The last MIST

[MasterOfGizmo]

Uhm, the Atari ST core supports 1280x1024 ...

Yes, I've already tested and used it, very nice feature, but a little too much for my actual needs.

Indeed i always planned to build something "in between" like a 4 colour 1024x768 mode. But I never finished that.

[Xyla]

This isn't a "decision". It's just that the MIST doesn't sell anymore (agreed, you obviously bought one ). So Lotharek has plenty of them in stock since quite some time now. The fact that they don't sell means that his stock will last pretty long. And if it doesn't last for long since for some magic reason his stock sells within a reasonable time then he might of course build more of them. After all he's a business man and it doesn't make much sense not to satisfy a demand. Currently there isn't demand so it looks this is the last batch to be produced. That's all.

That all makes perfect sense, and like I said I will probably buy more before they are all gone. But also consider that most people don't know what the MIST is. For a long time I wasn't quite clear on how well it would function as an Atari ST replacement. The direct SD feature that allows you to use the SD card like a hard drive rather than deal with disk images is what convinced me to buy one. I love the product, and I hope more people will discover it... and maybe sales will pick up! I haven't yet tried the 15khz video feature yet but I can't wait to see how MIST looks on an Atari SC1224. This doesn't feel like a cheap generic FPGA board, it feels like an Atari (even though it can be more than just an Atari!)

[sardine]

Thankyou.

Love my Mist, love the work everyone has done with the core and support.

I have everything except one core, Dragon 32/Coco2, I've just not got the talent or knowledge to create the core myself.

The list of cores are amazing and I have a bunch of SD cards here with full setup systems from BBC B, C64,Spectrum to Amiga,Atari, the list is huge.

thank you all.

I know the Dragon 32 Core is a dead duck...
but if anyone here could look at the spectrum core if they get any time and see if they can fix the "Ultimate Games" bug, any of the games run no issues, title screen, sound etc.. but when you start the game its a black screen, the game is still running in the background. its a known issue thats never been solved.. some of my favorite games are ultimate games.

thanks and maybe we will see a Mist 2 ??

[Total Eclipse]

I'd also like to add my thanks to Till, for getting the whole MiST project off the ground. It's been a fun thing to play with, and I still use mine fairly regularly.

It's unfortunate that development on most cores appears to have stalled now, especially on the ST and Amiga cores, but I guess that's the nature of relatively niche projects - the community has a dependency on a very small number of individuals. Many people can contribute to software emulator projects, but FPGA skills appear more rare.

Software emulation still, usually, offers more accuracy and compatibility than the current FPGA boards, but no emulator comes as close as the "real" feeling you get with a MiST.

The MiST has been (and still is) a fantastic little board, and since getting mine I've rarely used my real Amigas (despite a few minor niggles in the current Minimig core). The ST core still has a long way to go if it was to ever run a high percentage of demo's, but is great for commercial apps.

It's a shame to end production, but understandable. To everyone that's made a contribution to the MiST, be it core development, firmware, or even case designs (I use the red design from karabasbarabas), I thank you!

[Alessio]

Pity. Fully understandable of course, but a pity nonetheless. I'm very happy with the quality of mine bought from Lotharek. Here's hoping the community will still continue to exist and not slowly fade into nothingness...

[iceman]

Here's hoping the community will still continue to exist and not slowly fade into nothingness...

I guess that's up to us! I always wanted to learn how to program an FPGA...

[Wayne123]

It's unfortunate that development on most cores appears to have stalled now, especially on the ST and Amiga cores, but I guess that's the nature of relatively niche projects

At least the MiST has an ST core, I have a FPGA Replay Arcade and the ST core has been promised for a long time now, unless it is hidden somewhere I don't know about, it is nonexistent.

[AdvancedFollower]

I'd also like to add my thanks to Till, for getting the whole MiST project off the ground. It's been a fun thing to play with, and I still use mine fairly regularly.

It's unfortunate that development on most cores appears to have stalled now, especially on the ST and Amiga cores, but I guess that's the nature of relatively niche projects - the community has a dependency on a very small number of individuals. Many people can contribute to software emulator projects, but FPGA skills appear more rare.

