OpenCBM 0.4.99.102 Windows Package

[Spiro Trikaliotis]

Hello,

I just uploaded a Windows binary of OpenCBM 0.4.99.102 as a bugfix for
the problematic v0.4.99.101.

In order to use it, just download it, unpack it and run install.cmd with
your normal user account! It will automatically elevate ("UAC") in order
to get the rights!

Hint: You can use your normal user account for installation. The script
will ask for the necessary rights as soon as it is needed.

Hint: For now, I tested it on Windows 7 and 10. It requires the
Powershell, so everything after Windows 7 SP1 should work.

It can be found here:

https://spiro.trikaliotis.net/download/opencbm-0.4.99.102/opencbm-0.4.99.102.zip

While installing the USB drivers, a red alert will pop up that tells
you that we are not signed, and what bad things can happen. I cannot do
anything about it as I cannot sign the binary myself.

So, If you do not trust me, do not install OpenCBM!



Possible problems:
==================

0. You get a red alert (sorry, I only have the German text) (cf.
attachment):

"Windows Security

Der Herausgeber der Treibersoftware konnte nicht überprüft werden.
(Could not determine the manufacturer of the driver software)

--> Diese Treibersoftware nicht installieren
(--> Do not install this driver software)

[...]

--> Diese Treibersoftware trotzdem installieren
(--> Install this driver software anyway)"

This is not a problem. I cannot sign the software, so Windows insists on
this popup.

If you trust me, use option 2 ("install anyway"). If not, you cannot use
OpenCBM.


1. After installation of the driver and trying to perform an OpenCBM command
(cbmctrl, ...), you get the message:

Error loading plugin '(null)': The parameter is incorrect. (87)
Error loading plugin '(null)': The parameter is incorrect. (87)
NO PLUGIN DRIVER!: The parameter is incorrect.

It seems the installation was not done correctly. Did you get an UAC
prompt when trying to install OpenCBM? Please note that an administrator
account is not enough, you need the UAC prompt (unless it is disabled by
you, of course).

It is also possible that you did not install OpenCBM at all. It needs an
installation, or it will not work.

Please use the install.cmd script, as described above.



2. After installation of the driver and trying to perform an OpenCBM command
(cbmctrl, ...), you get the message:

error: Cannot open USB device: LIBUSB_ERROR_NOT_SUPPORTED
error: no xum1541 device found
error: Cannot open USB device: LIBUSB_ERROR_NOT_SUPPORTED
error: no xum1541 device found
libusb/xum1541:: The handle is invalid.

This happens sometimes if the device is already plugged in while
installing the driver. To solve this issue, unplug the device, wait some
2 or 3 seconds and plug it in again.



Changes w.r.t. 0.4.99.101:
==========================

- Windows:
. The USB access problem should be gone now:
- no error messages or blinking after issuing commands

. Installation failed if the path of the unzipped directory
contained a space

. Install detects if installing on Windows XP or older. In this case,
it does not try to automatically install the USB drivers, as this will
fail anyway.

. On XP and older, install will change to an administrator account
instead of trying the UAC dialog (which does not exist)

- Linux:
. Fix some wrong references to directories where the .so files are
stored

. Better integration of the nibtools build system

- Overall:
. If the command line parameters are wrong, cbmctrl sometimes output
some error message about USB access that was plain wrong. Fixed that.

. Fix libusb0 compilation


Changes of 0.4.99.101 w.r.t. 0.4.99.100:
========================================

- Remove the misleading error:
USB deconfig device error: -2 LIBUSB_ERROR_INVALID_PARAM

- the OpenCBM shell is removed on uninstall of OpenCBM

- updating an OpenCBM installation does not ask you over and over again
if you want to overwrite files

- fixed the internal path of the firmware-update.bat script, so it works
again

- remove the uninstall script from the path, so OpenCBM is not
accidentially uninstalled

- different small fixes by Martin Thierer. Thanks for these!


What is still missing for v0.5.0?
=================================

- For Windows XP and 2000, the drivers must be provided so xum1541 and
xu1541 can be installed there.

- excessive testing

- documentation updates


Regards,
Spiro.

