VOGONS


The quest for the perfect retro laptop: a saga

Topic actions

Reply 1160 of 1162, by Dodobird

User metadata
Rank Newbie
Rank
Newbie
wierd_w wrote on 2026-04-23, 07:19:
There IS a driver for that, but I have not tested it. https://www.bretjohnson.us/ […]
Show full quote
megatron-uk wrote on 2026-04-23, 07:10:

Joystick is probably one of the biggest sticking points.

Some (read: very few) older laptops do have a combined joystick/midi port, and of course with those you can use a real joystick. With usb-only laptops you may be able to get one of the usb enablers to detect the chip and something like Bret Johnsons USB tools which can map USB controllers to keyboard input.

But there's no way to use an analogue usb joystick with dos, natively, that I am aware of.

There IS a driver for that, but I have not tested it.
https://www.bretjohnson.us/

Naturally, a "Joystick port" joystick would need a HID class adapter to physically be connectable via USB.
I remember looking to see if that X401U could work with his USB HID Storage class driver, and found that it was not usable, since this thing is so budget, they cut USB1.1 hardware support. It only has OHCI and EHCI, not UHCI. (I wanted it for troubleshooting reasons. I have unusual mouse input if using the "Microsoft Mouse" driver for win3x--this is svardos afterall-- but found that switching it to the Logitech one fixes everything. The builtin legacy peripherals bios routines work fine knowing this caveat.)

So, I guess it has 2 strikes against it. It was exceedingly rare to even *FIND* a gameport joystick port on a laptop from the days of yore anyway though.

Coming late to the game, have you checked this YouTube channel RETROBECANES? It reviews a bunch of laptops with game port and even with wave table audio. https://youtu.be/tRU7DBwV7yg?is=vcyTieqKnN-5BBEy

Reply 1161 of 1162, by wierd_w

User metadata
Rank Oldbie
Rank
Oldbie
Dodobird wrote on 2026-09-10, 09:03:
wierd_w wrote on 2026-04-23, 07:19:
There IS a driver for that, but I have not tested it. https://www.bretjohnson.us/ […]
Show full quote
megatron-uk wrote on 2026-04-23, 07:10:

Joystick is probably one of the biggest sticking points.

Some (read: very few) older laptops do have a combined joystick/midi port, and of course with those you can use a real joystick. With usb-only laptops you may be able to get one of the usb enablers to detect the chip and something like Bret Johnsons USB tools which can map USB controllers to keyboard input.

But there's no way to use an analogue usb joystick with dos, natively, that I am aware of.

There IS a driver for that, but I have not tested it.
https://www.bretjohnson.us/

Naturally, a "Joystick port" joystick would need a HID class adapter to physically be connectable via USB.
I remember looking to see if that X401U could work with his USB HID Storage class driver, and found that it was not usable, since this thing is so budget, they cut USB1.1 hardware support. It only has OHCI and EHCI, not UHCI. (I wanted it for troubleshooting reasons. I have unusual mouse input if using the "Microsoft Mouse" driver for win3x--this is svardos afterall-- but found that switching it to the Logitech one fixes everything. The builtin legacy peripherals bios routines work fine knowing this caveat.)

So, I guess it has 2 strikes against it. It was exceedingly rare to even *FIND* a gameport joystick port on a laptop from the days of yore anyway though.

Coming late to the game, have you checked this YouTube channel RETROBECANES? It reviews a bunch of laptops with game port and even with wave table audio. https://youtu.be/tRU7DBwV7yg?is=vcyTieqKnN-5BBEy

I am more interested in what is "readily available."

there is no shortage of 'budget' laptops from the early to mid win7 epoch. These are reasonably powerful, and generally have Intel HDA audio. With the advent of VSBHDA / SBEMU, we can have a way to have soundblaster audio emulation on such systems. These are very cheap, and can often be found for 50$ to 70$ on ebay.

They also *MUST* have a CSM, since windows 7 requires one (under most ordinary conditions. there are ways to sidestep it, but they are painful.)
That means they can run DOS, and will have 'some semblance' of a VESA bios present. (However, the 'quality' of this VESA bios is often... problematic. Mine is "Sufficiently functional")

They are new enough to have USB ports, 99% of the time. Some of the time, they are *SO NEW*, that they dont have USB1.1 support.

As I pointed out, there *IS* a codebase for a USB1.1 hardware driver, to simulate the presence of a gameport joystick. This is because the gameport joystick is very simplistic in its design. It responds on a queried IO port, what it's current electrical status is. A software daemon can also respond on an IO port, and, if nothing is there (because there is no physical gameport), it can simulate the existence of that hardware.

The problem I have, is that my cheap win7 era laptop (that I got FOR FREE), only has USB2.0 and 3.0 functionality. No 1.1 functionality. This means I cannot use that driver.

Recently, Crazii released a USB2.0 driver framework, intended for USB keyboards, Mice, and USB mass storage devices. It would be very nice, if that work was utilized to update the earlier gameport emulation driver for USB1.1, so that it could work with such hardware.

I could then plug in a USB gamepad, and the driver would present its button statuses using the expected IO port. Games that expect a gameport joystick would then "see" it.

I have this preference for a number of reasons.

1) Period correct hardware tends to suffer price gouging on second hand marketplaces.
2) "Ideal" versions there of, will suffer even greater gouging, due to desirability and demand.
3) I am not some kind of purist, and dont care that I am using emulation. Half the time, you were using emulation BITD, ANYWAY. Did you have an SB64? Guess what, you were using emulation if you were playing a DOS soundblaster game. You are not fooling anyone but yourself pretending otherwise.

Reply 1162 of 1162, by MAZter

User metadata
Rank Oldbie
Rank
Oldbie

I decided to tinker with a ThinkPad i 1124 2609-53J (240x) laptop featuring the SM712 video chip. It’s not a huge issue, but there is a quirk: the "HW Expansion" setting in the BIOS—designed to scale resolutions lower than the native 800x600 up to 800x600—doesn't work for 640x480. For instance, 320x200 in games fills the whole screen, whereas 640x480 remains in a small window. I decided to ask ChatGPT to write a TSR program for an experiment with *Duke Nukem 3D*. The game runs smoothly even at 800x600, but I was simply curious if it was even possible to scale its 640x480 output to full screen.

In hindsight, perhaps I should have used Claude from the start; after fifty reboots of the test machine—cycling through suggested tests and every possible register—it feels like we're just going in circles. Early attempts involved a TSR using a laughable software-based method: the image scaled perfectly, but the frame rate dropped to a mere 2–3 frames per second! I had provided ChatGPT with everything available: the laptop's BIOS firmware, the game's source code, and the PDF datasheet.

The best result I managed to achieve is shown in this image: scaling worked only horizontally, and the FPS remained stable, but there were minor artifacts—specifically dots (and a repeated image at the bottom):

The attachment file-00000000a6fc81f58fe18fa83a2526c7.jpg is no longer available

Whenever ChatGPT attempts to correct the vertical aspect as well, the image does scale to fill the screen, but it appears to be cropped from the top-center area (shown here after pressing ESC):

The attachment PXL-20260920-230138487-MP.jpg is no longer available

It seems there was a reason the engineers didn't include scaling for the 640x480 resolution.

The assembly code for the TSR is attached. In addition to the actual scaling attempts, it generates a report after you exit the game and restart the TSR. Build command: `nasm -f bin 1.asm -o 1.com`

Doom is what you want (c) MAZter