Software emulation still, usually, offers more accuracy and compatibility than the current FPGA boards, but no emulator comes as close as the "real" feeling you get with a MiST.

The MiST has been (and still is) a fantastic little board, and since getting mine I've rarely used my real Amigas (despite a few minor niggles in the current Minimig core). The ST core still has a long way to go if it was to ever run a high percentage of demo's, but is great for commercial apps.

It's a shame to end production, but understandable. To everyone that's made a contribution to the MiST, be it core development, firmware, or even case designs (I use the red design from karabasbarabas), I thank you!

I agree, when running the Amiga core, my MiST "feels" like an Amiga, albeit a slightly incompatible one. Emulation is cool, and can of course emulate things far beyond the capability of the original hardware (AmiKit X, for example) but there's something genuine about having the same limitations as the original hardware. The choice for me was to either recap my A1200, get some kind of scandoubler, CF card adapter etc. or just get the MiST for a lower price than all those additions for a machine that could break next month...

It's just unfortunate about the cores - the ST core hasn't been updated since 2016 I think, and the latest Amiga beta core is now over a year old. I think there are many who would like to do something about it, but learning FPGA programming is a big undertaking.

[Xyla]

I'm glad I'm not the only one who thinks the MIST feels like a real whatever-core-you-are-using. It's really wonderful to have that real hardware feeling and also to have the power to make it whichever machine you want.

You know what would be really amazing?... If someone manufactured a laptop case for MIST. I've been thinking a lot about what it would take.

[Newsdee]

You know what would be really amazing?... If someone manufactured a laptop case for MIST. I've been thinking a lot about what it would take.

Close enough: viewtopic.php?f=117&t=32949#p336974

[Kujako]

Well.... for what it's worth I just bought a MiST. So there does seem to still be a market for such things. I did see the other similar systems out there, but I liked the MiSTs feature set the the best of the various options (DB9 joystick ports is a big plus). I did get the 1.3 revision, so am a bit disappointed that the old serial connection is the only way people have managed to get networking running, but perhaps in time that will change. Seems to me that a USB to serial to whatever method could be figured out... perhaps I'll take a crack at it if I get bored.

[philexile]

Hello,

I also bought a real MIST from LOTHAREK last February. I haven't used it as much as I would like because my job is quite busy. A friend, who also bought a real MIST from LOTHAREK, told me that development had ceased and that is what made me search out this forum.

First, I understand your frustration with some imitating your creation. I'm sure this imitation is substandard — which leads to issues for the user base. This also happened to Krikzz. His Everdrives were cloned and those clones are cheap and buggy. The users who bought the buggy clones then complain to Krikzz and his time is wasted. It is frustrating for everyone.

That all said, I don't think that development should halt completely. There is a community out there that invested in this product and they will continue to use and support it. This leads me to another thing I noticed:

NOT MANY PEOPLE KNOW ABOUT THE MIST!!!!! In the USA at least.

I've shown the MIST to many of my friends who are also retro game enthusiast and only one person has know what it was or even heard of it. People can't buy what they do not know about!!! I know there are a few videos on YouTube showing the capabilities of the MIST, but those are older and I don't think these are reaching parts of the ever-growing retro gaming community.

I would suggest that you or LOTHAREK send out review units to some of the more popular YouTube review shows. The first that comes to mind is My Life in Gaming and second would be MetalJesusRocks. Either of those shows would be impressed with the capabilities of the MIST and it would increase the visibility of the device significantly. Having a reseller based in the United States would also help.

In closing, thank you for the fantastic device and I hope this isn't truly the end.

Best Regards

It's great that you bought a real MIST!

This isn't a "decision". It's just that the MIST doesn't sell anymore (agreed, you obviously bought one ).....

