On Sat, May 18, 2002 at 10:39:29AM +0100, Roy Wood wrote:
> 
> >I detect from the tone of your response that you are a bit cheesed off with
> >Richard's comments on the proposed licence.
> I missed out on some of this because I have been trying a new spam 
> rejection program which was harder to configure than I thought so I 
> apologise if some of my comments are slightly off.
> >
> >I think that he has some valid points which you don't seem to have
> >understood. Think of it this way: Richard has done a good job with UQLX and
> >has it working on may flavours of Unix, on different platforms and
> >processors. It would benefit the community to SMSQ to have it supported by
> >UQLX, and have Richard's Unixy extensions within it as he has already done
> >for JS and Minerva. However, he does not have continous access to all those
> >platforms, and definitely not all the combinations of interface cards,
> >displays etc they come with. He therefore cannot guarantee support, or to
> >fix problems. He helps where he can, and in the spirit of GNU etc, he makes
> >the information available so that technically advanced users can help
> >themselves.
> This is surely not a problem because the technically advanced can have 
> the source code and do the fixes, pass these back to Richard and he can 
> get them into an 'official' UQLX SMSQ/E.

that is the optimist view. However there is nothing in the license
that would guarantee me that the source code would be continuously 
available in the future.
There is nothing in the license that would guarantee me any of my
changes will get back into official SMSQ.
There is nothing in the license to guarantee me that official
or inofficial binaries of SMSQ will be available.

> >
> >Under the current proposed licence, he cannot operate in this reasonable
> >way. He cannot merely do his best, but he has to give an open-ended
> >commitment to provide support -something that few if any software vendors
> >would do.
> Only in the PC world. In practice most of the QL authors have, in the 
> past, given pretty much open ended support. Bugs and problems get 
> reported either to the authors or to the vendors and get passed on and, 
> where possible, fixed. Marcel, in particular, has been pretty tireless 
> in this area. There will be some insurmountable problems with some 
> hardware but that is unavoidable, even on the best platforms. The 
> problem of the many different flavours of both UNIX and LINUX is one of 
> the things that this licence is setting out to try to avoid.
> >He isn't even allowed to provide effective support -emailing
> >patches, assistance over the phone of how to hack a config file ....outlawed
> >by the proposed licence.
> Patches and hacks always confuse the issue and, given the complex nature 
> of SMSQ/E a patch here may destroy something else there (This happened 
> several times to TT himself so I should imagine it would be worse for 
> someone who is not the original author). We want coherent and uniform 
> versions of SMSQ/E not hacked and patched ones. Of course this may not 
> matter if the patches and hacks are confined to UQLX but this is an 
> opening for the 'if they can do why can't I?'

things will not get better if there isn't any official version for
some platform and people will try to compile their binaries from
sourcecode.. perhaps sourcecode they received snail-mailed from
whomever I don't know with an unknown set of hacks.

> >This is not the way to encourage the few souls who are both willing and
> >capable of making SMSQ available and useful to a wider audience to harness
> >their talents to our mutual benefit.
> WE are only asking for discipline I think.

than the license is very badly engineered. It enforces discipline
by rather brute methods that will only hurt people who would like
to help and leaves too many important points wide open.
I have proposed alternatives to Wolfgang, something like this:

<< you are allowed to do anything with this code as long as
    - you accept this copyright
    - you leave this copyright message intact and don't
      place any additional restrictions on the code
    - you don't sell this source or anything derived from
      this source, including binaries
    - you don't branch the code.
   licensing for commercial purposes is available under
   following conditions:
    ...........
    ...........
>>

If discipline is all you want than this should do quite
well and still leave sufficient room for commercial
development. The formulation above may seem a bit naive
- it is. We aren't expecting to deal with criminals here,
are we?

Richard

Reply via email to