--
Spiro R. Trikaliotis
http://spiro.trikaliotis.net/

[rittwage]

I updated. So far so good under Windows 7 and 10 (amd64).

[rittwage]

Hello Spiro,

I see you pushed 103 into your repo over a month ago. Is there a binary release of it available?

[Spiro Trikaliotis]

Hello Pete,

* On Thu, Sep 03, 2020 at 12:17:35PM -0700 rittwage wrote:

> I see you pushed 103 into your repo over a month ago. Is there a binary release
> of it available?

Didn't I announce it?

It is available at
https://spiro.trikaliotis.net/download/opencbm-0.4.99.103/opencbm-0.4.99.103.zip

Regards
Spiro

[xAD / nIGHTFALL]

Hi Spiro

I have updated to the latest version of opencbm 0.4.99.103.

everything works perfectly (opencbm native tools) but not the tools: nib ** (nibread,nibwrite .. ) and so on.

I add an example screenshot here:

https://www.nightfallcrew.com/download/nibread_error.png

do you have any idea?

Thanks.

Dave.

[rittwage]

Bad IEC cable (missing SRQ line) or bad 6526A (or 8521) in the drive.

[Spiro Trikaliotis]

Hello,

adding to what Pete wrote:

* On Tue, Nov 10, 2020 at 05:19:57AM -0800 rittwage wrote:
> Bad IEC cable (missing SRQ line) or bad 6526A (or 8521) in the drive.

Are you using an original ZoomFloppy device? Or another xum1541 based
device? What firmware are you running? Give the complete output of
the command "xum1541cfg devinfo", please.

Regards,
Spiro.

--
Spiro R. Trikaliotis

https://spiro.trikaliotis.net/

[xAD / nIGHTFALL]

Hi Spiro,

for the avoidance of doubt i am a oldskool programmer and electronic repair man from 38 years :/  , look my blog: https://www.nightfallcrew.com  :-D

Thank you very much for your reply , i appreciate it.

I use ZoomFloppy with firmware 0.7 or .0.8 (tested both) , the firmware 0.8 found in the opencbm  0.4.99.103.

The IEC cable obviously has all the pins connected, i have here a hundred of IEC cables never found one without SRQ line :-D

I have tested the opencbm (0.4.99.99 and 0.4.99.103) with different config.

Old laptop with windows XP and usb 2.0

PC with windows 10 Pro 64bit and usb 3.0

PC with windows XP 32bit and usb 2.0

virtualbox under windows 10 pro 64bit with windows 7 usb2 or usb3

in all my tests the nibtools (tested: nibread / nibwrite) do not work. :-D

The other native tools of opencbm work perfectly.

I use a 1571 Floppy Drive in a perfect working state, just to avoid making the parallel cable on the 1541.

I am available to made all the tests you ask me, i only have time at night :D

thx

Dave.

[Spiro Trikaliotis]

Hello Dave,

* On Tue, Nov 10, 2020 at 03:38:00PM -0800 xAD / nIGHTFALL wrote:

> I use ZoomFloppy with firmware 0.7 or .0.8 (tested both) , the firmware 0.8
> found in the opencbm 0.4.99.103.
>
> The IEC cable obviously has all the pins connected, i have here a hundred of
> IEC cables never found one without SRQ line :-D
>
> I have tested the opencbm (0.4.99.99 and 0.4.99.103) with different config.
>
> Old laptop with windows XP and usb 2.0
> PC with windows 10 Pro 64bit and usb 3.0
> PC with windows XP 32bit and usb 2.0
> virtualbox under windows 10 pro 64bit with windows 7 usb2 or usb3
>
> in all my tests the nibtools (tested: nibread / nibwrite) do not work. :-D
>
> The other native tools of opencbm work perfectly.
>
> I use a 1571 Floppy Drive in a perfect working state, just to avoid making the
> parallel cable on the 1541.
>
> I am available to made all the tests you ask me, i only have time at night :D

Unfortunately, I am not the expert when it comes to the SRQ nibbling
protocol that is used with nibtools. This protocol is much more
stressfull and picky than all the other protocols in OpenCBM.

