Re: Looking for Nicolas Welte

From: Marko Mäkelä (
Date: 2007-07-10 22:25:01

Hi all,

On Tue, Jul 10, 2007 at 08:57:37PM +0200, Wolfgang Moser wrote:
> Hello list,
> Nicolas Welte wrote:
>> I'm alive and it's good to see you can't hide nowadays anymore :)
>> I only wonder by which name Whitewolf had contact with me?
>> Marko: I followed your emails and appreciate your efforts, but can't
>> help much at the moment. Of course, whenever possible, I will flash
>> the new firmware and give it a test. The other person who can do
>> tests is Womo, who already got a kit from me immediately after I had
>> the PCBs.

I understand.  I'm also suffering from lack of time, or perhaps having
too many other tasks.  Life was so much simpler as a student, when I
wasn't a father or a homeowner.

> well, I'm a bit overloaded currently due to several
> critical job related issues. The situation regarding
> the C2N232i is as following:
>    * I got the PCB and all parts from Nicolas
>    * The device is not assembled yet
>    * I recently updated my AVR-GCC toolchain to GCC 4.2
>    * Using a real C64 is difficult, I've got no monitor
>      anymore and would have to look out for an old TV
>    * low-level RS232 tools like RawTerminal would need
>      a re-installation

Given this situation, I would postpone the debugging of the serial bus
driver.  The C2N232I seems to work here on three platforms: Linux 2.6
with a PL2303 USB-to-RS232 adapter, and Linux 2.6 and Windows XP with
a 16550 based RS232 interface.

> Which tests are required to get an impression, if the
> port is working as expected?

I've tested with cisp, c2nload, c2n (LOAD and SAVE) and cbmlink.  I
don't think that a monitor is needed on the Commodore side.  Only
the c2nload and c2n commands require typing.  You can blindly press

For some reason, the Windows PL2303 driver from 2005 hangs with the
C2N232I, but it works with the C2N232.  I should try to mount an 8 MHz
crystal on a C2N232I and load the same firmware on both devices to see
if this is a timing problem (race condition in the PL2303 driver).
The PL2303 even causes a Blue Screen of Death when I attempt to interrupt
c2nload when it is hung while trying to talk to the C2N232I.

I think that we should test the C2N232I on some other operating system
and hardware platforms, such as FreeBSD, Solaris and Mac OS X on x86
and other hardware.  If the new programming interface works on all
these, the programming jumper could be omitted and replaced with a
trace that can be cut if needed (when the computer can't control DTR).
A friend of mine has (had) various *nix boxes that I might be able to
test with.

It's a pity that many new computers lack an RS-232 interface.  The drivers
of many USB-to-RS232 adapters are buggy (and different bugs on different
operating systems).

It'd be nice to know which adapters to recommend on which operating systems.
So far, I know that the Keyspan adapter works on Mac OS X (and possibly
others), that the PL2303 works on Linux 2.6, and that the FTDI FT232
has different kinds of flow control bugs on Mac OS X, Linux, and Windows.
I'd appreciate help in diagnosing the PL2303 problems on Windows, and
checking if the PL2303 driver works on Mac OS X.  PL2303 based adapters
are inexpensive (10 EUR).  Ideally, the same USB-to-RS232 adapter should
work correctly on all operating systems.

I haven't published the updated software and firmware yet.  I'll upload
it at a hidden place and email you (Nicolas and WoMo) privately.

If someone else is interested in testing (Patryk? Spiro?) I think that
I could send an assembled and tested C2N232I unit to each, perhaps at
the cost of parts and postage, if that's okay with Nicolas.  But let's
discuss this off the list.


       Message was sent through the cbm-hackers mailing list

Archive generated by hypermail pre-2.1.8.