From Newsgroup: comp.os.msdos.misc

On Sun, 10 Sep 2017 06:38:52 GMT, Harry Potter <rose.joseph12@yahoo.com>
wrote:

If I upgrade from EMM386 to UMBPCI, what can I expect from the memory
layout. I understand that it uses less RAM, but what *else* will it
do to the UMB memory layout? Will it make more memory available?
What about the EMS page frame? And what are the available switches?
I can't find the last one in the docs.

Are you going around in circles?


--
Bah, and indeed, Humbug.
--- Synchronet 3.17c-Linux NewsLink 1.110

From Newsgroup: comp.os.msdos.misc

On Wed, 1 May 2019 09:57:09 -0000 (UTC), Kerr-Mudd,John wrote:

On Sun, 10 Sep 2017 06:38:52 GMT, Harry Potter wrote:

If I upgrade from EMM386 to UMBPCI, what can I expect

Much trouble. I should say, just use EMM386, but you never listen.

Are you going around in circles?

Someone had to tell Harry about UMBPCI, didn't they. That's like giving
matches to a child and sending them into a dry forest.


--- Synchronet 3.17c-Linux NewsLink 1.110

From Newsgroup: comp.os.msdos.misc

On Wed, 01 May 2019 15:36:59 +0000, T. Ment wrote:

On Sun, 10 Sep 2017 06:38:52 GMT, Harry Potter wrote:

If I upgrade from EMM386 to UMBPCI, what can I expect

Much trouble. I should say, just use EMM386, but you never listen.

UMBPCI always worked perfectly for me. I do not use EMM386 at all, don't
like V86 mode.

Someone had to tell Harry about UMBPCI, didn't they. That's like giving matches to a child and sending them into a dry forest.

There's always the possibility that once the forest is on fire, the child
will perish there. We can hope, at least.

Mateusz
--- Synchronet 3.17c-Linux NewsLink 1.110

From Newsgroup: comp.os.msdos.misc

On 01 May 2019 17:23:03 GMT, Mateusz Viste wrote:

UMBPCI always worked perfectly for me.

I have one computer where it's perfect. All the others, DMACHK says
ISA-DMA fails in UMBs. It reports the ISA-DMA failure with both UMBPCI
and EMM386.

So it's not clear if DMACHK has a problem, or if UMBPCI is fully safe,
or not. They warn about the ISA-DMA problem in the readme though, so I
presume they know what they're talking about.

They say ISA-DMA in UMBs is trouble when you load SMARTDRV high and
cache the floppy drive. Maybe I will try that, and see what happens to a
(test) floppy disk.

I do not use EMM386 at all, don't like V86 mode.

What's wrong with V86 mode?



--- Synchronet 3.17c-Linux NewsLink 1.110

From Newsgroup: comp.os.msdos.misc

On Wed, 01 May 2019 15:36:59 GMT, T. Ment <t.ment@protocol.invalid>
wrote:

On Wed, 1 May 2019 09:57:09 -0000 (UTC), Kerr-Mudd,John wrote:

On Sun, 10 Sep 2017 06:38:52 GMT, Harry Potter wrote:

If I upgrade from EMM386 to UMBPCI, what can I expect

Much trouble. I should say, just use EMM386, but you never listen.

Are you going around in circles?

Someone had to tell Harry about UMBPCI, didn't they. That's like giving matches to a child and sending them into a dry forest.

My point was that he asked this 18 mths ago.



--
Bah, and indeed, Humbug.
--- Synchronet 3.17c-Linux NewsLink 1.110

From Newsgroup: comp.os.msdos.misc

On Wed, 01 May 2019 20:06:26 +0000, T. Ment wrote:

On 01 May 2019 17:23:03 GMT, Mateusz Viste wrote:

UMBPCI always worked perfectly for me.

I have one computer where it's perfect. All the others, DMACHK says
ISA-DMA fails in UMBs. It reports the ISA-DMA failure with both UMBPCI
and EMM386.

So it's not clear if DMACHK has a problem, or if UMBPCI is fully safe,
or not. They warn about the ISA-DMA problem in the readme though, so I presume they know what they're talking about.

They say ISA-DMA in UMBs is trouble when you load SMARTDRV high and
cache the floppy drive. Maybe I will try that, and see what happens to a (test) floppy disk.

Oh, I see - but I never load stuff that uses DMA high. That's probably I
have no problems. I do remember however, many years ago, getting into
problems with some CD driver (I think). Not loading it high resolved the issue, that's when I learned never to trust DMA when running from within
UMB.

I do not use EMM386 at all, don't like V86 mode.

What's wrong with V86 mode?

I don't like having some extra layer between me and the hardware. When my system crashes with an "invalid opcode" or similarly terrifying message,
I don't want having to second-guess whether it's my program that's buggy,
or the V86 monitor that went south. If bad stuff happens on bare
hardware, then at least I know who to blame (the hardware, of course).

Mateusz
--- Synchronet 3.17c-Linux NewsLink 1.110

From Newsgroup: comp.os.msdos.misc

I just tried UMBPCI on my network configuration, and the driver crashed with an exception during start-up, forcing me to do a hard reset. :(
--- Synchronet 3.17c-Linux NewsLink 1.110

From Newsgroup: comp.os.msdos.misc

On Sun, 19 May 2019 09:56:46 -0700 (PDT), Harry Potter wrote:

I just tried UMBPCI on my network configuration, and the
driver crashed with an exception during start-up, forcing
me to do a hard reset. :(

Did you read this?

https://www.uwe-sieber.de/umbpci_e.html

On AMD K7/K8 you should not load high drivers of PCI network cards.

--- Synchronet 3.17c-Linux NewsLink 1.110

From Newsgroup: comp.os.msdos.misc

On Sunday, May 19, 2019 at 1:37:40 PM UTC-4, T. Ment wrote:

Did you read this?

https://www.uwe-sieber.de/umbpci_e.html

I thank you for the tip: it's unsuitable for my needs. :(
--- Synchronet 3.17c-Linux NewsLink 1.110