byte 1 -----x--b the xb value is unused in this data
; byte 2 xxxxxxxxb the xxxxxxxxb value is raw data and is the value of the first pair of music notes. This value is identical to the middle
;  value in the original 3 byte layout. It is the NOTE value for NOTE 1
; The value extracted from the NOTE_SHIFT array is simply added to the value of NOTE 1
; the sum of NOTE 1 and the NOTE_SHIFT value is the value for NOTE 2

;            0     1      2       3     ; notice I have used 75 (which is the logical value to use)

      DB 25,  50,   75,   100



;              0  1    2   3    4     5     6     7     8     9    10   11   12   13    14    15

    DB      0,  1,  6,  8,  10,  11,  13,  20,  21,  24,  32,  43,  75,  77,  128,  154

DB 2+8*1,128
DB 2+8*1,102
DB 2+8*1,86
DB 1+8*1,86
DB 1+8*10,171
DB 1+8*3,43
DB 1+8*3,43
DB 1+8*10,51
DB 1+8*6,64
DB 1+8*6,64
DB 1+8*10,171
DB 1+8*1,128
DB 1+8*1,128
DB 1+8*1,102
DB 1+8*1,86
DB 1+8*4,86
DB 1+8*8,171
DB 1+8*4,43
DB 1+8*4,43
DB 1+8*8,171
DB 1+8*7,48 
DB 1+8*7,48 
DB 1+8*8,171
DB 1+8*1,136
DB 1+8*1,136
DB 1+8*1,114
DB 1+8*1,76
DB 1+8*1,76 
DB 1+8*8,171
DB 1+8*8,38
DB 1+8*8,38 
DB 1+8*8,171
DB 1+8*7,48 
DB 1+8*7,48 
DB 1+8*8,171
DB 1+8*1,136
DB 1+8*1,136
DB 1+8*1,114
DB 1+8*1,76 
DB 1+8*1,76
DB 1+8*10,171
DB 1+8*6,38
DB 1+8*6,38 
DB 1+8*10,171
DB 1+8*6,51
DB 1+8*6,51
DB 1+8*10,171
DB 1+8*1,128
DB 1+8*1,128
DB 1+8*1,102
DB 1+8*1,86 
DB 1+8*1,64 
DB 1+8*11,128
DB 1+8*5,32
DB 1+8*5,32
DB 1+8*11,128
DB 1+8*3,43
DB 1+8*3,43
DB 1+8*11,128
DB 1+8*1,128
DB 1+8*1,128
DB 1+8*1,102
DB 1+8*1,86 
DB 1+8*1,64 
DB 1+8*9,128
DB 1+8*2,32 
DB 1+8*2,32 
DB 1+8*9,128 
DB 1+8*4,38 
DB 1+8*4,38 
DB 1+8*0,0  
DB 1+8*1,114
DB 1+8*1,114
DB 1+8*1,96 
DB 1+8*1,76 
DB 1+8*13,76 
DB 1+8*1,77 
DB 1+8*1,77 
DB 1+8*13,76 
DB 1+8*1,91 
DB 1+8*1,86
DB 1+8*15,51 
DB 1+8*1,51 
DB 1+8*1,51
DB 1+8*15,51
DB 1+8*1,64 
DB 1+8*1,102
DB 1+8*1,102
DB 3+8*1,114
DB 1+8*1,76 
DB 1+8*1,86 
DB 1+8*12,128
DB 0+8*14,128  
DB 0+8*1,128
DB 1+8*12,128

DB  0 ;End marker

#10589 How Fast Can Willy Play

Posted by Norman Sword on 16 January 2019 - 05:43 PM

The missing room was room 47. ( was that the original "]" room ? )

The file has again been updated with the whole 128 rooms. I suppose I should delete the game file and update that as well (not done) 

#10587 How Fast Can Willy Play

Posted by Norman Sword on 14 January 2019 - 02:06 PM

The colouring is just another product of the game design.

Four pictures . 1 and 3 use "OR" for sprite draw. pictures 2 and 4 use "XOR"

We are looking at the pixels and not the colours. In picture 1 ("OR") Willies head is merged into the background, and the pixel pattern of the background is lost beneath his head. In picture 2("XOR") Willies head now exhibits the pixels of the background beneath the outline of his head.

The differences in picture 3 and 4 have been lost due to the collouring changes. But the black shape of Willies lower body in picture 4("XOR")  is a classic look due to "XOR" on the screen.  The area of note here is the conveyor as it passes through the body of willy. The conveyor "Xor"s his body. In the "OR" version we can not see the conveyor beneath his body outline.

I will find some better examples

ADDED pictures. again in groups of two. Pictures 5 and 7 use "OR" and pictures 6 and 8 use "XOR"

Notice the difference in First Landing (pictures 7 and 8) using "OR" the stair merges into the form of Willy. In the usage of "XOR" the stair can be clearly seen

Again the differences in pixels in "Ballroom West" (pictures 5 and 6) the rabbit platform merges into the outline of Willy with "OR" picture 5. In picture 6 "XOR" the outline and Willies shape are now undetermined.


Attached Files