Once the next release is out of the way, I'll deal with the tracks.

An XMS request to himem. Sys can be at most 65535 Kb in size. I don't think icr2 goes to the trouble of making multiple requests and manually managing multiple, possibly discrete, blocks. The request size in KB is a 16 bit unsigned value.

At least that's how it's implemented in dosbox.

[EDIT] it's in the spec
https://www.phatcode.net/res/219/files/xms30.txt

The size in KB is passed in a 16 bit register.

**************************************************************
RReady Alpha, January 7, 2025, Build #001 (Full tested stable build)
**************************************************************
https://nirvtek.com/downloads/RReady.Alpha.20250107.001.7z
MD5: 6ef91529aa5bf287b88fd5179f36c489
- Support for a full 16 MB of VRAM for Speedy3D apps. This is the upper limit for Speedy3D (24 bit).
For Speedy3D apps, VRAM is configured in dosbox-rendition.conf (or ...-staging.conf) with
[rendition]
MaxVRAM =nn (nn is in MB, range 4-16, default 16 if setting missing)
For RRedline MaxVRAM=nn can be set in rendition.cfg and it defaults to 128 MB
- vQuake beta 1.06 is once again playable and joins betas 1.07 and 1.08* (*recommended beta)
- Higher resolutions may be selectable in vQuake 1, but currently have issues. This will be resolved in a future release.
- Bilinear filtering can be disabled by adding nofilter=1 to rendition.cfg. [This should work correctly now]. Some games might not render correctly with this setting.
- vHexen II additional resolutions can be unlocked by selecting the game in the WinApps tab of RLauncher and clicking "App Settings."->Unlock resolutions. There may be rendering bugs with higher resolutions. This will be fixed in a future update.
- Regression ucube.exe (SDK demo) fails to render correctly in full screen mode.
- Corrects OpenGL GLSL shader "error C7011: implicit cast from "uint" to "int'" error on some hardware.
- 60 fps Frame Generation/interpolation supported for Grand Prix Legends and "Die by the Sword." Note GPL will have to be run at default resolution/4:3 (fill screen aspect) aspect as the game has a mouse lock. Higher resolutions can be selected from within the game, Your GPU's internal aspect ratio controls from its control panel should be used for aspect ratio selection. "Die by the Sword" can be upscaled without issue.
- Fixes Eurofigher 2000+ texture corruption when switching between menu and active mission. This was caused by the game writing to 8 MB of VRAM and choosing textures within the 4 MB range. Rounding write addresses modulo 4 MB, effectively restricting writes to the first 4 MB for this game, corrects the issue.
- Whiplash minor fixes.
This update should be available in the Microsoft Store within the next 3 days.
**************************************************************
[EDIT] ICR2 antialiasing and line drawing is disabled in this build. There should be no need to hit [ALT+A] to fix AA issues.

Crappy mobile captures:
https://youtu.be/fBs9ZIbHIG0

V1000 vs V2000 vs RReady running 1982 Cart.

Raven-05, we need you to post a decent vid.

I think I've fixed my RRedline problems on the 3D Blaster PCI. I had the machine connected to the network. Disconnected that and it's stable. There is one problem though, the monitor goes to "no signal" for around 30 seconds at Windows login. It might be my VGA to HDMI adapter, but the V2000 has no problems with that.

I tested V_EnumDisplayModes on both the V1000 and the V2000. The V1000 doesn't support the display1 extension, so it fails. The V2000 does return a full set of results:
320x200/8@0
320x240/8@0
512x384/8@0
640x400/8@0
640x480/8@0
800x600/8@0
1024x768/8@0
1152x864/8@0
1280x1024/8@0
1600x1200/8@0
320x200/16@0
320x240/16@0
512x384/16@0
640x400/16@0
640x480/16@0
800x600/16@0
1024x768/16@0
1152x864/16@0
1280x1024/16@0
320x200/32@0
320x240/32@0
512x384/32@0
640x400/32@0
640x480/32@0
800x600/32@0
1024x768/32@0

The call takes a structure with 4 params (x-res, y-res, bit depth and refresh rate) which can restrict the resolutions returned. vHexen II sets bit depth to 16 and everything else 0. When I use that with the call, it returns nothing and then probably queries display modes directly using the Windows API calls.

