FX CAST Atari ST core

tatouine · Post by **tatouine** » Mon Dec 31, 2018 3:52 pm

Thank you and Happy New Year!

cacophony · Post by **cacophony** » Mon Dec 31, 2018 6:47 pm

Hey Ijor, any update on when you expect to release the source code for FXCAST so that it can be added to MiSTer-devel ?

nico24 · Post by **nico24** » Tue Jan 01, 2019 4:20 am

I have to say through VGA and separate sound output this core is VERY good. It runs better than the version on the MiST, from a video output perspective. The only thing missing is support for hard disks which that one has. This core actually makes me like an ST, which I don't really have with my real hardware.

DanyPPC · Post by **DanyPPC** » Tue Jan 01, 2019 10:51 am

You have to try the Component YpBpr Out on MiST, 50Hz well syncronized and good compatibility with STE games.

BBond007 · Post by **BBond007** » Wed Jan 02, 2019 5:08 am

Ijor,

Thank you so much for the MIDI feature.

This is how it sounds with MUNT running on the ARM HPS for people who don't own MIDI hardware.

Conquests of Camelot (hybrid sound) --> http://y2u.be/cM3YELrRRsY

Only inexpensive USB sound dongle is required (for now).

Hopefully writing to the floppy is soon

some of these Sierra games won't continue if the disk is write-protected.

mattsoft · Post by **mattsoft** » Wed Jan 02, 2019 5:58 am

Been trying out (and enjoying) many Atari ST demo scene disks. So far, everything just works. Good stuff. Thanks Ijor!

SaschaFFM · Post by **SaschaFFM** » Fri Jan 11, 2019 2:36 pm

@Ijor

Do you already have a feeling how much work it takes to get 15khz-VGA output to run? It would be so great if I could enjoy the ST on my CRT.

Thanks again for this great core!
Sascha

Shamus · Post by **Shamus** » Fri Jan 11, 2019 5:08 pm

It would be so great if I could enjoy the ST on my CRT.

... and please keep in mind DVI ... (I am still eager to try FX CAST

)

matt · Post by **matt** » Mon Jan 21, 2019 8:15 am

Hi Ijor,

I was wondering, is there any way for this core to influence software 68000 emulation? Not so much automatic conversion form the definition to C but some of the details that software 68000 currently lack.

Post by **ijor** » Mon Jan 21, 2019 1:23 pm

matt wrote:I was wondering, is there any way for this core to influence software 68000 emulation? Not so much automatic conversion form the definition to C but some of the details that software 68000 currently lack.

Software emulators are quite mature and pretty good. I emailed Toni Wilen about this some time ago. He is the WinUAE maintainer, and the WinUAE 68K emulator is also used in Hatari. According to him the main weakness is interrupt timing. Although I suspect there are other issues, especially in some obscure areas such as some minor details of exceptions, that WinUAE is probably not cycle accurate yet.

Interrupt timing is a bit of a PITA to implement on software emulators because it depends on the specific instruction microcode. So there is no straightforward method to implement for emulators, unless they emulate microcode of course, but usually they don't.

At this time Toni is much more interested in improving 68020 emulation. From the point of view of the Amiga, it is of course, very understandable.

I also emailed the Genesis Exodus emulator. I actually emailed him for asking him about some details about the Genesis 68K interface. In reply he told me he also would be very interested in improving his 68K emulator, but after that initial reply I didn't hear back from him.

Edit: corrected couple of typos

matt · Post by **matt** » Mon Jan 21, 2019 9:14 pm

Thanks, that's interesting. I'll do some reading!

BlastEm is the latest fast emulator using a 68K core that I've seen, it does not use WinUAE

https://www.retrodev.com/blastem/

laxer3a · Post by **laxer3a** » Thu Jan 24, 2019 4:05 pm

Hi !

First of all, congratulation on your work on the FX CAST cores.
This is a really impressive amount of work.

I must say that I hesitated to write this, as I do not want to look pushy or anything...

I'd like to know if you have any ETA about release the sources of the FX CAST ?
As you can probably see, a lot of changes happen on a daily basis by putting the core source open.
(I know from hands-on experience that such project are exhausting and after a release you want to relax for a while, do not touch the code, swap to another funny project, etc... May be you feel your source could be better, that you want to refactor something or document it before release the baby in the wild.)

I've been lurking for 2 months now and seen no changes around about the code release... but I am afraid that now people have your binary and get satisfied with that (well, it runs floppy images so).
There are probably a lot of small features that people would like to see (ie mapping SDCard to HDD, better continuous integration with the MiSTer environment as the open cores are doing, may be some people adding the STe feature ?).

Happy new year anyway if it is not too late

Regards,
Romain

bitfan2011 · Post by **bitfan2011** » Thu Jan 24, 2019 5:53 pm

laxer3a wrote:Hi !

First of all, congratulation on your work on the FX CAST cores.
This is a really impressive amount of work.

I must say that I hesitated to write this, as I do not want to look pushy or anything...