[AdvancedFollower]

Marketing the MiST is tough. It costs several times more than a Raspberry Pi 3+case, which is hard to justify for the "casual" retro gamer, especially as the Pi emulates more systems. FPGA is arguably more "authentic", but most people probably don't care, when they just want to mess around with a few childhood games one evening over a beer. There's also the problem that emulation (software or FPGA) is a bit of a grey area. With companies like Nintendo capitalizing on the current retro trend with their own ARM-based recreations, it's probably wise to keep a low profile (even though the MiST itself doesn't infringe on any copyrights or trademarks).

Of course if Lotharek had the economics of scale, the price could come down, and more units would be sold, so it's a bit of a catch 22.

[philexile]

Marketing the MiST is tough. It costs several times more than a Raspberry Pi 3+case, which is hard to justify for the "casual" retro gamer, especially as the Pi emulates more systems. FPGA is arguably more "authentic", but most people probably don't care, when they just want to mess around with a few childhood games one evening over a beer.

There is a market for this. Look at other expensive devices like the Framemeister, NeoSD, Super SD System 3, or any of the premium Everdrives. The EverDrive-GB is a $130 device that sits alongside cheap alternatives that cost 10% of that price.

With companies like Nintendo capitalizing on the current retro trend with their own ARM-based recreations, it's probably wise to keep a low profile (even though the MiST itself doesn't infringe on any copyrights or trademarks).

That is crazy! Look at what Analog is doing with the Analog NT Mini and the Super NT! High-end, FPGA devices that cost $500 and $200 respectively! The Super NT was released a few months after Nintendo's SNES Classic.

A low-profile is not required.

Of course if Lotharek had the economics of scale, the price could come down, and more units would be sold, so it's a bit of a catch 22.

More units could be sold if more people knew the units were out there!

Also, I happened to find this YouTube review of the Mist: https://www.youtube.com/watch?v=RW9b22D0nWY&t=3s

Its the only recent review of the MIST and the reviewer had a lot of problems, many of which I think were due to him not knowing the ins and outs of the MIST. Maybe someone here could reach out to him for suggestion solutions to the issues he was having  — mostly with the Atari ST and MSX.

Another person to reach out to would be The 8-Bit Guy — he has over half a million subscribers: https://www.youtube.com/channel/UC8uT9c ... u7ITLGo9Ww

[Kujako]

I found Lotharek on eBay here in the states while looking for parts to build an Amiga. Not sure how one could better spread the word... perhaps some sort of tie in with Amiga Forever? Would be interesting to see them provide a ready to boot SD card. Not yet received my MiST (Poland is far away) but definitely looking forward to it.

[Foxie]

Here's hoping the community will still continue to exist and not slowly fade into nothingness...

I guess that's up to us! I always wanted to learn how to program an FPGA...

I'll probably have to learn VHDL or Verilog some day. You can program FPGAs and CPLDs with schematic entry, which I prefer since I come from a hardware background. I don't think in lines of code. But it seems most FPGA projects use VHDL or Verilog since it's friendlier towards programmers.

The trickiest part for the beginner is soldering the tiny FPGA and meeting its very strict power requirements. Some FPGAs need three voltages at considerable current, which all need to be sequenced correctly during power up and power down. You could just buy a dev board, but properly designed boards cost quite a bit.

I do prefer CPLDs because they only need a single supply rail and have internal configuration storage. They're also still available in big, easy to solder packages. The problem is the number of registers is tiny, which rules out lots of applications. They also still have strict requirements about the behaviour of the power supply during power up and power down.

I did get the 1.3 revision, so am a bit disappointed that the old serial connection is the only way people have managed to get networking running, but perhaps in time that will change. Seems to me that a USB to serial to whatever method could be figured out... perhaps I'll take a crack at it if I get bored.

Does the MIST actually implement an Atari modem port? I was under the impression it had no legacy ports at all (apart from joystick).

For my use, I need at a minimum a cartridge port. Preferably also a printer and modem port. Without those, I may as well just run Hatari on a Raspberry Pi - it has better demo compatibility, same speed, and higher resolution support. The main advantage of an FPGA implementation is the ability to easily use legacy interfaces - that's very difficult to do on an emulator. The lack of cartridge port keeps the MIST from entering the hands of musicians, sadly.