RReady can't do that because modern Windows machines return a shedload of resolutions and the game crashes. Disabling the display1 extension (by adding nodisplay=1 to rendition.cfg) causes the game to crash at start up. So I don't know whether it's possible to support higher resolutions than 1152x864 (or 1280x800 as Raven-05 pointed out).

Also on the V2000 (4 MB) 640x480 is the highest available resolution.

RReady Alpha, January 7, 2025, Build #002 (Untested build for ICR2)
Updated build:

https://nirvtek.com/downloads/RReady.Alpha.20250107.002.7z
MD5: d975cfc9f583d10b1de9a121e514a194

- Fixes all issues raise by @gplaps for ICR2 (two of them I'm not sure, but they seem fixed)
- Massive performance improvement in ICR2 for mobile/integrated GPUs. Achieved by disabling Edge AA. I disabled only Line AA and not edge AA.
- supports nofog=1 and nofilter=1 fully.
- This won't be on the store for at least a weeek.

[edit] even runs on my Intel uhd 620 based i3 7xxx dual core+ht laptop. It's never rendered more than a single frame before crashing.

shouldn't the config be named this way?
dosbox-staging-Rendition.conf

i got other idea

nofilter=-1

which experimental and forces every graphic to be filtered , even plain 2d , menus etc

robertmo3 wrote on 2025-01-07, 14:45:

shouldn't the config be named this way?
dosbox-staging-Rendition.conf

It should. I should change that. Thanks for point that out.

RaVeN-05 wrote on 2025-01-07, 14:56:

i got other idea

nofilter=-1

which experimental and forces every graphic to be filtered , even plain 2d , menus etc

Could do that as well. Would be fun to do a v1k_drawflat as well 😁

why not just
filter=[0,1,2]
with 2 being what raven said
also
fog=[0,1]

@robertmo3, what do you make of this:
https://www.icr2.net/forum/showthread.php?tid … =14261#pid14261

robertmo3 wrote on 2025-01-07, 15:22:

why not just filter=[0,1,2] with 2 being what raven said also fog=[0,1] […]
Show full quote

why not just
filter=[0,1,2]
with 2 being what raven said
also
fog=[0,1]

That makes sense. What's happening to dosbox-staging over at icr2.net?

I'll add the forced filtering support in the next next release.

i noticed one joystick too.
what i did before it started to work:
- i visited Devices and Printers to check both are there
- i tried 4axis_2 to see second one is detected
- than i tried 2axis and both were detected
- auto started to detect both now

btw there are many fixes for joysticks in:
0.81.1

robertmo3 wrote on 2025-01-07, 16:48:

i noticed one joystick too. what i did before it started to work: - i visited Devices and Printers to check both are there - i t […]
Show full quote

i noticed one joystick too.
what i did before it started to work:
- i visited Devices and Printers to check both are there
- i tried 4axis_2 to see second one is detected
- than i tried 2axis and both were detected
- auto started to detect both now

btw there are many fixes for joysticks in:
0.81.1

I'll mirror port those changes across as soon as possible.

https://nirvtek.com/downloads/RReady.Alpha.20250108.001.7z
MD5: 87c8bf0af4e331026c8e9716618ca482

- Dosbox-staging should use dosbox-staging-rendition.conf now.

Dosclient. Exe is being flagged as malware by Windows Defender and my false positive reports are being ignored. This applies to the store based version. As well.

The only thing I can think of is to get as many people as possible to report it to Microsoft.

https://www.microsoft.com/en-us/wdsi/filesubmission

Nofilter and nofog finally working correctly with ICR2:

https://youtu.be/OXvHB-hB_ko

Components of RReady are being flagged as malware by Windows Defender (and nothing else): Trojan:Win32/Wacatac.B!ml or Win32/Formbook!m. These are really dangerous forms of malware . Over the past year I've had every single component of RReady being randomly flagged as malware including at one point a text file. At the moment Dosclient is being flagged.

MS ignore reports unless there're a significant number of users complaining. if you would like to help, the link to mark an incorrect detection is over here:

https://www.microsoft.com/en-us/wdsi/filesubmission

And one more boring old vQuake video with d_mipscale 0 and a_mip 0. A playthrough of the first 2 (or 3) levels:

https://youtu.be/PXfvbG0FBQA

There is a quite of bit of slowdown, but before with only 4 MB of VRAM it was unplayable.

There will be a patch for dosbox-x and dosbox-daum at some point, but not right at the moment.