PCem

It's largely the same as for Linux. You need to install sdl2, wxmac and openal-soft through brew (https://brew.sh/), then run ./configure and make.

Spectre/Meltdown mainly affect code that makes huge numbers of kernel calls, mainly IO heavy stuff. PCem doesn't really fall into this category, hence it shouldn't be too badly affected.

I bought a shiny new Ryzen laptop the other day. Preliminary results - in software Quake under PCem v14, a Ryzen 5 2500U at 3.3 GHz gives about 2/3rds the performance of a Core i7-7700k at 4.4 GHz. Make of that what you will!

PCem v14 is now available. Changes from v13.1 : New machines added - Compaq Portable Plus, Compaq Portable II, Elonex PC-425X, IBM PS/2 Model 70 (types 3 & 4), Intel Advanced/ZP, NCR PC4i, Packard Bell Legend 300SX, Packard Bell PB520R, Packard Bell PB570, Thomson TO16 PC, Toshiba T1000, Toshiba T12...

Fixed in rev 1175.

I'd love to know why Award thought trying to determine CPU speed with the cache off was a good idea.

leilei wrote: ↑

Fri 13 Apr, 2018 8:52 am

- Darkened Skye (crashes PCem when you access one of the menus)

Code: Select all

Bad CMDFIFO packet 3b03126e 001f27ac
CMDFIFO packet 5 has byte disables set 3f0301ad

Can't reproduce this one. Does it need a specific configuration?

This looks to be a result of trying to convert between Windows and SDL views of window coordinates. I've just had a go at trying to fix this and largely failed. I think rewriting wx-sdl2-display-win.c to only use the Win32 size & position APIs is probably the long term solution.

What are you running in these screenshots?

Seems to work okay here. Are your ULTRASND and ULTRADIR environment variables set up correctly? I was able to reproduce your error if either variable was wrong or missing.

2.5.1 also seems to be fixed now.

2.4 with AHA-1542C fixed in rev 1162. I think that rev may have also fixed the IDE issue.

Fixed in rev 1158.

Fixed in rev 1156.

Are you drunk? You sound more rambling than normal.

Should be fixed in rev 1154.

Which drivers were you originally trying to get working? The drivers I've been using (CMS Game Blaster 720KB Driver Disk off vogonsdrivers) detect the card okay and seem to work fine.

Fixed in rev 1153.

leilei wrote: ↑

Fri 06 Apr, 2018 4:41 am

- Magic Carpet 2 has 320x200 VGA issues under Windows 95 on S3 Trio64. Changing to 640x480 by hitting R or windowing and back alleviates this.

Can't reproduce. What exact configuration are you using? What drivers? What do the 'VGA issues' look like?

Not a PCem bug. Pictionary tries to initialise the mouse via INT 33h. Your mouse driver would normally hook this, but you don't have one loaded. DOS 3+ will install a dummy handler, as will AT BIOSes. If you use DOS 1.x/2.x with an XT machine and no mouse driver then nothing will initialise it and P...

That looks more reasonable. Committed at rev 1152.

Fixed in rev 1150, mainly by adding the monitor type configuration...

A bit late responding to this, but this issue is the fault of bad UI design on my part. Once you've configured a CD-ROM drive in the configuration dialogue, it won't show up in the emulated machine until you've selected something other than 'disabled' in the CD-ROM menu in the main PCem window. This...

What version of SDL do you have installed? Those errors imply that it's something prior to 2.0.4. PCem needs something more recent than that.

Can't reproduce this. What version of Windows 95 are you using? What driver? Could you post the emulator configuration you're using?

Is the card showing up in Windows Device Manager as a PCI device?

That's unhelpfully vague.

readme.txt?

What conditions are causing this crash? I'd be extremely dubious about any situation where _mem_exec was set but _mem_mapping_r was not - that would imply that memory mapping has not been set up correctly.

Fixed in rev 1148/1149.

Can you provide the configuration you're using please? As I said above, this works fine for me.