I'd like to know if you have any ETA about release the sources of the FX CAST ?
As you can probably see, a lot of changes happen on a daily basis by putting the core source open.
(I know from hands-on experience that such project are exhausting and after a release you want to relax for a while, do not touch the code, swap to another funny project, etc... May be you feel your source could be better, that you want to refactor something or document it before release the baby in the wild.)

I've been lurking for 2 months now and seen no changes around about the code release... but I am afraid that now people have your binary and get satisfied with that (well, it runs floppy images so).
There are probably a lot of small features that people would like to see (ie mapping SDCard to HDD, better continuous integration with the MiSTer environment as the open cores are doing, may be some people adding the STe feature ?).

Happy new year anyway if it is not too late

Regards,
Romain

i'd love to see VGA output options for native rez, filters, etc..

djmartins · Post by **djmartins** » Thu Jan 24, 2019 10:15 pm

bitfan2011 wrote:I must say that I hesitated to write this, as I do not want to look pushy or anything...

I'd like to know if you have any ETA about release the sources of the FX CAST ?

Must be tough to release one's sources knowing that a lot of people will be pawing through them but
we have seen a lot of good stuff come out of open source projects.
It IS impressive work either way, but getting it in line with the other MiSTer cores would be great!

Post by **ijor** » Fri Jan 25, 2019 12:15 am

laxer3a wrote:I'd like to know if you have any ETA about release the sources of the FX CAST ?

Sorry, not ready yet and I would prefer not to commit an ETA. But I have no plan to keep this closed for very long.

There aren't too many news because I hadn't have enough time lately to dedicate to my "retro" projects. But I certainly don't feel "satisfied" with the core as it is now. There are actually several features that are already implemented but they were disabled on the public release because they had some issues, or because I couldn't test them well enough. Hard disk support and selectable TOS are two examples of these features, as many others that I plan to implement as soon as I have some time. This have no relation with releasing the source code.

I must say that I hesitated to write this, as I do not want to look pushy or anything...

Don't ever hesitate to write anything that you mean in good faith

You didn't sound pushy at all.

laxer3a · Post by **laxer3a** » Sun Jan 27, 2019 9:26 am

Hi ! Thanks for your kind reply.

I know how it is, given that I have done some HDL programming myself. And have like 10 things sitting on my HDD that I should have put on GitHub and are not, for various C/C++ projects (from emulator work,script languages to libraries).

(even I do not have any plan to keep them for me)
So I am not even following the advice I am giving

It can always be annoying and a kind of mental workload to think about going to GitHub, prepare everything and then release the thing.
But honestly, as a whole, take 10 mn, just commit the whole thing even as it is, push the folder and be done with it for now, is something that is most likely the best thing to do... As an example, I don't think that people will try to appropriate your work anyway and given the size of the community, everybody knows who did what

. Or that people will look at your stuff and say it is badly done (honnestly, how many people look at the HDL sources in detail ?

. It is a huge workload to look at those things in detail, and people with no intent of really helping do not go that far ) But as I said, I respect your choices, and do not want to be pushy in any way.

People may pick it and do work for it, may be not, but at least it is backed up in the wild, one more step secured for the future. (which are already 2 big advantages vs sitting on an HDD somewhere, and I hope you have backups or a private repositery at least, or better, have already some people with access to it too in any case. I do not remember if you wrote about this, sorry)
Moreover, even if people do not touch the HDL or core, most likely all the things around (integration, system features, etc...) are going to be maintained automagically by Alexey and tested by people around.

Torlus (the guy who made the original Megadrive and PC Engine project, met him IRL) committed the sources code of his project like nearly 10 years ago now, and did not work on it anymore, and it was picked up only like 2/3 years ago by others, and that is thanks to putting things in the open.
In my case I put a Snes PPU core open around 2008, and got only 3 or 4 mails in 5 years from univ. student but thats about it.
But I have a feeling the scene is getting a LOT of traction since MiSTer is here more than any platform ever. (I had 6 different FPGA board until now, and getting good enough specs to things like an ST or a Snes was talking about 1200$ dev board at least, but today for 120$, anybody can consider buying one, even if it means not touching it in 6 month and I feel a lot of tinkerer are joining the club)

In my case, on MiSTer side, I am very busy at work and I do not think I can help any project until this summer most likely...
(I planned to look at the system, document things for others and may be improve things if I could. Would be nice to help on a few cores (I know the Snes well, I may help there... Or look at the STe, that could be cool too))

Anyway, I am waiting on the side line until the day I can probably help (and some people around probably feel the same), hope to see your updates soon, and wish you to have the motivation to continue, even if you are busy now.

Romain

Sorgelig · Post by **Sorgelig** » Sun Jan 27, 2019 10:52 am

If Torlus would wait for perfection before share the sources, such cores as Genesis and PCE wouldn't exist today. And both have perfect compatibility now.

vanfanel · Post by **vanfanel** » Sun Jan 27, 2019 1:19 pm

Sorgelig wrote:If Torlus would wait for perfection before share the sources, such cores as Genesis and PCE wouldn't exist today. And both have perfect compatibility now.