To be honest, you have already done what I would have suggested:
- Use OpenCBM 0.4.99.99 (which uses libusb0) and 0.4.99.103 (which uses
libusb1 and many recent fixes), to rule out any problems introduced
lately;

- Use FW 7 and 8, as most probably, 7 is the mostly used one in the
wild, while 8 has some fixes which also might be problematic in your
case;

- use different OS (although VirtualBox might not be the best choice, as
SRQ nibbling is really picky).


Things you could still do:

- Try another 1570 or 1571 (there have been reports from people who
could make SRQ nibbling work on one 1571, but not on another one; most
probably, due to the CIA 6526);

- Try other IEC cables, especially of other lengths! (I remember reports
here that some people had problems with cables that were too long,
while others had problems with cables that were too short!)

- Try Linux (to rule out any HW based problem)


Other than that, I am a little bit out of ideas.
You could try to connect a logic analyzer (if you have one) to the IEC
lines and send the output, though.

@Arnd, any other ideas?

[xAD / nIGHTFALL]

Hi Spiro,

I did two tests:

- very short IEC cable (all pins connected)

- a new 1571.

Same results. :(

short cable: https://www.nightfallcrew.com/download/nibread_error_1.jpg

dos error: https://www.nightfallcrew.com/download/nibread_error_2.jpg

my 1571 mobo (both 1571 uses the same mobo and same chip version - one with the jiffydos the other not): https://www.nightfallcrew.com/wp-content/gallery/commodore_floppy_1571/img_0286.jpg

Dave.

[Nate Lawson]

I’m suspicious — 8521R0 with an “A” stamped on it. Is that the 2 MHz part?

What does the other 1571 have on its board?

-Nate

> --
> You received this message because you are subscribed to the Google Groups "ZoomFloppy Users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to zoomfloppy-use...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/zoomfloppy-users/7b1d2de3-d18c-4e7c-98fa-9bcbdefadb4en%40googlegroups.com.

[xAD / nIGHTFALL]

Hi Nate,

The letter "A" of the CIA 6526 indicates the 2 MHz version, i deduce that is the same thing for the 8521.

Both motherboards use the same CIA.

Dave.

[Raj Wurttemberg]

Hey Dave,

Silly question... What "terminal" is that?

I'm also on Windows 10... ZoomFloppy (v7) and a plain 1571.  I ran the same commands you did and mine appears to work:

https://www.c64.us/pix/opencbm-w10-1571-works.jpg

Is there anything I can compare for you?  I'm just a user... and I think my xum1541cfg is older than yours because the 'devinfo' command does not work on mine.

/Raj

--

You received this message because you are subscribed to the Google Groups "ZoomFloppy Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to zoomfloppy-use...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/zoomfloppy-users/9d6c9f72-0a10-48f9-92f0-6c9271e16872n%40googlegroups.com.

[xAD / nIGHTFALL]

Hi Raj,

The Terminal is the ConEmu a fast and reliable terminal window where you may host any console application developed either for WinAPI (cmd, powershell, far) or Unix PTY (cygwin, msys, wsl bash).

Obviously i did the same tests with the very BAD native windows CMD, absolutely nothing changes.

The xum1541cfg is the new version , like written in my previous message.

I have used these two version for the test:

0.4.99.99

0.4.99.103 ( https://spiro.trikaliotis.net/download/opencbm-0.4.99.103/opencbm-0.4.99.103.zip)

...and the latest nibtools.

Dave.

[Spiro Trikaliotis]

Hello Dave,

* On Wed, Nov 11, 2020 at 11:47:54PM -0800 xAD / nIGHTFALL wrote:
> Hi Nate,
>
> The letter "A" of the CIA 6526 indicates the 2 MHz version, i deduce that is
> the same thing for the 8521.
>
> Both motherboards use the same CIA.

Do I read the datecode correctly as 4485?

I know, it is a shot in the dark, but: This is exactly the datecode that
has been identified here: https://sourceforge.net/p/vice-emu/bugs/1219/
as being different from other CIAs, and this test has been done to
identify differences in the serial data processing of the CIA - exactly
what we are using!

You do not happen to have 1570 or 1571 with a CIA with another datecode? ;)

[xAD / nIGHTFALL]

Hi Spiro,

ALL is FIXED!

