It's no more a MiSTery

[Albator]

Hello, thanks for all your effort on this core.
I still have a problem with the "Fantasia" demo from Dune ...
https://demozoo.org/productions/10013/
The display is stuck at the "Dune" logo at the beginning of the demo
It works on the old ST core.

[slingshot]

I still have a problem with the "Fantasia" demo from Dune ...
https://demozoo.org/productions/10013/
The display is stuck at the "Dune" logo at the beginning of the demo
It works on the old ST core.

Works for me (ST mode, TOS 1.04). What's your setup?

[slingshot]

Don't think so. Shifter can't latch on the CS leading edge. Data is not yet available.

Just checked this in the sim, true, that cannot happen (CAS goes low at the same time as LOAD).

Sorry, I wrote nonsense. LATCH asserted after CS goes down, so the data is not available.

Upd.: just thought further - palette registers are latched during the whole asserted state of CS, that might store an invalid value for a moment (I'm not sure it is correct). What if the same happens with res?

[Albator]

I still have a problem with the "Fantasia" demo from Dune ...
https://demozoo.org/productions/10013/
The display is stuck at the "Dune" logo at the beginning of the demo
It works on the old ST core.

Works for me (ST mode, TOS 1.04). What's your setup?

By default I'm using STE, Tos 1.62, 4MB ram.
I've tried ST, tos 1.02 and 1 Mb ram, no luck either.
I simply download the .msa file, convert to .st using hmsa utility, and boot straight to it (no hard disk).

[slingshot]

By default I'm using STE, Tos 1.62, 4MB ram.
I've tried ST, tos 1.02 and 1 Mb ram, no luck either.
I simply download the .msa file, convert to .st using hmsa utility, and boot straight to it (no hard disk).

On your link, I've downloaded a zipped TOS file only. That worked (used zip2st from Hatari to put it on a disk).

Upd.: the other version (in the MSA) needed a small fix in the FDC.

[ijor]

Don't think so. Shifter can't latch on the CS leading edge. Data is not yet available.

Just checked this in the sim, true, that cannot happen (CAS goes low at the same time as LOAD).

Sorry, I wrote nonsense. LATCH asserted after CS goes down, so the data is not available.

It is neither CAS neither LATCH, it is WDAT what matters most. Although the main issue is just that the slot timing doesn't allow to latch too early.

Upd.: just thought further - palette registers are latched during the whole asserted state of CS, that might store an invalid value for a moment (I'm not sure it is correct). What if the same happens with res?

Palette registers are complex cells based on an asynchronous transparent latch. No way the REZ registers could use async latches. They must be synchronous.

Btw, please don't abuse editing old posts to add new information at a later time. Many times it get unnoticed. In this case I noted it just by chance because I didn't read the thread before you edited the post.

[slingshot]

Sorry for abusing the edit function. I think editing shouldn't be allowed when it's not the last message in the thread.

[ijor]

Sorry for abusing the edit function. I think editing shouldn't be allowed when it's not the last message in the thread.

No need to apologize. Whas just a comment

Allowing or not editing old messages is probably the most controversial topic in most forums. I believe there were some limitations here before, guess that configuration probably got lost with the forum update.

[jamesrc]

I've got five straight days off and I've hooked up my MiST so I'm grabbing the latest core and then will continue with PoV disks.

I will say this -- having used the Spectrum Next for a couple of months, I *hate* the cheap keyboard I'm using with the MiST and think a new one is in my near future.

[jamesrc]

POV 26.

The Ultimate Demo -- This is the most broken I've seen something on this core in a long while! And it's not that broken. Hit Space to get past the credits and greets and then use the left/right cursors and return to select The Sprite Demo.

I think I could fix this finally. Also the BLT3 in SkidRow. Interestingly the proper fix (at least what I believe is the proper fix) still allow de-syncing in some cases (about 2 times out of 10), and this is ack'ed in the scrolltext, too (that's why I believe the proper fix is found). Maybe that was the last MFP problem (or not, hehe).
Now just why Closure doesn't work in ST mode...

