Re: [ql-users] SMSQ/E Source Code, general

[wlenerz]

Hi all,

Thank you all for your patience whilst waiting for me to read and digest all of your comments. I hope most of your concerns will be addessed below. I now have had enough time to read all your replies, requests comments and criticisms very carefully.

I'll be replying to some emails in particular, to get rid of some points and fears expressed, they are also on this list. This here missive is just to set out the general points.

First of all, I must admit that the criticisms were less sharp than I feared that some might be, which indicates, broadly speaking, a general agreement with most of the principles set out earlier, if not necessarily with the implemenation of them. I recognize, of course, that some people are not happy at all and, whilst I also recognize that I can’t please all people, I still would like to try, even if, as Phoebos Dokos pointed out, this is not necessarily the point of this exercise.

I have more or less amalgamated all of your comments and criticisms, to avoid having to reply to each e-mail individually. I hope you all don’t mind. If I have forgotten to reply to a particular point you have made, please let me know.

Before I start out, I would just like to define the words ‘original sources’. These are the sources of SMSQ/E as I shall (hopefully soon) get them from Tony Tebby. ‘Official sources’ are those distributed by the registrat, which comprise the original sources and any modifications/additions that will hopefully be made.

Before going into the details, I want, agan, to stress a point. A pretty high degree of control has been retained over what will be an ‘official release’ and what will not be. Many people have objected to some aspect or other of this control. Please be aware that this wasn’t done because I, or anybody else, is a control freak, nor that I, or anybody else, revels in the “power” that control supposedly gives us. I am really concerned about the fragmentation of SMSQE. There are already 3 QDOS related systems out there:
- QDOS and QDOS Classic (as available on the Q40/Q60)
- Mineva
- SMSQ/E.
If we split this up even further, we will only divide our community even further, getting to the point where there will be programs that run only on one machne, and only on one OS on that machine. THAT IS SOMETHING WE MUST ABSOLUTELY AVOID.

Hence, the system of a registrar, whose main job, as I perceive it, is to attempt to bully the individual –and highly competent!- software authors to work together, so that one change on a machine is at least made in such a way that it also becomes possible on the others. The price for this attempt at coherence is a degree of control. I do not find this unreasonable.


Remember, the PC world has always had the motto ‘divide and conquer’. Let’s NOT fall into that trap.

Ok, now let’s get some of the legalities out of the way.

I know that some part of the future licence, as tentatively set out in the offcial statement, are a bit convoluted, and various persons have pointed out to me that a GPL licence would have been so much easier (especially for me...). That is true, but would have led nowhere, as several other people would simply not hear of it, since they want to retain some control over their own software. And there is the aspect of distributing the binaries, to which I will come later, which also made this impractical..

I also know that, as Timothy Swenson pointed out, the licence as set out in the official statement is much perfectible. Please only remember that the official statement is not the licence, which will be sharpened up a bit. However, I do not propose to make it into the multi page documents used by M$soft et al., for various reasons, one of which is the futility of it all. The licence is there to keep you honest. I can’t see anybody really suing over it, unless you want to make YOUR lawyer rich. (If it really came to it, I’d probably act as Tony’s lawyer, and that would not cost him any money – just so you know... (ha!)).

Moreover, if I really want to be perfectly legally coherent, I would have to draw up one licence per country. I’ll make it easier, and draw up one licence - and to make it more difficult for you, I may draft it in french... (with an english translation, perhaps). (ha, again!)



Copyright status of additions/modifications.

Some concern has been voiced (notably by Joachim van der Auwera) about the copyright status of the additions/modifications brought about in the future by various authors to the original sources.

The official statement said the following:
"
Authors retain copyright over their additions/modifications, but when submitting their additions/modifications, they agree that, if they are accepted in any official distribution (under the statements as set out above), the may be included in all other furture distributions (in other words, you can't submit something, which is included, and then some months later attempt to
withdraw it).

Let me clarify this. Copyright of the modifications will ALWAYS belong to, and continue to belong to, the authors of these modifications.

HOWEVER:

If you submit your modifications to be included in the official release of SMSQ/E, then you also give the registrar permission to:
- effectvely incude your modifications in the next official release
- effectively include your modifications in ALL FUTURE official releases.
- include in future versions a modified version of your code (so that changes, such as bugfixes, may be made to your code!!!)

This is made with the purpose of avoiding the problems that arise when I include a modification/addition in one version, and then not having it any more in the next version. I think you should be able to live with that. Of course, your copyright will be acknowledged, even if not specifically
(it will somewhere say:
Copyright (c) Tony Tebby
Some parts copyright (c) whoever
Some other parts copyright (c) somebody else
And so on.)

If I can, and if you REALLY WANT to, I can even try to be more specific.