I'm sure the MIST has its particular niche, but in order to be a suitable music tool I'd need to learn enough about VHDL to somehow implement my own cartridge and printer ports. I'm not sure if the existing MIST board has a way of connecting TTL signals to a pin header though? Perhaps if the Cubase dongle reverse-engineering project succeeds then the dongle and MIDI expanders can be built right into the FPGA eliminating the need for a cartridge port. You'd still need a few I/O pins for getting a few ports of MIDI data out of it.

I wonder if the joystick ports can be re-purposed as digital I/O lines? If you combine two joystick ports, you might have enough I/O lines to implement a printer port. That would at least be useful for the cracked copy of Cubase, if you can live with the crack's shortcomings.

I found Lotharek on eBay here in the states while looking for parts to build an Amiga.

Are you building an Amiga from the chip level? I heard of someone attempting that, they obtained all the necessary custom chips and designed their own PCB. I'm not sure if they ever succeeded.

[Kujako]

Does the MIST actually implement an Atari modem port? I was under the impression it had no legacy ports at all (apart from joystick).

The docs for the Atari core say they support ethernet using a DLink DUB-E100 USB-Ethernet-Dongle (http://www.atari-forum.com/viewtopic.php?t=29684). In past hardware versions of the MiST people have used the onboard serial pins (where the MIDI daughter board connected) to get the Amiga core online by having a modem emulation device (such as a Raspberry Pi). However, with the 1.3+ hardware the MIDI is built in and the board no longer has the pins for direct serial.

I don't see why a USB to serial adaptor couldn't be gotten to work, or worse case doing a MIDI to serial adaptor (MIDI is serial, just at odd bit rates).

Are you building an Amiga from the chip level? I heard of someone attempting that, they obtained all the necessary custom chips and designed their own PCB. I'm not sure if they ever succeeded.

I was going to get a 1200 motherboard, replace the capacitors, upgrade it and build a new power supply. But the price of the project was going to be double what the MiST cost.

[Martinus]

Here's hoping the community will still continue to exist and not slowly fade into nothingness...

I wonder if the joystick ports can be re-purposed as digital I/O lines? If you combine two joystick ports, you might have enough I/O lines to implement a printer port.

I do remember that on my Atari 400 I had a special cable + driver that connected to the 2 right joystciports and had a parallel/centronic(?) plug for printers on the other side. So similar is possible,

[Foxie]

I don't see why a USB to serial adaptor couldn't be gotten to work, or worse case doing a MIDI to serial adaptor (MIDI is serial, just at odd bit rates).

The problem is USB is unsuitable for MIDI, because of its latency/jitter (the 1ms framing and bulk transfers). So there needs to be some more direct method of connecting to the core without the overheads of USB.

Can the serial I/O be configured for higher baud rates, like 115.2k or even higher from the Atari side? This would be acceptable for driving about four MIDI outputs, which is better than nothing. I've already designed a MIDI expander that can accept serial data at 102.4k (standard ST) or 115.2k (TT/Falcon, everything else) and sends it out to four ports. To get the best performance you really need two handshaking outputs and an input in addition to the serial I/O, but I've also implemented a protocol that eliminates the need for the handshaking outputs. This might be suitable if there's only TX/RX available. The handshaking input could be done via the joystick port at a pinch.

The serial I/O would need to appear to the Atari independently of the MIDI ACIA. The reason is because Cubase completely takes over handling of the MIDI ACIA, so there is no possibility to write a Cubase driver to send high baud rate data out of the modified ACIA implementation. Cubase drivers can only address non-ACIA hardware like the MFP, SCC, printer port and cartridge port.

I do remember that on my Atari 400 I had a special cable + driver that connected to the 2 right joystciports and had a parallel/centronic(?) plug for printers on the other side. So similar is possible,

The real question is how the joystick ports are implemented on the MIST. If they just go straight through to the FPGA with nothing more than ESD clamping, it would be fairly easy to convert them into a printer port. Configure them as I/O rather than inputs and then add some external circuitry to level shift to 5V. If the inputs are buffered on the board, then it won't be possible unless the board is modified.

