[License-review] For Legacy Approval: TOPPERS License

[Carlo]

Dear Yutaka,

some replies inline.

Something went wrong with the mailer, the thread is not really readable
as per the standard way of nested discussion.

On 22/07/2015 08:07, Yutaka MATSUBARA wrote:
> Hello Carlo and other members,
>
> Sorry for my late response.
>
> On 7/9/15 20:53, Simon Phipps wrote:
>> Hi there Yutaka,
>>
>> On Tue, Jul 7, 2015 at 8:37 PM, Yutaka MATSUBARA <yutaka at ertl.jp
>> <mailto:yutaka at ertl.jp>> wrote:
>>
>>     Hello Carlo,
>>
>>                  (a) The above copyright notice, this use conditions,
>>             and the
>>                      disclaimer shown below must be shown without
>>             modification in
>>                      the document provided with the redistributed
>>             software, such as
>>                      the user manual.
>>
>>
>>         The example is clear, but why is it not sufficient to
>>         accompany any
>>         distribution in such a form with a copy of the license itself?
>>
>>
>>     The form with a copy of the license itself meets (a) clearly.
>>     Do you have a concern?

So why you simply don't use one standard language (like the one in the
GPL v.3 for instance) saying that any distribution must be accompanied
with a copy of the license, and be done? I understand sometimes you have
to deviate from standard practice, but here I don't see a rationale, and
any unnecessary complication can lead to unnecessary confusion.

>>
>>                  (b) How the software is to be redistributed must be
>>             reported to the
>>                      TOPPERS Project according to the procedure
>> described
>>                      separately.
>>
>>
>>         If this was the only option available, this would make the
>> entire
>>         license strikingly non-compliant, IMHO. In addition, reference
>>         to a
>>         specific project ought to be omitted, otherwise the license
>> would
>>         practically be limited to a single use -- a one-way free
>> ticket to
>>         proliferation.
>>
>>     > I would advocate that, because the condition under b) is
>> non-compliant,
>>     > it should be removed for the license to be approved.
>>
>>     I think that if a user can do (b), then the user can do (a) as well.
>>     What kinds of situations do you imagine?

This is what I mean. In any case, b) is useless, and why give an
exemption to a), which is the standard way to ensure certain notices are
preserved, to those who comply with b), which does nothing to foster
openness and freedom, and is uncompliant with OSD as well?

Therefore, I advocate that it must be removed.
>>
>>
>> I am still concerned that the license could result in situations where
>> a developer is unable to ensure a document is included with the
>> software they are distributing (imagine a portion of the code is
>> excerpted and used in an embedded function in a hardware component
>> used for constructing other products). In this event, (b) would be the
>> only option and it is clearly not OSD compliant. As such I don't
>> believe the license as a whole can be OSD compliant.
>
> I understood that the event (a portion of the code is excerpted and
> used in an embedded system) can be occurred. But, other OSD-compliant
> licenses such as BSD Licenses could result in the same situation
> (where a developer ...).
>

BSD is not the best example IMHO, and in this case it is ambiguous as to
what happens if no documentation is distributed with the software. But
one can argue that in this case only the copyright notice should be
preserved somewhere.

Just to show how BSD was bizarrely written, the advertising clause was
conceived to apply even separately from the actual distribution (and for
how long?). Wisely it has been recalled.

Anyhow, this is a sort of ad hominem argument, I don't like to follow
this way of discussing.

> If your concern is common, I guess that a developer is also unable to
> ensure a document is included with the software in case of BSD
> Licenses. What do you think about it?

Ditto. I can read that license in a way that in this case it is exempted.

To be clearer, I do not think that requiring that the text of the
license is distributed along with the software is non compliant with
OSD. I speculate about the fact that in some environment it encourages 
following a different rule, which is non compliant.

>
>> Remember, open source code is frequently excerpted outside the scope
>> of the original project, even if you are not aiming for that. In
>> addition, OSI approves licenses for general use not just for your use.
>
> I understood the above intention.

Fine
>
>>     Several licences including names of specific projects/companies
>>     have been already approved. I think that these licences seem to be
>>     limited to use cases. What differences between these licenses and s
>>     TOPPERS license do you see?
>>
>>
>> Indeed, licenses were historically approved with specific names in
>> them. Since the anti-proliferation discussions OSI hosted nearly a
>> decade ago, we have been avoiding new licenses that refer to specific
>> projects. The license should be genericised for anyone to use before
>> being approved.
>
> Is it an official and unified comment from the OSI committee that OSI
> does not approve new licenses that refer to specific projects? Please
> let me know where it is noticed in the OSI website.

I think this is clearly stated many times in this discussion list. But I
think this is a matter best left to the Board to discuss with you.


All the best

Carlo