However finally the forum is back!

The Sprite Demo on The Ultimate Demo is still very very flickery here with a large number of the sprites winking in and out near constantly.

I'm on MiSTery-200621.rbf.

I can post a video if it would be of use.

[slingshot]

The Sprite Demo on The Ultimate Demo is still very very flickery here with a large number of the sprites winking in and out near constantly.

I'm on MiSTery-200621.rbf.

I can post a video if it would be of use.

Maybe it's happening on original HW, too, just less obvious with a CRT display. Either increasing or decreasing the DE -> Timer B delay makes it worse.

[jamesrc]

Maybe it's happening on original HW, too, just less obvious with a CRT display. Either increasing or decreasing the DE -> Timer B delay makes it worse.

My MiST is on a CRT. I'm moving on with other disks, but at some point I'll post a video in case reallarry has a moment to test on his always-to-hand STe.

I just finished with PoV 30, which works perfectly on the MiSTery core but which left me swearing. The "CST Demo" gives four bombs when you hit Space. Thinking it must be loading a second screen and crashing, I tried multiple machine types, memory and TOS settings to no avail. I tried on Hatari and go the same result. So than I tracked down another copy of PoV 30 and got the same result again.

Finally I tracked down the original demo and found there is no second screen and the PoV version just doesn't exit cleanly. PoV 30 took an hour of my life that I want back.

[slingshot]

Finally I tracked down the original demo and found there is no second screen and the PoV version just doesn't exit cleanly. PoV 30 took an hour of my life that I want back.

But it's a good sign that bugs from the real world also replicated

[jamesrc]

Yes!

I'm currently working on PoV 35.

Have I mentioned that I'm reading all scrolltexts on these disks? I might make an exception for a 6 hour scroller or whatever it is on the B.I.G Demo.

[jamesrc]

P.O.V 37 - Ste Presentation 4

There's a flickering group of pixels on the bottom line of the main picture, towards the right, were it splits for the scrolling section below.

On Hatari the pixels momentary flicker when the magnifier above scrolls to the right, but it's otherwise stable.

I think this is one to be tested on actual hardware as I'm just as likely to trust the MiSTery timings as I am Hatari's. So I'm mostly leaving it here for later.

[jamesrc]

But it's a good sign that bugs from the real world also replicated

Reallarry has been very generous and tested all the things I've found where the emulator behaviour does not match the MiSTery behaviour.

In this reply, I'm going to pioneer something I'm calling the "Good news sandwich". The bread of our "Good news sandwich" is going to be two really exciting things where the MiSTery surpasses emulation. Hidden away in the middle as the filling is going to be more news about that Ultimate Demo that I'm sure you'd wish I'd stop talking about.

Very Good Thing #1

There's something on POV 37 called the "Nitrowave Demo". This demo consists of four screens -- a menu, and three screens you can access from the menu by pressing F1-F3. On both Hatari and Steem in STe mode, the full-screen menu renders. It also works properly in STFM mode on the MiSTery core. However, if I run it on MiSTery in STe mode, the menu screen is utterly unreadable (although F1-F3) still work.



Reallarry tested on a real STe and the behaviour matches MiSTery and not Steem or Hatari! He recorded all his tests, so the proof is right here, starting at around the 23 second mark:

https://youtu.be/7wJ_HKGHX8w

(Embedding YouTube videos in the forum isn't working for me, so I'm just giving regular old YouTube links).

Ultimate Demo Incompatibility

Here's what that Sprite Demo section of The Ultimate Demo looks like for me on the MiSTery:

https://youtu.be/zmZkbe9UMxo

I've tried both STFM and STe with different TOS versions. This was recorded in STFM mode because this demo suffers from the known incorrect colours bitplane issue we've seen before when run on the MiSTery. I can confirm that the flickering (of both the screen itself AND of the sprites showing the background behind) still happens in STe mode.

