On 21 May 2002, at 23:38, Jeremy Taffel wrote:

> A lengthy response, please don't flame, I would appreciate a considered
> response.
And a VERY long reply...
As long as you don't flame me, I don't flame you.... I don't think I 
did, at least, obviously you feel different... :-)


> Because instead of answering in a civil an unemotional way you get provoked
> into escalating the flame wars, and often don't address the legitimate (in
> their minds) concerns of some of those that ry to debate this.

What kind of reaction do you expect when I'l being called a 
racketeer? (not by you!)

(...)
> > So he wants to program something, but not support it later on.
> > Nice.
> 
> That is not what I said. With the best will in the world, the key word here
> is "guarantee". I know from personal experience that Richard, (and probably
> the others giving you grief on this list) provides exceptional support. You
> must know, however, how many platforms (and OS's) UQLX runs under, and any
> problems arising from unusual combinations of hardware and OS could take a
> long time for any individual to bottom out, especially if he hasn't got the
> access to that hardware. I imagine that similar guarantees are probably just
> as problematical with QPC given the variety of hardware and software drivers
> etc running Windows. Any such guarantees are essentially dishonest, and are
> only ever "best efforts". That should be recognised, otherwise we should be
> asking questions like "how quickly can we expect to get the bugs fixed?" We
> don't because we are reasonable people. However some people when they read
> the licence interpret it literally.

Well isn't that a legitimate question? If you buy on OS that is 
bugged in some fundamental way, isn't it your right to expect the 
bugs to get fixed?
In the situation as it was until now, when a new version of SMSQ/E 
came out, it sometimes did have bugs. The users then contacted 
the person they bought their SMSQ/E from, most probably Jochen, 
Roy or Peter.
They passed on the reports to Tony (or Markus, if the problem was 
QPC related) and the bugs got fixed. Ok, they got fixed sooner or 
later only - but they did get fixed.
(At least I'm not aware of any major bug in SMSQ/E as it stands 
now - and please peole, I'm talking about bugs, not missing 
features!).

> I was making a serious point and received an unwaranted (in my opinion)
> sarcastic glib response. This really is not necessary.

I didn't mean to offend you. The reply wasn't meant to be sarcastic, 
but reflects what I understood from your posting.

However, the problem remains: How do you implement any kind of 
bugfix scheme in something like SMSQ/E if it becomes entirely 
free? Then there is no legal relationship whatsoever. However, 
many users require support. Hence the restriction on distributing 
the binaries.
The reasoning I have always had is as follows:
If anybody makes a change in the sources, then how will this be 
distributed? 
There is nothing that forces you to give your change to the 
registrar, if you don't want to - but then, you can only distribute 
your change as source code (if it contains original SMSQ/E code - 
if not, this licence doesn't concern you). If you give it away as 
source code, then, if the recipient can compile this and make 
himself a new SMSQ/E, then there is a fait chance that the 
recipient WILL NOT NEED ANY TECHNICAl SUPPORT, or at 
least, will know what the problems are.
If the recipient can't compile everything, then he is more of a 
"simple user" - and he should not get untested binaries. He should 
buy SMSQ/E, or get an upgrade, from a reseller, who can supply 
support.

(snip)
> UQLX is distributed as source, and in my experience most if not all Linux
> users are familiar with make-files. So lets see..
>  provided the  developer provides the necessary compiler/cross-compiler and
> makefile(s) for the platform, he can freely distribute it as a set of source
> files. Sounds like open source to me. Only leaves the problem of how to get
> you to accept it into an official version (don't flame -see later comments).

See above - if they can compile it, then they are probably 
sufficiently "advanced" to tinker with the system. There is 
ABSOLUTELY no problem in distributing the source code in this 
way  - the restriction lies in the distribution of the binaries.

> We are still waiting on this list for a definition of "support". It seems to
> be absolutely essental, but totally undefined! It seems, however to underpin
> most of your defence of the approach being taken.

Ok, lets address this question here:
What kind of support would you, the simple user, like?
According to you, who should supply it?

(snip)

> > That's true. What would be my interest in doing so?
> >
> But what's to stop you? 

Nothing. But, again, I don't see why one should suddenly change a 
licence that we have had so much trouble in setting up in the first 
place. That would only lead to outcries and rejections

> I think that part of your role is to provide the
> reassurance that you are aiming to act in an inclusive manner, not
> exclusive; also to provide objective criteria for inclusion of new
> offerings. e.g.  coding style, completeness, compatibility, maintainability,
> documentation, personality (just kidding - but without any of the other
> information how do we know what criteria will be used?).
Well, my role is actually being defined here and right now, and it is 
a much broader role that I had initially in mind.
Essentially, I just wanted to act as some kind of clearing house - 
get individual author's ideas and contributions, try to make sure 
they are implemented accross the range as much as posible, 
release the new versions of the source and, possibly, supply the 
resellers with compiled versions.