[Sorgelig]

joysticks are connected to MCU.

[Kujako]

I don't see why a USB to serial adaptor couldn't be gotten to work, or worse case doing a MIDI to serial adaptor (MIDI is serial, just at odd bit rates).

The problem is USB is unsuitable for MIDI, because of its latency/jitter (the 1ms framing and bulk transfers). So there needs to be some more direct method of connecting to the core without the overheads of USB.

Not seeing that as an issue, since the onboard MIDI negates the need for support over USB. To be clear, this is directed at the 1.3+ MiST model which has no serial pins but has the MIDI daughter board built in. To get the Amiga core online, people have been using the serial pins that the newest model lacks. So I'm looking at if I can use a USB to serial adaptor to then connect to the internet using a modem emulator, or building a MIDI to serial adaptor to do the same.

Also not an issue with the Atari ST core, since it has support for USB ethernet (if for only a single model adaptor).

[Foxie]

Not seeing that as an issue, since the onboard MIDI negates the need for support over USB.

Unfortunately, the single MIDI output that comes standard with the Atari is nowhere near enough for the needs of even a very basic music setup. It's pretty typical to need something like 8 independent outputs, which can be achieved with devices like the Midex, Unitor and Log3 (cartridge port), SMPII, Midi3, Soundpool MO4 and Friend Chip MM1 (printer port), Export (modem port) and on the MegaSTE/TT/Falcon you can connect a Macintosh MIDI interface to the LAN port. In all, you can connect up to 15 additional MIDI outputs on a 1040ST under Cubase - and 8 additional outputs under Notator Logic.

That's why - even if you do succeed in running a crack without it crashing - you almost always need a cartridge port or printer port at a minimum. Although the printer port has better performance for MIDI I/O, the cartridge port is probably the most important of the two - because so many programs rely on a dongle and have never been cracked reliably. The Cubase crack is probably one of the more reliable cracks out there, but even that's on the ragged edge of usability. It's not uncommon for it to crash, and certain features don't work.

Running the serial pins at a very high baud rate would be enough to give you an extra four MIDI outputs. That may be enough for some musicians running a small setup, providing they're willing to use the Cubase crack.

I wonder if there are any other ways to pump medium-bandwidth data out of the MIST with guaranteed latency/jitter figures? I thought about the DMA audio as a possible data conduit, but it might be a bit tricky. If you have access to the analogue output from the DAC before it goes through any capacitors, I reckon you could encode four bits per sample per channel. That would be easily sufficient to drive 8 additional MIDI outputs. If the MIST does any resampling to the DMA audio (for instance, to feed a 48kHz AC97 DAC or something similar) then it becomes impossible. Also, any added audio buffering inside the core would render this unusable due to added latency.

So I'm looking at if I can use a USB to serial adaptor to then connect to the internet using a modem emulator, or building a MIDI to serial adaptor to do the same.

You could use a modem port MIDI interface device in reverse for this. There are some pitfalls though. Multi-output interfaces interpret certain data bytes as commands to select particular outputs, so you'd need a completely dumb single-output interface device. The other problem would be flow control, since the serial data is likely arriving faster than the MIDI port can handle - or vice-versa. The device attached to the modem port would need to implement hardware flow control.

If you can set the MIDI ports to run at a standard baud rate like 19.2k, all you need is a simple electrical interface. An Amiga MIDI adapter is all you need, but you will need to provide +/- 12 volts to it from an external source. You won't have any flow control though so the modem emulator might not be happy.

[AdvancedFollower]

Marketing the MiST is tough. It costs several times more than a Raspberry Pi 3+case, which is hard to justify for the "casual" retro gamer, especially as the Pi emulates more systems. FPGA is arguably more "authentic", but most people probably don't care, when they just want to mess around with a few childhood games one evening over a beer.

