thanks for a well considered reply. Apologies if in my haste I missed
some of the nuances of your points.
2011/12/31 Mats Lundälv <[log in to unmask]>:
> -----Open Source Assistive Technology <[log in to unmask]> skrev: -----
> 2011/12/29 Mats Lundälv <[log in to unmask]>:
>> Well, my comment to that quote and the trends it builds on would be this
>> "Our perspective, as always, should be that closed source and proprietary
>> methodologies in principle present a definitely disruptive, though in the
>> short term (and particularly for small minority groups) sometimes also
>> valuable, asset which we should persistently be trying to minimise in
>> of FLOSS".
> Hmm. if the value is reflected in high cost and low innovation for
> users through constant re-invention of the wheel then I say it's very
> limited value. As they say, let's minimise it. In my view that is the
> current situation in AT and the way out is wide spread collaboration
> between companies and projects.
>> In this perspective it is bad news that permissive licenses are pushing
>> forward at the expense of strong copyleft GPL kind of licenses. This may
>> some cases have good reasons, but is probably largely part of a strategy
>> infuential parts of the industry (reflected in the original 451 group
>> to fight the trend of FLOSS by temporarily and partly joining the too
>> trend to gradually dfismantle it by fighting its main successful tool for
>> long-term growth; the copyleft condition.
> The shift we are seeming and the the article highlights is that
> companies are beginning to realise that collaboration and open
> innovation is better for their bottom line than exploitation. I don't
> believe this is a wide spread conspiracy to dilute OSS licensing. Many
> companies will not go near GPL code, but the reason is more to do with
> the cost of extra due diligence and potential loss of their IP, rather
> than malice as you suggest. That said, plug in architectures can solve
> much of the risk of invoking copyleft on company IP
> Steeve, I don't live in any gloomy fantasy world of conspiracies. But I
> believe it is a clear and basic fact that there are conflicting interests
> here, and that different players act out of their actual and/or perceived
> interest, and there is a very obvious on-going struggle developing through
> different stages between these economic/political interests (also) in the
> domain of conditions for development and use of software, and now in the
> still growing sub-domain of free and/or OSS.
I agree about there being conflict of interests and complex
interactions. All to be expected in times of change (for the better, I
trust). However I also see evidence of some parties looking for ways
forward that resolve at least some of these conflicts.
> The success of the FLOSS trend has naturally attracked or even forced people
> and organisations who are pragmatically indifferent or basically sceptic or
> even hostile towards the growth of free software to act within its realms.
> This is a natural result of the success.
Quite - the alleged words of Gandi come to mind 'First they ignore
you, then they laugh at you, then they fight you, then you win' But I
don't see this as all out war with only one winner - as I now
understand is your view too.
> However it is naive to believe that
> these actors will fight to protect the freedom of software against their
> short, medium or long term interests if they see a chance to benefit from
> proprietary software hegemony in some domain. They will be likely to act
> within the FLOSS domain to minimise the threat of free software growth and
> maximise the opportunity to maintain or (re)gain proprietary software
While no doubt some will act this way, Others are exploring less
myopic views of how to get the best value for share holders. To some
extent unless we show some trust nothing will change.
> The existence and growth of liberal licensing is not a problem in itself. It
> may sometimes be well motivated, productive and causing no harm. But I
> believe that the balance between copyleft and liberal non-copyleft is
> important. If the liberal non-copyleft free software would push back
> copyleft software to relative insignificance, I think we may anticipate that
> this will at some stage create the preconditions for a quick and strong
> backlash in favour of proprietary and even more restrictive licensing and
> patenting rule in the software industry.
That would be unfortunate but to me it seems there may be an
assumption that proprietary is wrong, or worse. I personally think it
is misguided, restrictive and causes many problems from a users
perspective, and also for companies bottom line. I would much prefer
all software was FLOSS. But it *is* a way of licensing and funding
software development. The fact it became incumbent is a real problem
and I believe is now being explored by enlightened parties.
> I don't suggest that the now strengthening pressure against copyleft
> licensing and towards liberal non-copyleft licensing within the free
> software domain is caused by either "wide spread conspiracy" or "malice" -
> that's what you suggest on my behalf, please don't! I don't see this in
> black and white, as a struggle between evil and good.
