Aitor has been doing some testing with the FreeDOS Beta6 pre-release and some other FreeDOS software, and has this to report:


Hi all,

This is the result of the testings of the Pre-Beta6 of FreeDOS.

Machine: IBM PS/1 with a HD of 20Mb (with bad sectors, so good for testing FORMAT ;-))

Testing results:

Beta5 (FreeCOM + Kernel): cool! No problem at all. It was working smooth (the fix for cd.. was good ;-)). By the way, it seems that we don't have a MOVE command, do we?

Even that awful failure of Beta5, where Abort, Retry Ignore? was preceeded by an entire page of bullshit seems to have disappeared. What I did: I booted from MINI (path=a:\) then removed the disk. After several aborts, I got error #16. I assume this is correct, although such a dull message (error #16) was not too user friendly. I imagine this was due to the lack of STRINGS.DAT. However, shouldn't COMMAND have at least ONE message, for the case in which STRINGS.DAT can't be reached?

Another one about the Kernel: NUMLOCK=OFF is NOT working, and it is so easy...

#define BIOS_KBF  ((unsigned far)* MK_FP(0x40,0x17))
*BIOS_KBF |= numl*32;
union REGS regs;
regs.h.ah=1;
int86 (0x16, &regs, &regs);

where numl=1 IF =ON, and =0 if =OFF.

Similar code for CAPSLOCK (if we wanted to implement CAPSLOCK= and SCROLLLOCK= in CONFIG.SYS), just replacing 32 by 64 and 16... (should I post this to the kernel mailing list?)

FORMAT: I formatted a 1.4Mb floppy disk (only quick format available), and had no problem with it. Just one thing: if the disk is write-prottected, then FORMAT just displys the disk information summary (as if it had finished formatting). I think it should say 'disk is write prottected, please insert a new disk'. M$ does that.

One more thing: I tested the FORMAT included in the Pre-Beta6 MINI, but I noticed that in ibiblio, there is another FORMAT 0.8, which doesn't seem to be from Brian, what do you know about this? (I haven't tried it).

xKEYB 1.6: Works fine!

CuteMouse: Again my mouse was not recognised: it says COM1 Mouse Systems mouse, and I don't think it is mouse systems. Be it as it may, it doesn't work, perhaps I have a strange kind of mouse. I am ready to test any kind of new versions or programs to try and detect this. Something curious: I loaded M$-MOUSE without rebooting (or exiting cutemouse), and I got: 'Existing mouse driver enabled', but it didn't work anyway. [CuteMouse was not included in the Beta6 pre-release, so I assume this is an older version of CuteMouse? -jh]

WINDOWS 3.1: Strange behaviour indeed. INSTALL started installation without problems (it passed the 'installing HIMEM.SYS' (??286?), and checking existing the software running on your PC', then started copying files.

It freezed on disk3 (95%).

At this point, I rebooted the machine. C:\WINDOWS had few files, being one of them INSTALAR.EXE (INSTALL.EXE of the Spanish version). I executed it, and got: INSTALL can't run an extended memory manager (XMS), and can't continue without a compatible XMS manager' (translation on the fly).

Other issues:

Why 'compatible'?

Why should it try to install XMS, if it is a 286??

Well, that's it for now. I hope that freezing at disk 3 was nothing accidental. The next time I have the ocassion to test FreeDOS in a 386+ machine I will retry Windows3.1, and FDCDEX (my 286 hasn't got any CD-ROM).

I might continue trying some apps, I will tell you if I am successful again...

-Aitor


Also, these notes on installing Windows on FreeDOS Beta6:
This is as far as I've been able to get with Windows 3.1 + FreeDOS:

After installing Windows 3.1 via MS-DOS and then booting the system with FreeDOS b6, loading Win3.1's own HIMEM.SYS, and then installing the setverb6.zip (as found in Simtel.net), the following things happened:

win (386 enhanced mode)
Starts to load but then freezes and outputs Unsupported MS-DOS version

win /s (Standard Mode)
MCB chain corrupted
System halted

Only thing in the config.sys was:
device=himem.sys

We're getting there! It used to just hang the machine with earlier versions of FreeDOS.

-wkl