PC-98 is it's own hardware entirely.

Yeah, but look at my link, the Windows thing is quite confusing, did they make windows95 (or even 3.1) versions for pc98 hardware? I erroneously thought that being the software made for Windows 95, it would be compatible with all Windows 95 devices.

They did. There are PC-98 specific versions of Windows 3.0, 3.1, 95, and 98. There are also PC-98 specific versions of Windows NT, apparently.

The game Magical Squadron (discussion about this game is linked in one of my previous posts) is a pc98 dos game afterall (it has got a Windows installer and all, you can even install it without errors in a standard Windows 95-98-me machine but it will hang (black screen) when running the game.

It has got a dos installer too, but for some reason only minimum install works. The game seems to run fairly good in dosbox-x pc98 emulation though.

TheGreatCodeholio wrote:

They did. There are PC-98 specific versions of Windows 3.0, 3.1, 95, and 98. There are also PC-98 specific versions of Windows NT, apparently.

Up to Windows 2000 PC98.

SA1988 wrote:

TheGreatCodeholio wrote:

They did. There are PC-98 specific versions of Windows 3.0, 3.1, 95, and 98. There are also PC-98 specific versions of Windows NT, apparently.

Up to Windows 2000 PC98.

WindowsNT4.0 too.
https://sites.google.com/site/np21win/setup/win2000
https://sites.google.com/site/np21win/setup/winnt4-0

danrevella wrote:

TheGreatCodeholio wrote:

Are you saying that VGA state is not being saved/restored properly?

http://www.ctyme.com/intr/rb-0222.htm

yes, so many cheat utility like game buster and Infinity machine are unable in working...
when you return to the game the screen is black.

Ehm..
Maybe you have an answer for me?

AnnaWu wrote:

WindowsNT4.0 too. https://sites.google.com/site/np21win/setup/win2000 https://sites.google.com/site/np21win/setup/winnt4-0 […]
Show full quote

SA1988 wrote:

TheGreatCodeholio wrote:

They did. There are PC-98 specific versions of Windows 3.0, 3.1, 95, and 98. There are also PC-98 specific versions of Windows NT, apparently.

Up to Windows 2000 PC98.

WindowsNT4.0 too.
https://sites.google.com/site/np21win/setup/win2000
https://sites.google.com/site/np21win/setup/winnt4-0

Up to Windows 2000 PC98 includes NT 4.0 for PC98 😀

Can someone please comment on this thread at DosBox-X regarding XP support:

https://github.com/joncampbell123/dosbox-x/issues/403

Thanks

The development tools I have make it difficult for me to support Windows XP. Specifically, Visual Studio 2017 has a Windows XP target but certain key C library functions like stat() are broken in a way that prevents them from working on Windows XP (but work fine with Vista and higher). Apparently this has been a known issue since VS2015.

I state that I personally won't support Windows XP, but patches from others who can compile for XP are welcome, because if they don't, continued development may break Windows XP support.

TheGreatCodeholio wrote:

The development tools I have make it difficult for me to support Windows XP. Specifically, Visual Studio 2017 has a Windows XP target but certain key C library functions like stat() are broken in a way that prevents them from working on Windows XP (but work fine with Vista and higher). Apparently this has been a known issue since VS2015.

I state that I personally won't support Windows XP, but patches from others who can compile for XP are welcome, because if they don't, continued development may break Windows XP support.

First of all, you do fantastic work, kudos for everything...

Don't take this the wrong way please, I know it is a privilege to use your build and I don't want you to think I'm self-entitled, but what are your thoughts on the last comment by aybe in that thread?

If you are able to follow his instructions to compile for XP, then go for it.

If anything I do breaks VS2013 or Windows XP, let me know and submit a patch to correct it.

I don't have enough free time to check everything against Linux, Windows XP and WIndows 7/8/10 all at once.

Thanks for the great effort toward dosbox-x development.

However Windows 98 runs very slow on dosbox-x pulled from github on 2018-01-31, although I have set core=dynamic
I got this message when I run make:
checking whether x86 dynamic cpu core will be enabled... no

is that what makes windows 98 runs slow in my machine?

Yes. I had to remove the dynamic core code, because it has slowly bitrotted over the years. It's also not compatible with the way emulation and page fault handling works now. Even if you did manage to run Windows 98 in DOSBox-X today, the dynamic core would mishandle page faults in a way that would eventually cause WIndows 98 to crash.

TheGreatCodeholio wrote:

Yes. I had to remove the dynamic core code, because it has slowly bitrotted over the years. It's also not compatible with the way emulation and page fault handling works now. Even if you did manage to run Windows 98 in DOSBox-X today, the dynamic core would mishandle page faults in a way that would eventually cause WIndows 98 to crash.

Thanks for the quick reply. Would you please point me to a version which has dynamic core available? Also is the removal of dynamic core the cause of slowness in Windows 98SE?

I removed the dynamic core code on Dec 28th, so the last build with it still in 32-bit builds is on this page of the releases, the Dec 25h Christmas release:

https://github.com/joncampbell123/dosbox-x/re … sbox-x-20171228

Yes, Windows 98 runs slower without the dynamic core.

As I said though, the dynamic core doesn't quite handle page faults the way they need to be handled. Windows 95 and higher require the emulation to support task switching even during a page fault, which dynamic core (and anything other than normal) are not designed to handle.

TheGreatCodeholio wrote:

Yes, Windows 98 runs slower without the dynamic core.

As I said though, the dynamic core doesn't quite handle page faults the way they need to be handled. Windows 95 and higher require the emulation to support task switching even during a page fault, which dynamic core (and anything other than normal) are not designed to handle.

Is this correct that dynamic core never exist on x64 build? I'm using x64 build of DOSBox-X right now, thinking about revert to earlier build...

Dominus wrote:

There is no reason. It's just that people assume 64bit is better on a 64bit system. And on linux it's a bit more of a hassle to build 32bit when the system is 64bit.
Dosbox SVN has a working 64bit core but it's still not up to 32bit dynamic core, having problems with some games and still being a lot slower on OS X.
Dosbox-x forked before that improvement and just doesn't allow dynamic core on 64bit (AFAIK).

The dynamic core was 32-bit only. 64-bit never had it. Since the change, neither 32-bit nor 64-bit have it now.

Weird ass behavior for super street fighter 2 turbo in latest dosbox-x. Installer detects hard drive space as zero and refuses to continue. I managed to install the game with an older build from 2016. But when running the game, at the "initializing cd-rom please wait" message, the cdrom drive spits out the disc and dosbox-x hangs. I'll try to make bin+cue of the original disc and retry with imgmount instead of mount command.