Subject: Re: Cumana SCSI II Card
To: None <port-arm32@NetBSD.ORG>
From: Markus Baeurle <emw4maba@gp.fht-esslingen.de>
List: port-arm32
Date: 10/06/1996 02:50:31
Hello all!

In message <28622.199610041037@mailhub.ggr.co.uk>
          Mr F H Baylis <fhb22266@ggr.co.uk> wrote:

>   Is anyone aware whether the original members of the Cumana driver
>   development who signed the Non Disclosure Agreement are still working on
>   the project? If not is it not a simple task to get the current developer(s)

Yes they are. But they're all quite busy and I think it might help if the code
could be shown to more people than only the RiscBSD kernel team.

>   to sign another agreement with Cumana. I understand that NDA's are a pain,
>   but I work with major computer harware/software suppliers who are much less
>   willing to divulge information. I can understand Cumana's situation, they
>   aren't saying they can't help, but will under their terms; it's not
>   intended to make life difficult for developers of new OS's, but to keep the
>   competition from stealing their ideas/code.

Oh come on, the Connect32 (it's also called Yucani F1, don't know if it's
available in the UK yet and under which name) is faster than the Cumana
nowadays, available (at least in Germany), stable and proving more and more
popular. (Jan-Uwe who is also on this list bought one after having problems
after flashing the Cumana card, but this was more due to ESH providing no
support. Ask him if you want to know more about the Connect32.)
Obviously they as well as PowerTec were able to produce their own cards without
information from the Cumana RiscBSD driver.
It's also strange nobody else has to fear this except Cumana.

I'm not sure but there's probably also not to gain much from the RiscBSD code
because this only contains rudimentary information and none of the advanced
features which makes these cards quite fast under RiscOS. That's at least the
impression I gained from having a quick glance at the PowerTec driver.
As I said I'm not sure because I don't know which information was provided by
Cumana. Maybe someone from the kernel team can clarify this.

I also received a reply from Mark Gillman arguing like you, culminating in the
following:

MG> > So, please please remove this nerving NDA from the Cumana driver for
MG> > RiscBSD!
MG> 
MG> I'm sorry, but this cannot be considered at the current time.

This was already a few days ago and I wanted to announce this before on this
list, but maybe it was good that I didn't get round to do it after all, because
in a more recent message, Mark Gillman told me that they were discussing the
NDA further and that there might be some change. (Sorry, but I can't quote atm
because the message's not yet on my home machine.)

It's probably a good idea if everybody of you mails to support@cumana.co.uk to
tell them your opinion.
Please think about where the free Unix community would be today if every
information about computer hardware (look at the PowerMac, there was no free
Unix available for it until recently because Apple also didn't release the
information) was only available under NDA.
Consider if you'd rather buy from a company who makes life for users and
programmers of free software easy or from one who insists on NDAs, for whatever
reasons.
Think about this and tell Cumana about your opinion.

Yours, Markus