True thing, yes. This how it happened.

laxer3a · Post by **laxer3a** » Sun Jan 27, 2019 1:36 pm

Sorgelig wrote:If Torlus would wait for perfection before share the sources, such cores as Genesis and PCE wouldn't exist today. And both have perfect compatibility now.

Yes, that's the whole point of my post. Just push things. And don't care much about how much code has to be, or features...
By having it in the wild, it can *only* evolve faster than being closed.

When I met Torlus, we discussed about this, he told me that his sources have been in a "sleep mode" for quite a long time until some other motivated people picked it up and asking usefull question about the source, instead of saying "hey, that games does not work, plz fix this or fix that" and took over the baton for the relay.
Putting things available to other (especially given the nature of those kind of project) is a huge multiplier.

But for that to happen in the first place, things have to be accessible.

I already said it on the Facebook MiSTer group, but again, thanks to you so much Sorgelig, you are a real MonSTer

/ workhorse for this platform and it would not be there today, so quickly without your work.

PurpleMelbourne · Post by **PurpleMelbourne** » Mon Jan 28, 2019 1:06 pm

ijor wrote:
laxer3a wrote: I must say that I hesitated to write this, as I do not want to look pushy or anything...
Don't ever hesitate to write anything that you mean in good faith You didn't sound pushy at all.

I have to disagree. Speaking from experience, hesitation is warranted around here!

However there is a lot of talent and good stuff come out of the place too. So it is worth persisting, and great if you get a good reaction

paulrobinsontkd · Post by **paulrobinsontkd** » Thu Jan 31, 2019 8:08 am

Hi,

Have just stumbled across this and very excited about the future possibilities and what is to come from this core.

I have yet to buy a Mister, but will be placing an order over the next few days with it's primary use for this core.

I however have a few questions

1.) Do you require an SDRAM module to use this?
2.) Do you require an I/O board? is it recommended to buy one anyway?
There also seems to be quite a few different boards about which is recommended (5.5 XS / XL /AIO?)
3.) Any advice where to order from the UK

Thanks

Paul

PiddlyD · Post by **PiddlyD** » Thu Jan 31, 2019 3:26 pm

Paul, I bought my DE10 from Amazon. I bought my Super Expansion from <removed>... but I'm in Arizona and so are they, so that was a no-brainer. I'm sure someone makes an expansion board with similar features in Europe.

You can start with just the DE10 - but to do most of the cores you probably want, you need memory expansion. The thing I liked about <removed> board was that it had memory, RTC, VGA, an extra SD slot, audio out and a fan and heatsink for a reasonable price.

Sorgelig · Post by **Sorgelig** » Thu Jan 31, 2019 3:42 pm

Stop to advertise the board which is NOT supported by MiSTer!
There is I/O board which is fully supported and available from different sellers.

Post by **ijor** » Thu Jan 31, 2019 6:23 pm

paulrobinsontkd wrote:
1.) Do you require an SDRAM module to use this?

Yes.

2.) Do you require an I/O board? is it recommended to buy one anyway?

It is not required at all.

bitfan2011 · Post by **bitfan2011** » Thu Jan 31, 2019 7:03 pm

paulrobinsontkd wrote:Hi,

Have just stumbled across this and very excited about the future possibilities and what is to come from this core.

I have yet to buy a Mister, but will be placing an order over the next few days with it's primary use for this core.

I however have a few questions

1.) Do you require an SDRAM module to use this?
2.) Do you require an I/O board? is it recommended to buy one anyway?
There also seems to be quite a few different boards about which is recommended (5.5 XS / XL /AIO?)
3.) Any advice where to order from the UK

Thanks

Paul

Buy everything you can that looks well-made and comes from sellers listed here. If you can afford it, I would recommend getting two sets of MiSTer hardware going, as there are various case options which are distinctly different now. Plus, it will be easy to sell the backup set as people learn about FPGA. I try to at least buy (2) of everything in case (1) is DOA. Then you don't have to be out of luck and waiting on shipping another.

You definitely need:
DE10 mainboard
USB hub
SDRAM module
64 or 128GB high-quality micro SD (DE10 comes with a small one)
Case or board to mount parts (USB connection is very fragile)

If you really want to get into 240p/15KHz oldschool gaming, then you want the full I/O board for the VGA port. You'll need a CRT display and the desired cable type to connect it. You'll also need the secondary SD slot for the a800 and a5200 cores, but maybe that has changed?
I have the native 240p going into an OSSC so I can easily see all the signal info during this phase of MiSTer development, but an external scaler/multiplier is probably overkill. I just don't have the needed CRT in the same area as everything else and I still want side-by-side HDMI vs. VGA for testing everything.

Check out the case/hardware threads to see what people are using.

Eventually, I will be using the full stack of official boards, but we're still waiting on the LL Cool Joy board to see where that's going to fit in?

Atari-Forum

FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core

Re: FX CAST Atari ST core