My apologies, I did not mean to imply that was your position, just
highlight that taken to an incorrect extreme your words could be
interpreted as leading that that way. I think it is more the rise of
liberal, though of course that may be experienced as pressure given
limited resources (projects). I've no doubt both styles will be around
for a long time.
> Such views don't help
> explaining and understanding anything, but analysing conflicting
> business, socio-economic and political interests, as well as
> philosophical arguments, may do.
Absolutely. There is so much to be said for analysis and reflextion
and I see it happening in some quarters, It's a complex situation and
invokes much passion. I do find it interesting that software
development appears to highlight this so strongly. I guess it
indicates the issue is about people as well as code and business.
> I do agree the choice between licensing with liberal or copyleft
> depends on your attitude to the likely hood of leeches using without
> contribution. Counteracting that is the potential limitation of
> community size and health.
> So are liberal non-copyleft based communities by definition larger and
> "healthier" than copyleft based ones? A current example: Is the liberal
> Apache license based OpenOffice.org branch larger and "healthier" than the
> LGPL copyleft (though weak) based LibreOffice branch? I don't think that's
> evident, and I don't buy your picture there
I don't think the OOo/LO comparison is good. It is has a fairly
complex history with many parties having a wide variety of view
points. These have complicated and prolonged the process that OOo is
undergoing to adjust to Apache Incubation process. I understand that
LO was founded to address the closed community nature.of OOo and
copyleft was/is seen as a key part of protecting that. We need more
time to pass before we see a clear result.
The original Apache HTTP server is a better example. It was started as
a collaboration and is now used in many products and I think it has a
'healthy' community. Linux kernel is also a good example of healthy
GPL community - also being used in lots of commercial appliances
(though I'd like stats on how many of those manufacturers like Netgear
> This is an interesting example in another way, as it is a case where there
> are lot's of reasons to both defend the copyleft branch, but to do it
> without fighting the permissive branch and instead work for a good
> cooperation and sharing of tasks between the two. Both should have an
> interest in maintaining basic compatibility for core functionality, that is
> it should ideally happen in the Apache branch, but allow involved parties to
> make the reasonable choice of adding higher level functionality either in
> the proprietary (based on Apache) at their own expence, or in open
> cooperation and with shared resources (based on LGPL). I don't see the
> OO.org Apache branch as a great problem as long as the LibreOffice (weak)
> copyleft branch thrives.
Wise words. Open standards such as ODF are important for many reasons
- one is that a wide range of products/projects can implement them. An
open source reference implementation of a standard also adds value.
One possibility for ODF would be AOO as a reference liberal-licenced
'core' project and others, including LO, deriving from it and also
contributing back. Assuming agreement can be reach between all
interested parities and licence compatibility and suitable
architecture can be achieved. However I see several parties are
finding common ground almost impossible to adopt..
> But it would be a whole other business if
> LibreOffice would be fading away.
> For an interesting blog in relation to this see:
> In my view, the main problem with liberal non-copyleft licenses becoming
> dominant is not my or anyones attitudes against leeches; It is the longer
> term vulnerability of the freedom of software, as major players may more
> easily walk away with the code and major parts of the know-how back into a
> proprietary strategies as soon as seems favourable, and quickly leave
> wing-cut remainders of free resources behind in necessary development. The
> exit of main players will of course happen also for copyleft based projects,
> but the price for abandoning such projects will typically be higher.
I'm not convinced that would easily happen now that FLOSS has become
so well accepted. Especially if proprietary became dominant because
key players did it that way and others simply followed as it obviously
'worked'. I hope we've all moved on and a mixed economy will always
exist. Not least as it contributing offers commercial benefits too.
Hopefully enlightened users will have their say too, when they realise
open development means they have more options for use and support.
> I'm not convinced AT/a11y space is any
> different to main software, though there may be a general lack of
> awareness of the value in open source collaboration amongst the
> smaller vendors. We should be educating them and so enabling
> innovation for users. How many almost identical AAC devices do users
> want, when they could be some real innovation enabled by using a
> shared open source platform?
