Hi all,
I’ve been keeping quiet on the licence front for some time, to let the
comments accrue a bit. Here’s what I have to say on most of them,
some of which resulted in some minor changes in the licence.. I
haven’t included the entire new licence, in view of the comments of
many of you that this debate is just too much and taking too much
bandwidth from this list – of course I will email the text to anybody
who wants it!!!!!
Payments: I’ve amended this part of the licence a bit to show who
pays what to whom, since this seemed to be a bit unclear in
RWAP’s mind. This now reads:
“The distribution of the binary versions of the software is made
against payment by the reseller to the software author(s). Each
reseller is free to set the prices he wishes for payment from his
clients/customers, provided however, that for each new copy sold,
an amount as determined in an annex hereto is paid to the
software author(s).”
The amount of money to be paid is determined in the annexe, it is
thus possible to set different amounts for upgrades etc… - if it
really comes to that.
(rest unchanged – please be reminded that the rest does allow the
resellers to provide for free upgrades, unless I require payment on a
new version). Richard had two problems with that:
“you read it correctly, however there are 2 problems.
- the license fee may change anytime and Wolfgang can then
prohibit free upgrades, the license explicitly says this.
- nobody (not even official resellers) is allowed to sell or
otherwise distribute modified (say bugfixed ) versions of the
binary”.
Both statements are correct as far as they go. The licence fee may
change at any time, if there is a reason for it (i.e. an author wants
money, and this is accepted). I can prohibit free upgrades. It is
possible that bugfixes are included in new versions of SMSQ/E.
However, there is absolutely no reason why these would then be
linked: every update does not entail additional payment. This would
only happen if there were bugfixes incorporated into a new version
that also has new features. I’m not sure that this is very realistic,
since one will mostly attempt to keep new features and bugfixes
separate. But the possibility exists.
Having it Richard’s way (i.e. bugfixes are always free) would
effectively prohibit the possibility of payment for new versions,
since any new (to be paid for) versions will incorporate all earlier
bugfixes. This is not what the licence tries to set out.
Richard also stated:
“we are not speaking about different versions, merely different
configurations. For technical reasons (mainly testing) it should be
obligatory that for every new feature added it should be possible to
exclude it from the built binary. Once you have this ability than it is
no extra effort.”
The emphasis being on the “once you have this ability”. This ability
isn’t there. It would require effort (yours, Richard?) to build it into
SMSQ/E.
Richard also stated:
“I think you are trying to make everything possible and this makes
the license extremely complicated and could cause some
problems for no good reason. Your above 2 points are good
examples of this.
If the code can be distributed completely separated than don't
make special provisions in the license for it. If it must be (for
technical reasons) distributed together with SMSQ as a single
monolithic binary than the author must accept SMSQ license,
anything else doesn't make sense for me.
What makes more sense is having parts of the code double
licensed, ie SMSQ and some other license. Again this does not
require any special provisions in this license, it is the natural right
of the
author to license his code under any number of licenses.”
This provision (i.e. that code may be distributed alongside official
versions) was included so that people (like you, hmmm?) may have
their code ”close to” an official version, even though they object to
the licence, as i heir code would not fall under the licence. I agree,
that this is bending backwards, but it is a compromise solution. I
take note that you don’t even want that.
RWAP wrote:
“Just noticed a problem with the licence - under the heading
Compilation of Source Code, this should read:
Any person may compile the source code and thus obtain a binary
(compiled) version and use it for itself. However, binary versions
may NOT be distributed other than pursuant to clauses C, D and E
below.”
I’ve amended the licence on this point to read:
“
Compilation of source code
Any person may compile the source code and thus obtain a binary
(compiled) version and use it for itself. However, unless otherwise
provided herein, binary code may NOT be distributed.”
RWAP also wrote:
“However, what if you could not be contacted or needed to clarify a
bit with the original author - this could delay matters substantially.
I agree that the code should still be submitted to you and approved
for inclusion, but on the other hand, why should something specific
to the Q40 need your approval
if it is of no use to any other version (for example)??”
Well, pray tell me how can there be this big important bug that
needs to be fixed within the next five minutes in an official version?
The purpose of official versions is (also) to try to test them
beforehand; and I do hope that the software author(s) have also
made some tests. Now, this doesn’t mean, of course, that official
versions will be exempt from bugs, but (hopefully!) not something
vital, and, apart from the times I’m on holidays, I will generally be
easy to reach …
The problem of software authors disappearing:
This is a problem that is only acute when the source code isn’t
supplied. In this case, this will be pointed out.
Finally, Ian Pine wrote:
«zzzzzzzzzzzzzzzzzzzzzzz»
This is duly noted….
Wolfgang
