Just Get 9 has just been published!

[BZHGames]

Otherwise, I had a question, the values after 9 are J, U, S and T if I understood correctly (J is sure but since I have never done better, I am not sure of the other values).
Are there others after?
And why these values? I tried to find the logic but without success.

I'm so stupid, rereading my message I just understood "JUST".
On the other hand, does it continue afterwards with "G", perhaps?

[eto]

If you want a ranking, there is one at https://bzhgames.xyz/?CPCGAME=CPC_Just%20Get%209 

Something's off there. 

Due to your post I gave it another quick try and came to level 10 - but now my previous record (level 12, about 1350 moves) is gone and has been replaced by a level 26 record with a mere 374 moves, which I definitely didn't achieve. Not sure what it is measuring now. 

[BZHGames]

Hello eto,
In fact for this game I could not automate the recording of the score. I only did it recently.
But I had not managed well when we exceeded 9.
Thanks to your games, I was able to correct the problem.
I also restored your old score (level 12) and corrected your level (10 and not 26) for the year 2025.
So levels 10, 11 and 12 are now managed correctly, I do not know if we can do more.

[cpcitor]

Familiar faces on this thread.

Have just downloaded the latest build, its good but for speedrunning I think the previous build feels slightly faster and is a touch more responsive.

Hi everyone.

Very interesting question, and the answer is, surprisingly, a little subtle.

Here are my conclusions

1. Is anything delayed, even a tiny bit, in patch3 vs patch2? Yes: the shake motion.
2. Does it slow down the game, even a tiny bit? No.
3. Is the latest build as good for speedrunning as before? Yes.
4. Is there an option that makes patch3 behave the same as patch2? Yes, two actually.
5. Is the game good for speedrunning overall? Yes!

Details

1. Is anything delayed, even a tiny bit, in patch3 vs patch2? Yes: the shake motion.

Patch 3 now works by default in strict video timing mode: all frames are complete, 312 lines, exactly 19968µs apart. This implies to change CRTC register R4 one frame before the start of the shake (done in both cases using R7). This implies that the shake motion is delayed one-frame (20ms).

2. Does it slow down the game, even a tiny bit? No.

* The shake motion is only decoration: you can press direction keys or joystick before the shake subsides and they are processed.
* The cursor motion has no delay in patch3 compared to patch2.
* The tile disappearance, tile fall, sound, behave exacty the same in patch2 and patch3.
* The time between the moment you confirm a move and the moment you can play the next move is exactly the same.

3. Is the latest build as good for speedrunning as before? Yes.

One might argue that a shake motion delayed one frame might hinder the player. Given the size of the graphics you can recognize the numbers well enough. Plus, given what is explained below, I think there is a difference that only a really good speedrunner can see (proof: you saw something) but it doesn't actually delay the processing of your inputs, absolutely not.

4. Is there an option that makes patch3 behave the same as patch2? Yes.

You can disable sound in the main menu, which also disables screenshakes. With this option, patch2 and patch3 behave exactly the same.

5. Is the game good for speedrunning overall? Yes!

This section applies to all versions of the game.

What delays the appearance of the cursor? In-flight tiles. This is short (only when many tiles disappear at once it can become a little longer). As soon as all tiles are packed as a grid, the cursor is shown and can move (irrespective of whether the screen is shaking). In-flight tile do introduce some latency (you can't see a moving cursor during that time) but an advanced player can compensate for it (I do and I'm sure as a speedrunner you do).

More precisely, you can send your inputs at any time. During tile flight they are still processed and buffered (by the firmware).

If your buffered inputs requested only cursor motion, you will see the cursor quickly move right towards the target tile. If your buffered inputs actually confirmed a move, you may not see the cursor at all, as your move is processed right away (played tile disappear, in-flight tiles again).

All in all, this makes opportunities for speedrunning: if you can recognize figures, think your move, press your keys at extreme speed before the tiles have finished falling, you're ahead of other players.

You cannot view this attachment.

Extra information

Thanks again @Longshot for hinting me that hardware scrolling and strict video timing was possible.

Due to some subtle interplay between gate array generating interrupts, firmware looking for VSYNC signal to tell whether this interrupt is a new frame, and changing CRTC registers in the previous frame, sometimes in patch2 and before on huge screenshakes the interrupt was not recognized as a new frame, leading to a slight slowdown in the music. This was not a big issue at the time, but it made the perfect video timing code ineffective. I fixed this by counting interrupts.

Just wondering also Mr CPCitor, how fast have you seen 9 gotten in? I've seen it done in sub 2 minutes and am just wondering is there a leaderboard anywhere, or do you know of any ridiculously fast times?

2 minutes?

I think (not even certain) I did below 4 minutes.

I'd be interested in seeing a video of some of your speedruns.

There is a public speedrun video on

[THE_Chris]

Just wondering also Mr CPCitor, how fast have you seen 9 gotten in? I've seen it done in sub 2 minutes and am just wondering is there a leaderboard anywhere, or do you know of any ridiculously fast times?

2 minutes?

I think (not even certain) I did below 4 minutes.

I'd be interested in seeing a video of some of your speedruns.

There is a public speedrun video on

All very interesting... fascinating to see the levels you've gone to to look into this, so thanks a million for summarizing that.

Yes, sub 2 minutes has been done by my girlfriend. Craigsbar got me into Just Get 9, then we got her into it and she immediately slapped the bejesus out of any of our times. I managed about a 2m30s once, I think, but she has gone sub-2minutes. The two of us made sure she got it on my WinAPE emulator and on Craigsbars genuine hardware. You need to write another special CraigsBar build that can tell if she's playing and adds a minute to her time.

I only took one photo but we'll get her out of Just Get 9 retirement and I'll try to get phone footage of her going sub-2mins again.

[cpcitor]

All very interesting... fascinating to see the levels you've gone to to look into this, so thanks a million for summarizing that.

I'm glad you appreciate. Yes there are rabbit holes in many places.

Yes, sub 2 minutes has been done by my girlfriend. Craigsbar got me into Just Get 9, then we got her into it and she immediately slapped the bejesus out of any of our times. I managed about a 2m30s once, I think, but she has gone sub-2minutes. The two of us made sure she got it on my WinAPE emulator and on Craigsbars genuine hardware.

Amazing! You can congratulate here very much!

Just to gauge with one data point, I just did this:

You cannot view this attachment.

You need to write another special CraigsBar build that can tell if she's playing and adds a minute to her time.