> I very much agree! This is what we in principle concluded already in
> 1989-1990 when we started the Nordic - British phase of the Comspec project
> - but long before we knew anything about the free software movement. We were
> thus far too early out have any real impact.
But your experience then is valuable now and adds weight. It's been a
long struggle but I hope we are starting to see benefits from the work
of you, Compsec and others. If AT users get great deal and innovation
> And the majority of the AT industry have all the time so far been generally
> anti free software. This is definitely not out of malice. They generally
> have very tight margins and tough conditions to survive, so you cannot blame
> them being cautious to enter into free software experiments risking to erode
> the vulnerable platforms they may have established in these often harsh
> conditions. But I'm not convinced it will be very much easier to involve
> these companies into contributing to liberally licensed projects than to
> copyleft ones
That's interesting - my thought was some won't contemplate using
strong copyleft unless it is clear their IP won't be effected, and so
never get to the point of contribution. If that is not an issue in
either case then the choice is likely depend on the company attitude
to 'softer' considerations such as self interest and collaboration. It
will be interesting to find out more.
There could be pros and cons for both from their points of
> view. Many will prefer liberally licensed or weak copyleft free software if
> others have invested and they could benefit from it for their proprietary
> development. It may be motivated to provide them with some of this e.g. for
> promoting support for common open standards etc., but this will not
> necessarily help involving them in free software community building and any
> real contributions.
> In OpenDirective's view if you want to have maximum usage and
> collaboration around code, including companies, then go liberal (eg
> BSD or Apache). If you want to stop the [perceived] leeches taking
> your code and doing something without contributing back got copyleft
> (eg GPL). If you have online service code that you want to protect
> from use then look at GPL Affero. Bear in mind many companies just
> wont go near copyleft code so it is likely to negatively impact the
> potential for community growth and so your sustainability (and as I
> said innovation for users).
> Wise companies value expertise and so will go to those who demonstrate
> it in their code. Once they come your governance model ensures no one
> party can take over, but all have a fair say.
> I find this analysis rather shallow and lacking many of the more fundamental
> perspectives, no differentiation between shorter, medium and longer term
> advantages and disadvantages for free software development etc.
Yes, it was a rushed response, highlighting one aspect. I see it comes
across as flippant. The point I was trying to make is that there is
currency in demonstrating you have expertise and some will value that
and will seek it out, look for ways to use that for their own ends.
Some are finding collaboration works as well as the methods that
traditionally come with a purely proprietary 'work style'. If we
accept that rather than a default position of looking for exploitation
only, then the risks you describe when not using copyleft are likely
to be less. True some parties will exploit if they can, even copyleft.
At least we now have court cases to set the presedence
> I also lack a discussion about the option of fair and constructive
> cooperation between free software and resources projects and proprietary
> developers based on dual licensing - instead of just choosing between one
> licence form or the other.
Agreed - see below
> As I wrote, I have good experiences from our work
> with Blissymbolics based on a dual licensing combination of free copyleft,
> and traditional commercial licensing for proprietary use (rather than
> considering some permissive licensing). It has not been difficult to argue
> for the reasonable and fair choice of either using free resources under free
> copyleft conditions or signing a royalty based license agreement for
> inclusion in proprietary packaging. The free resource policy has helped
> providing information and support transparently to all, very much
> appreciated not least by proprietary developers, and these in general find
> it very reasonable and not at all threatening to contribute back to the
> maintenance of the resource through some royalty payments.
That is good to hear, and I agree dual licensing is appropriate in
many cases.I'm slightly confused here as these are resources not code?
Though that effect the applicable licences and I guess you apply a
share alike style licence to act like copyleft. Also you previously
expressed concern with 'non commercial' clauses and isn't that what
you are using here? I may have missed something in your original post?
Perhaps you were highlighting the opportunity for problems?
> I also think that both your general argumentation here Steve, as well as the
> description and only examples given in the paragraphs about permissive and
> copyleft and strong and weak copyleft, in the first link to the OSS Watch
> resource below, present an implicitly favourable picture and arguments for
> permissive licensing, while at the same time a corresponding implicitly
> negative image with missing or partly irrelevant and misleading arguments
> for (strong) copyleft conditions - all presented as a "neutral" perspective
> over the field.