There is a market for this. Look at other expensive devices like the Framemeister, NeoSD, Super SD System 3, or any of the premium Everdrives. The EverDrive-GB is a $130 device that sits alongside cheap alternatives that cost 10% of that price.

With companies like Nintendo capitalizing on the current retro trend with their own ARM-based recreations, it's probably wise to keep a low profile (even though the MiST itself doesn't infringe on any copyrights or trademarks).

That is crazy! Look at what Analog is doing with the Analog NT Mini and the Super NT! High-end, FPGA devices that cost $500 and $200 respectively! The Super NT was released a few months after Nintendo's SNES Classic.

A low-profile is not required.

Of course if Lotharek had the economics of scale, the price could come down, and more units would be sold, so it's a bit of a catch 22.

More units could be sold if more people knew the units were out there!

Also, I happened to find this YouTube review of the Mist: https://www.youtube.com/watch?v=RW9b22D0nWY&t=3s

Its the only recent review of the MIST and the reviewer had a lot of problems, many of which I think were due to him not knowing the ins and outs of the MIST. Maybe someone here could reach out to him for suggestion solutions to the issues he was having  — mostly with the Atari ST and MSX.

Another person to reach out to would be The 8-Bit Guy — he has over half a million subscribers: https://www.youtube.com/channel/UC8uT9c ... u7ITLGo9Ww

This video (https://youtu.be/CVq_jzj_u8U) for example, has over 50,000 views. This did not result in 50,000 sales. People know about the MiST, they just can't justify buying one. Even though it costs €200, it's not a "premium" device. It doesn't come in a fancy box, isn't housed in a machined aluminium case, doesn't come with a bundled high-end controller. It's expensive because it's niche. It comes in a project case and requires you to hunt on Github for firmware and cores that may or may not have been updated in the last 2 years, then go to shady "grey area" sites to find ROMs and games.

Just achieving the kind of availability that e.g. the Raspberry Pi has would take a huge investment compared to the tiny batches the MiST was being manufactured in. The Pi foundation had expenses of £6,315,535 in 2015, but this has allowed them to create the necessary eco system to support the wide availability of the product. Even in my small home city, there are several stores that have boards, custom cases, heat-sinks, accessories etc. Achieving this broad availability isn't trivial. Marketing isn't as simple as sending a couple of units to some Youtubers (especially when, as you noticed, the product isn't super user-friendly so Youtubers give it a bad first impression). Marketing campaigns require a huge starting capital and investments in areas like manufacturing and logistics, market research, planning and strategy, presentation etc. It's the same problem companies like A-Eon face. They just don't have the starting capital to kick-start the market, so their products will forever be confined to their niche.

[Kujako]

This video (https://youtu.be/CVq_jzj_u8U) for example, has over 50,000 views. This did not result in 50,000 sales. People know about the MiST, they just can't justify buying one. Even though it costs €200, it's not a "premium" device. It doesn't come in a fancy box, isn't housed in a machined aluminium case, doesn't come with a bundled high-end controller. It's expensive because it's niche. It comes in a project case and requires you to hunt on Github for firmware and cores that may or may not have been updated in the last 2 years, then go to shady "grey area" sites to find ROMs and games.

On the last point, my plan is to use a KryoFlux to image real floppies. It is true that there is no Atari equivalent of Amiga Forever giving a legitimate way to source TOS.

[Eero Tamminen]

It is true that there is no Atari equivalent of Amiga Forever giving a legitimate way to source TOS.

You can just use EmuTOS, its sources are available under GPL license:
https://github.com/emutos/emutos

It has has a builtin HD driver, boots much faster, and its disk handling is in latest version faster than with "real" Atari TOS.

There are some dirty programs, mainly old floppy-only games & demos that don't work with it: https://github.com/emutos/emutos/blob/m ... atible.txt

But most of the Atari ST/STE/TT programs *do* work with it:
https://hg.tuxfamily.org/mercurialroot/ ... emutos.txt

(Also Falcon programs that don't need HiColor mode or DSP, work with it, but that's not relevant for MiST as it doesn't emulate Atari Falcon.)