PCem

I'm intending v14 to be a relatively small release, basically the features I intended for v13 didn't make it, the backlog of patches that built up, and bugfixes. I'm setting feature freeze for the end of Saturday the 17th of March, to be followed by a few weeks of bugfixing & code hardening. Any feature patches posted after then won't be considered for the release - obviously bugfixes are still welcome! Once v14 is out then I'm going to hunker down for a bit to rework the dynamic recompiler, which will probably be the only major change from me for v15.

I would also appreciate any help with testing & building a compatibility list. PCem's grown to the point where I can't really do this alone anymore; full testing for v13 took weeks. leilei did a great job helping with v13 here - viewtopic.php?f=2&t=885. I'd be extremely grateful to anyone who can help do something similar for v14!

I'll see what I can do about cleaning and sorting some updated documentation in time. This next week will be busy for me but it should slow down afterwards.

I don't know a lot of C but this is something I can help out with.

Thanks for all the work that you've put into this thing.

A higher cylinder limit could help. There are times I wish I had a bit more space so I wouldn't have to make 15 8gb images to test across many configurations. Most of the games dumped in them are from hard drive migrations and often have redundancy

I can probably accommodate that, assuming you're happy using Disk Manager/EZ Drive/Max Blast/whatever to work around BIOS limitations.

Great work. Keep up the good work!

About that compatibly list...
Is it OK to create a new thread and put everything there? Or you prefer something more elaborate?

A new thread's fine with me.

Now feature frozen.

As I'm new here apologies in advance if this post is in the wrong place or is in any way inappropriate.

Firstly PCEm is a great effort and for me running a PS/2 in emulation on my Lenovo T430 is immensely pleasing.

As background about your typical followers I have a small collection of IBM PS/2s an RS/6000 and vintage ThinkPads. In terms of the real thing I am currently implementing SCSI2SD on my Model 70 due to it original HD expiring – yet the HDs in my P70s are still fine.

In terms of other comparable emulators , I like IBMulator as it has some clever features like machine sound effects and nice graphic appearance, which enhances the retro experience. Its rougher edges are its configuration and it emulates only two machines.

I understand PCEm is work in progress and is a background free time project.

One minor in your face v14 “bug” is that “Disc” at the top of the second menu should be Disk – optical is disc, magnetic is disk. Hey in ye olde IBM language we used “floppy diskettes” but “Compact Disc ROMs”.

Minor v14 issues encountered are very occasional mouse focus discontinuities on repeated floppy disk image changing (eg installing OS/2 2.11), a couple of strange PS/2 cmos corruptions (have to run ref disk again for no obvious cause) and occasional hangs on the reference disk splash screens (press enter to continue and when you do nothing happens and it ignores further input)

Known v14 bugs encountered are Model 70 486 POST error codes and apparent failure to read IBM XDF images (DOS-7 and OS2-3+)

Back in the day for some of us there were three types of computers: IBM, clones and Apple. Those of us who played in the IBM world may have paid more but we got a lot out of our bubble because support was excellent and everything just worked.

While PCEm has to be a broad church, could I request some more IBM world tweeking in future versions. Firstly natively supporting the IBM .dsk floppy image format. Some of us have created large floppy software libraries in this format with the free IBM tools. Secondly the program’s directory structure needs to have a structured place for the PS/2 reference and option disks.

Dare it be said the whole PCem rom directory structure could be reworked so everything is more categorised – i.e. graphics, machines, disk interfaces etc.

Could disk activity indicatiors go in the empty part of the menu bar ? (Dont then have to open report window.)

While where the software goes in future versions is up to you as the developer (and yes saw your reply), some of us would find working serial and parallel ports really useful (mapped via USB ports so e.g. plugging in a USB to serial cable to the host PC produces a working legacy serial port on the emulator). Inevitably all our working PS/2s will fail at some point, so having an emulator that can run equipment on these ports will become essential. Secondly native file system access would be good (so running emulator can see host machine’s drives - maybe implement as expansion card “adapter”) so as not to need programs to externally manipulate hard and floppy disk images, which can't be done live. Finally for the PS/2 microchannel emulations we urgently need an emulated SCSI card so optical drives can be accessed.

If it assists, spending more development time to deepen the support for existing emulated machines, rather than expanding the range would be seen positively by some of us, but obviously its your call. Certainly expanding the host platform base (eg to ARM Pi) is both amusing and useful.

Keep up the good work and best wishes.

Hugh Burns

Model70 wrote: ↑Fri 15 Mar, 2019 6:55 am While PCEm has to be a broad church, could I request some more IBM world tweeking in future versions. Firstly natively supporting the IBM .dsk floppy image format. Some of us have created large floppy software libraries in this format with the free IBM tools.

That's the format used by LOADDSKF / SAVEDSKF? I had a go at implementing that in LibDsk, but I never got the LZW compression right.