MOREOVER:
When you submit your additions/modifications, you may tell me whether they are to be :
- distributed as part of the official release source code and binaries, which means that anybody can have a look at them, or
- distributed only as part of the official release binaries, that way you keep your source secret (except from the registrar).

As a personal note, I must add that this latter option is not completely to my liking, but concern about it was very clearly expressed in some quaters, so this option had to be included. I personally would prefer every change to be included in the source as well, but that's life.

The reason I don't like this "secrecy clause" so much is, again, coherence. Suppose any author make a useful addition to SMSQ/E, but asks me not to publish the source thereof. I would, of course, abide by this request. How do I then go about attempting to get this new feature onto the other machines?

Please also note that some parts, which are distributed with SMSQ/E are not part of SMSQ/E.. This, for example, is the case of QPC itself. QPC is the software that (roughly speaking) is the 68K emulator runing under wndows. SMSQ/E runs on top of it. Neiher Tony Tebby nor mself have any say nor right to QPC itself.


Resellers and binaries.

This is probably THE thorn in the side of most. Frankly, I anticipated more problems on other aspects, but the discussions seem to have focussed on this point.

To clarify the situation, this is as follows:

Source code:
You may get the official release, in source code form only from the registrar, i.e. me. Once you have the source code, you may compile it, or modify it or add to it. If you want, you can pass your modifications/additions to me (see above), but you don’t have to. You can keep them for yourself (why?).
You may also distribute the SOURCE CODE ONLY, provided that
-this is done entirely free of charge
-not done via a Website or FTP.

There was some grumbling about the “free of charge” bit. The reason is that we felt that nobody should make any money on Tony’s source code. The reasons set out by J.Grimbert in this respect are also valid:

"
For instance, the QLCF library has been running along similar lines:
- the requestor must provide both the media and stamped return package.

The reason might concern the income tax of the Librarian:
- if the requestor make a payement to the librarian,
the income of the librarian grows. Which mean more taxes for the librarian.
Therefore, the taxes should be reported to the requestor, but
why would a requestor pays the librarian's income tax ?
(Ok, if you want, you can pay all my income tax, you are very welcome...).
- Sending cash is forbidden by the postal rules here.
- Sending unlabelled cheque is dangerous
(the librarian could not label it to the seller of media, but might
label it to "Tresor Public", which is the tax collector. But paying
income tax with a lot of small cheques is rather strange and may be
not welcome by the tax people...)

So, avoiding any money transaction is the best thing.
And as a requestor, I know exactly where my money goes: back to me.
(because I got back the media, and the parcel... even if the postal service destroy the value of the stamps!)

"
The reason I’m doing it by avooiding the Web is that I hope that this will allow the registrar a bit more control over the sources, something I don’t make any secret of.


Binaries:

YOU MAY NOT GIVE AWAY, nor sell, the Binaries, i.e. the compiled version of the sources. You can only obtain them from the official resellers (more on which later).



Here is one issue which still must be resolved, and it concerns testing. The points raised in this respect by Peter Graf and Dave are valid, in that people who develop new additions/modification should be able to give test versions of the binaries away, so that they can be tested. If this could (should) be done before the changes are brought to my attention, the development cycle would be much faster.

I will come back to this point later this week. PLEASE BEAR WITH ME FOR THAT.

Resellers:
How do you become a reseller?
Ask me. (You can also ask Tony Tebby directly, but I would suggest that going through me you get a faster rply). I’ll coordinate with Tony Tebby and the other resellers.

In this respect, again, a justified comment was made, saying that it isn’t fair that those who are already resellers should have a say in who else can become one.

That is true, and it would also be perfectly illegal in the E.U. Please consider that, if anybody asks me to be come a reseller, I WILL ASK those already being resellers about their opinion. HOWEVER, the ultimate decision will lie with Tony Tebby!


Actually, the scheme will be as follows: If all, or the majority of the current resellers agree on the new reseller, there is no probem. If they do not, I’ll ask Tony Tebby, and his decision in that respect will be final.

To be quite frank, I don't really foresee any problem with that. Roy Wood has made quite a valid point about the support you are expected to extend to the user as a reseller - and that obligation of support WILL be part of the licence. I, for one, wouldn't WANT to be a reseller for SMSQ/E. If YOU want to be, ask me (I think Phoebus Dokos muttered something in this respect...).

On the other hand, Peter Graf has asked whether he can become a reseller – sure you can. As I said, consent to become a reseller will not be unreasonably withheld - I mean why should it be? Just remember, you will have to pay 10 EURO for each copy sold – you can pay this directly to Tony or pay it to me and I’ll pass it on to Tony Tebby.



I hope these clarifications are ... clear!

Wolfgang