MiST - FPGA ST

[ericgus]

I should have added the 715v is direct VGA connection.. so far i've yet to find a core that doesnt work .. {I have not tested every core though}

[Newsdee]

All cores works with my old Eizo screen in 50hz or 60hz, as long as the vertical refresh is kept at 31khz. But some cores are subject to the vertical artifacting several people reported (some lines of pixels are the wrong size). I can reduce it by playing with the screen controls but it needs to be tweaked per core. Only using a GBS or XRGB-mini gets rid of this problem.

[vebxenon]

Is there any news about this main core?

[vebxenon]

Any news? Timings on STe for example must be fixed. Some images on high res look scrambled.

Regards,

Salva

[Estrayk]

It seems MasterOfGizmo has abandoned the core? We will never be able to play Lethal Xcess.

[vebxenon]

It seems MasterOfGizmo has abandoned the core? We will never be able to play Lethal Xcess.

I was wondering the same some days ago

[vebxenon]

No news?

[jcarvalho]

sorry to ask but the atari st core development stopped?
That sucks, looks like the fpga thing is a group of unfinished emulation projects...
not even the 20 Years Atari STE Megademo works... I like atari and amiga but what i can say is the only decent core is the zx spectrum 128 one...
I finished with an zx spectrum very expensive in my hands...

[Newsdee]

The problem is not so much this core but the fact that nobody else is interested in following up. At least its open source and future enhancements are possible by somebody else.

[farvardin]

That sucks, looks like the fpga thing is a group of unfinished emulation projects...
not even the 20 Years Atari STE Megademo works... /.../ I finished with an zx spectrum very expensive in my hands...

Sorry but I don't agree at all. Most Atari ST games work very well. Many cores are working great too (Amiga, Amstrad, Nes, Atari 8bit, Sega Master System, and ZX for course etc)
So you have a great multi machine FPGA which can already do much. It's quite rough for the developers to tell them it's "unfinished", like a bad work. Of course there is always room for improvement, but many demos use undocumented features (or bug) so it's quite unlikely all of them will work one day.

[Gehstock]

If Everyone who complains about the Quality of the Cores himself some time in the Mist Community we are already much further.
What annoys me here is that nobody wants to invest some Time, everything should be free and of course perfect run (with desire features).
The Developer has no other life and nothing better to fulfill your Wishes. This is no fun anymore. ( Sorry for bad English written with Google Translator)

[jcarvalho]

If Everyone who complains about the Quality of the Cores himself some time in the Mist Community we are already much further.
What annoys me here is that nobody wants to invest some Time, everything should be free and of course perfect run (with desire features).
The Developer has no other life and nothing better to fulfill your Wishes. This is no fun anymore. ( Sorry for bad English written with Google Translator)

I never said that the cores should be free... It could be like a kickstart project. Donations to fund a heavy project... Could have "free" cores, with bugs and so but with funding present the developer can make efforts to get it working 100%. On zx spectrum128 core it is perfect and its free, if there was(or exists) a fund raise to help the author I will donate. who pays the Mist price for sure as 20 euros to donate to help developer and of course pay for the job. If I go to the garage to change the oil's car I pay. Why not pay for a core?

[Gehstock]

That is not was i mean, we dont need your Money but a little of your time.
Make Helpfiles, bring order in MainRepo or http://docs.google.com/spreadsheets/d/1P1LPndBDkcSSG9Cxco8NGMnihYyO-4FIvz1kcyZ7vi4
or make a Homepage

[jcarvalho]

That is not was i mean, we dont need your Money but a little of your time

I didn't want to be rude... If it is my time that the developers need, say it... I gave my time to the Portuguese linux Mint translation team for free (I started the team with Mr. Clem approval) so I can give my time to the core team... say what is needed and if it is on my skill level I will gladly give my effort... In my case I don't want any fancy features, just be able to run games and demos with mist.

[Gehstock]

So it should be but you are an Exception and Mist uses no Linux Mint.

(Mister is another construction site)

[Sorgelig]

Buying a MiST board you pays only for HW manufacturer. It's not a secret that HW manufacturer donates some boards for active developers and it's good help for those who spend a lot of time for open source projects.
But saying that some part of MiST price goes to core developers is not correct. Most developers are just doing it for free, for fun.

If you won't count the MiST board itself, then MiST project is simply open source project where anyone can participate and no one is obliged to do what he doesn't want to.
Thus complains about core quality is not correct or welcomed here.
Don't ask what MiST community should do to you, ask what you can do for community

Talking about paid cores is not correct as well because creating the core requires very precise knowledge about emulated hardware. In most cases there are no such info. Developing the core is like a science where you try to guess something or look for info by small pieces and try to solve the puzzle. And at the end you even don't know how precise it will be emulated. And if it will cost the money then only more angry people will come to complain and demand the quality for money.