The same demo works correctly on Reallarry's STe (With the part in question starting at the 1:22 mark):

https://youtu.be/CP0r47kyfiY

Note the sprites don't wink in and out to reveal the background behind.

Very Good Thing #2

On that same POV 37 there's the STe Presentation Demo 4. There's a group of pixels between the image and the scrolly bit that filcker continuously on the MiSTery:

https://youtu.be/nZtRqZTPdHI

This same set of pixels flicker only occasionally (when the magnifier quickly flies right) on Hatari, and I haven't tested on Steem.

The real ST once again matches the MiSTery behavior! MiSTery trumps Hatari, again!

The proof is here (with the pertinent section starting at 1:16):

https://youtu.be/homEuVslCVc

Acknowledgement

Thank you, thank you, thank you Reallarry. I think my wife is already quietly frustrated with the amount of real estate my retro computing hobby takes up, and if I had to pull out both my ST and the 20 year old Pentium 75 which is the last PC I have with a real floppy drive for writing disk images, I think she very well might kill me.

I would probably have to abandon this demo testing project without you.

I'm going to push onwards through POV 38 which I started last night. I'll probably take another short break after POV 40, but I've got another long weekend coming up in August as I've got unused vacation time.

[RealLarry]

Well done, James - and thanks for the many thanks of you, but I've the easyiest task: You're checking every disk, I've to check the only handful disks your're concerning of. Not to mention that my STe is always ready to hand and nearly every game or demo disk is already residing on my UltraSatan or Gotek drive.
So to wire, powering on and selecting the right stuff to go is fast and easy to manage that no one at (our) home is registering this

[slingshot]

Good to see you only found one difference in 40 disks
The answer is easy: the "good things" show that the GLUE+MMU combo is created after the real schematics, the other flickering one that the MFP is a functional recreation, which might still has imperfections. However not much, I think.

[jamesrc]

Good to see you only found one difference in 40 disks
The answer is easy: the "good things" show that the GLUE+MMU combo is created after the real schematics, the other flickering one that the MFP is a functional recreation, which might still has imperfections. However not much, I think.

Oh yes! While I'm being thorough, I'm incredibly grateful for all the work you've put into this core. One weird edge case in nearly 40 disks is, indeed, phenomenal. In just about every way that matters, the experience is indistinguishable from having an actual ST on my desk.

So again, thank you.

[slingshot]

For the new release:
https://github.com/mist-devel/mist-bina ... 200725.rbf
I've added USB-RTC support (https://github.com/mist-devel/mist-board/wiki/UsbRtc)
It's a nice little device, which is not only useful for MiST, but can be used in RPis and other gadgets, where one don't want to make an internet connection for NTP.
Now I built it on a breadboard, but I'm planning to order proper PCBs. Maybe I'll order about 10 pieces, I'm curious if somebody need it, then I can sell the excess ones (fully built of course).

[DanyPPC]

Very good !

So, is it supposed it will be supported only on MiStery core ?
Or are you planning to add support to the Amiga Core as well?

[slingshot]

I've already added to Minimig-MiST and the Archie cores, too.

[DanyPPC]

You are Great !

[jamesrc]

I'm now completely through PoV 40, which represents a quarter of the PoV collection.

Imposing a break upon myself, now. I've got a bunch of physical ZX Spectrum Next games that have shown up that I want to try.

But I'll be back at this soon. I'm taking some holiday time at the end of the month.

[krupkaj]

I have just tested the real Atari mouse. It is great there is an implementation of the real IKBD. But I have a trouble with right button of the mouse. It does not work the right way. Right button does nothing but when I press the button on the joystick it does what I expect.

I know that the right button is connected to the joystick port in real Atari. Would it be possible to make it this way also on MIST?