I have replaced the CIA 8521R0-A with a 6526A.

But this C1571 with the CIA 8521R0-A worked perfectly in burst mode on the C128.

So ... in summary, not only the 6526 (not A) are not good but neither are the CIA 8521R0-A

https://www.nightfallcrew.com/download/nibread_ok_6526A_not_8521R0-A_DOS.jpg

https://www.nightfallcrew.com/download/nibread_ok_6526A_not_8521R0-A_PCB.jpg

Many thanks all.

Dave.

[Spiro Trikaliotis]

Hello Dave,

* On Thu, Nov 12, 2020 at 01:43:59PM -0800 xAD / nIGHTFALL wrote:

> ALL is FIXED!
>
> I have replaced the CIA 8521R0-A with a 6526A.

Thank you for this feedback. So, there is a first indication that my
suspicion might be correct.

> But this C1571 with the CIA 8521R0-A worked perfectly in burst mode on the
> C128.

Well, SRQ nibbling puts the serial port on its physical limit: The timer
is programmed with the smallest possible value (1), the bps is PHI2 / 4;
because we are on 2 MHz, this is 500 kbps.

This is an order of magnitude higher than what the burst mode of the
1571 uses.

> So ... in summary, not only the 6526 (not A) are not good but neither are the
> CIA 8521R0-A

Well, from other sources, it seems that dies from a 8521 were put in a
6526, and vice versa, so this is no big surprise.

Does your data code read 4485 or 4486? The later picture is not clear to
me.


> Many thanks all.

If we come up with some more tests, would you mind trying to execute
them with the non-working CIA? Who knows, we might find out how to fix
the problem for them, too.

[xAD / nIGHTFALL]

This is a better photo:

4486 ; https://www.nightfallcrew.com/gallery/commodore_floppy_1571/img_0294.jpg

I'm sorry but these are historical photos of my blog taken in 2008 :D

nice idea, no problem, i could do it by capturing the states with a logic analyzer;  first with the working CIA and then with the not working one.

D.

[Raj Wurttemberg]

Silly question... What "terminal" is that?

I'm also on Windows 10... ZoomFloppy and a plain 1571.  I ran the same commands you did and mine appears to work:

https://imgur.com/razb2Lc 

Is there anything I can compare for you?  I'm just a user... and I think my xum1541cfg is older than yours because the 'devinfo' command does not work on mine.

/Raj

 

--

You received this message because you are subscribed to the Google Groups "ZoomFloppy Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to zoomfloppy-use...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/zoomfloppy-users/7b1d2de3-d18c-4e7c-98fa-9bcbdefadb4en%40googlegroups.com.

[Nate Lawson]

> On Nov 13, 2020, at 4:22 AM, Spiro Trikaliotis <an-zoomfl...@spiro.trikaliotis.net> wrote:
>
> Hello Dave,

>
> * On Thu, Nov 12, 2020 at 01:43:59PM -0800 xAD / nIGHTFALL wrote:
>
>> ALL is FIXED!
>>
>> I have replaced the CIA 8521R0-A with a 6526A.
>
> Thank you for this feedback. So, there is a first indication that my
> suspicion might be correct.
>
>> But this C1571 with the CIA 8521R0-A worked perfectly in burst mode on the
>> C128.
>
> Well, SRQ nibbling puts the serial port on its physical limit: The timer
> is programmed with the smallest possible value (1), the bps is PHI2 / 4;
> because we are on 2 MHz, this is 500 kbps.
>
> This is an order of magnitude higher than what the burst mode of the
> 1571 uses.

Right. CBM DOS burst is about 3.5 KB/second and only for reads. Nibtools is 62.5 KB/second (almost 2 orders of magnitude) for both reads and writes.

The raw media rate is 40 KB/second of GCR so we need to be able to keep up with that to do a track in a single rotation. This covers nearly all protection schemes and is the same thing the 8 KB drive expansions like RAMboard enable.

[RETRO Innovations]

Since we have a few of these problematic CIAs running around, is there
an option to "loan" one or two to you and/or Spiro to expedite testing?

Jim


>

--
RETRO Innovations, Contemporary Gear for Classic Systems
www.go4retro.com
store.go4retro.com