[Sorgelig]

I think, the project model of MiSTer is more fair for all parties. It's clearly visible that developer(s) of the project gets no money from project. And the board is completely from 3rd party which is not bound to MiSTer project at all.
So, i hope there will be less complaining people who think they pay for core development.

[Newsdee]

What we need is more developers and people willing to learn. Unfortunately to go from learning HDL to write a core is quite a leap (except maybe if you have EE experience to rely on). There are other ways to help, though.

One bottleneck is undocumented features. When game X or demo Y doesn't work its usually for some reason. Looking for simpler test cases, and for example writing small programs to replicate the problem quickly and easilty can help developers test and fix the issue. Especialy if the FPGA dev is not familiar with the original machine.

Such programs would also help emulator developers so the benefits are not confined to the MiST.

[jcarvalho]

Buying a MiST board you pays only for HW manufacturer. It's not a secret that HW manufacturer donates some boards for active developers and it's good help for those who spend a lot of time for open source projects.
But saying that some part of MiST price goes to core developers is not correct. Most developers are just doing it for free, for fun.

If you won't count the MiST board itself, then MiST project is simply open source project where anyone can participate and no one is obliged to do what he doesn't want to.
Thus complains about core quality is not correct or welcomed here.
Don't ask what MiST community should do to you, ask what you can do for community

Talking about paid cores is not correct as well because creating the core requires very precise knowledge about emulated hardware. In most cases there are no such info. Developing the core is like a science where you try to guess something or look for info by small pieces and try to solve the puzzle. And at the end you even don't know how precise it will be emulated. And if it will cost the money then only more angry people will come to complain and demand the quality for money.

But saying that some part of MiST price goes to core developers is not correct. -> I never said that
Paid cores could be by funding system, sure that developers have to pay the bills, they need to put food on the table and so on. What I was thinking is that money could be used to help developers to dedicate more time in development, like full-time job on it or to ask for reverse engineering of some funky chip to reproduce it on fpga, a SID chip or something else. The motivation is higher if we got payment. I develop web services in php, mysql, sql server (web dev.) Sure nothing goes 100% at first try, but the motivation is the paycheck on the end of the month AND the feeling of a job done. I dont start a project before finishing the first one, so core devs could go the same way... And put food on the table with funds on those kickstart websites

[Sorgelig]

It's up to core developer. If there is developer with a lot of knowledge and he can create a perfect core, let's say much better Amiga core, but the money is the only issue to accomplish this then he is free open a kickstart project.
In most cases the money is not a main issue.

[ericgus]

I think people offer to contribute money because its really the only way they **CAN** contribute .. perhaps they lack technical skills (if everyone could code an FPGA core we would be drowning in them) .. and other people lack the time to do debugging .. but yet they still **WANT** to contribute.. I don't see a problem with a Pateron or something **if** the developers are ok with that.. again yes its totally up to the core developers.. but please don't dismiss people wanting to contribute financially to core development if thats the only means they can actually be of help.

[Sorgelig]

I've already explained that main problem is NOT money. If you will run around offering the money to everyone, then someone will eventually get your money and will go away. Money will help if there is a demand of money at the first place.
If you think you can solve everything by money, then you are wrong.

[Newsdee]

Look at all those kickstarters that successfully raised money and then... went nowhere. There's a big share of disasters out there. I think there was even one case where the guy ran away with the money!

You need to bear in mind that even if you had enough money to pay somebody full time for a year to work on things, it might not be enough to get the core perfect unless they invest a lot of time investigating and documenting the original hardware. And it's easy, even with the best intentions, to spend time on the wrong thing and fail to make progress. Personally I think it's best to keep things casual without having to worry about dissapointing backers.

There is one guy with a Patreon looking at NeoGeo though, you can check it out but I have no clue how far ahead the project is:
https://www.patreon.com/furrtek

[mfro]

I totally agree that money isn't the problem. Good HDL developers can gain safe and high-salaried jobs in the industry that desperately searches for such rare breed.

You'd need to find somebody that has the skills and is crazy (or let's say "dedicated") enough to abandon this in favour of a hobby project with dubious future. Unless you'd be able to find somebody like this, you'll probably need to stick with us humble spare-time amateurs (not meant to insult anybody in particular other than myself ).

[Newsdee]

For those that want to help but have no knowledge of HDL, I can suggest testing and documenting the results.

I've tried every single ROM for NES that I have so I can now tell to any interested developer which games to look at, or which features would need to be added:
https://docs.google.com/spreadsheets/d/ ... p=drivesdk

There is no guarantee somebody will pick it up, but at least we can show clearly how compatible the core is and what things aren't working.