I don't know what other reassurances than those I have (vainly, it 
seems) tried to give here in the past I could still give you. I cannot, 
and will not, guarantee that nothing will never change, to do so 
would be absurd.
I can only state that I still intend to make sure that every platform 
on which SMSQ/E runs now will continue to have up to date 
sources (and this binaries). I actually see this as my main work - 
for a very simple reason: the way this debate has been going, I'm 
pretty much afraid, being a pessimistic sort, that there will be 
several "camps" of development. I fear that some might not resist 
the temptation to have things done in such a way that only one 
machine can profit - I hope this will not be the case, several people 
have told me that, even though they could act in this way, they 
agree not to do that. But, seeing how some of the debate here has 
become polarized, I do have this fear...
I thus see it as my main work to try to make sure that this doesn't 
happen. ALL OF THE REST, including this debate about the 
licence, is, to my mind, pretty much secondary - but it does show 
how deep the feelings run, and how difficult my job will be made 
because of them.

As to acting in an inclusive manner, I'm not sure what you mean by 
that. Do yo mean that I will try to include all proposed changes into 
SMSQ/E? YOU BET I WILL. I can go on record here for that.

But, to be quite honest, I must also state something that will 
probably make Richard howl with dispair: I don't believe that I will 
get many contributions. I also belive that most contributions I will 
get will be from MArkus Kingus, who has a record of supporting, at 
least, QPC, and also SMSQ/E.
I WOULD LIKE TO BE PROVED WRONG! Oh boy, how I would 
like to be proved wrong.
Bit I have had, until now, not one single suggestion of what 
anybody would actually attempt to change (not what they would 
like to see changed, but what they would guarantee that they 
would change).

> It is very clear to me that you accept input from some people on this list
> (Tim Swenson, for example) extremely civilly, and do actually take some of
> their comments on board. Others, you seem to dismiss out of hand, and none
> too politely. Perhaps you have a past history of disagreement with those
> characters; I know they can be rather blunt, but I can't say that I blame
> them for getting paranoid.

I don't thinbk that this is true. I don't reject anyone 'out of hand'.

> > Oh? What wider audience? DO you really mean that letting an
> > unsupported OS float around the shareware scenen would make for
> > a wider audience?
> 
> An open source operating system with as many useful and different ideas as
> this would get a wider audience provided that some essential difficiencies
> (filesystem, internet oriented applications etc) were sorted. So, perhaps
> not in its current guise, but if it were to develop the way that I (and I
> imagine, you) would hope, then yes. Otherwise it is dead, and all this
> discussion is just so much pointless hot air!

Well this is here we disagree. Things will have to be changed first 
before a wider audience will accept the OS, but in order to get 
things changed, it should be given to a wider audience

(...)
> > I know exactly how many : none.
> 
>  I don't agree -see above

Ok, so we have a fundamental disagreement here..

> However, I am an eternal optimist, and I do hope that a compromise can be
> found which will enable us to keep (get back) all our hardware and software
> developers - We just cannnot afford to lose them.
This is a question far from rhetorical: Have we lost them (in the 
sense that they would have been there otherwise?).

> Nasta's proposal that we split out a core common to all platforms from the
> rest, is worthy of consideration . Perhaps this sounds too much like the
> Linux to you (you taking the Linus Torvalds role as "custodian of the
> kernel", but potentially losing control of the rest)?

No, I don't see myself in that way. The question of what is a core 
and what not will give rise to much debate... However, 
fundamentally, isn't that what I'm trying to achieve - have one single 
COHERENT OS for all machines?
OK, so perhaps we don't agree on what is the "core". Is the Pointer 
environment part of the core? Is the filesystem? Is the scheduler? 
Is the Basic interpreter part of the core? Most probably not - but try 
to take it away, and see the outcry.
etc. etc. etc.
To make things easier, for me, the "core" simply is the OS as it 
stands now. Even though this may not be grammatically correct, it 
has the advantage of simplyfying the debate - something we can all 
use now.



>  and also please, please, please
> 1)  A definition from you of the sort of support expected of developers and
> resellers. Without it, it seems that much of your rationale for wording of
> the licence as you have,  vapourises.

No. The support, as I see it, must be supplied by the resellers. 
This is also one of the rationales behind having "official versions" of 
the source - at least I, or other developpers and the resellers, will 
have the possibility to test the new code!

> 2) the objective set of criteria that you proposet to use in determining
> which code gets included.

There is no objective set of criteria.
But the questions I'll ask myself will be simly as follows:

Is there any reason NOT to include it?
Subsidiarily:
It is useful?
Will it profit anyone?
Will it profit all machines?  If not, why not?
Is it likely to lead to incompatibilities?
Does it clash with the OS as it stands now?
Is it likely to clash with other developments?
Is the developper willing to fix bugs?
Is the code understandable?

> I personally believe that if those had been addressed early on, the paranoia
> would not have set in.

I'm not sure about that. paranoia doesn't need reasons to set in - 
not that I'm accusing anyone of it:
Wolfgang

Reply via email to