I possibly made a poor choice there. The following would have been
much better landing page given this context as it mentions and links
to Dual Licening
There are several suitable resources and I quickly picked one to go
with the licence tool.
Note that OSS Watch make every effort to ensure their advice is
impartial - I know having been present at many of the lengthy and
regular quality reviews. However there is a trade off between covering
all angles and the length of an individual note. There are links to
related resources so any one article can be treated as a starting
If you still feel there is a bias that could and should be addressed
in the original article, can I encourage you to contact the OSS Watch
team [log in to unmask] They will be only too keen to discuss your
concerns as quality is of great importance.
> The whole area is reduced to a matter of (short term) practicalities,
> efficiency and convenience. The fundamental issue of the importance of
> promoting and defending the long term growth of free software, and reducing
> the dominance of proprietary, seems to be banned, and if you bring it up it
> you will be implicitly or explicitly depicted as secteristic, fundamenalist,
> nourishing conspiracy theories, or being against wide cooperation and
> efficient innovation in general. I find this somewhat disappointing :-(
I really don't read this in the article at all. Inference based
omission can be tricky. Some may have take the stance you describe but
certainly not OSS Watch. Your points about different time scale
perspectives interests me though.
You touch on one *very* important point - reading a few articles or
using a tool will almost certainly fail to give one a good
understanding of the how the complex world of open development best
applies. Issues of licence, governance, community health, business and
long term effects etc are all complex people based issues and only
discussion and mentoring can help the most appropriate solution be
found. This is at the heart of OSS Watch services and now
OpenDirective too. It is also what is behind the Apache way as
project's in the Apache Incubator learn. A large project like OOo with
plenty of cultural 'baggage' needs a lot of time for the mentors to
learn and advise - hence my argument before about it not being a good
example of Liberal community health - not yet anyway :)
> With these somewhat gloomy reflections I wish you and us all a very Happy
> New Year!
Well in summary I think we both agree it's a complex situation and
there are pitfalls and traps for the health of FLOSS, both real and
projected. Discussion, example
Happy new year. I think it will be an interesting one for opena11y.
> Cheers, /Mats
> OSS Watch  have a number of good articles on the various types of
> licence, community and sustainability. And REALISE has a licence
> selection tool  that can be used to explore these issues.
> 1: http://www.oss-watch.ac.uk/resources/licdiff.xml
> 2: http://is.gd/AJUShz
>> So in my view we should work for free software, against too restrictive
>> counter-productive conditions like the non-commercial, but also for the
>> protection of the copyleft condition in all cases where there is not a
>> strong specific case for more permissive licensing.
>> I think it's generally better to use a copyleft license in combination
>> traditional restrictive licensing to allow cooperation with propietary
>> software developers, rather that just adopting a permissive license
>> proprietary software developers to grab free resources without giving any
>> contributions back. This is what we do for Blissymbolics, and it seems to
>> work fine.
>> -----Open Source Assistive Technology <[log in to unmask]> skrev:
>> Further to my comment on the rise of non copyleft licence use, here
>> are some figures from The 451 group
>> "Our perspective, as always, is that open source methodologies present
>> a potentially disruptive, and also valuable, asset that complements
>> the way both vendors and enterprise IT organizations conduct their
>> Our analysis indicates, however, that open source methodologies are
>> increasingly being employed by ‘complementary vendors’ with a leaning
>> towards more permissive licensing."
>> Steve Lee
>> Programme Leader (Open Accessibility)
>> OpenDirective http://opendirective.com
>> On Thu, Dec 22, 2011 at 12:04 PM, Garry Paxton
>> <[log in to unmask]> wrote:
>>> In the past few months, I've seen quite an increase in the number of
>>> people asking to use Mulberry Symbols in commercial apps, particularly
>>> Android and iPad/iPhone mobile apps.
>>> We need to make it more clear on our website (there is a ticket raised to
>>> update our FAQ) that the CC BY-SA allows developers to go ahead with
>>> commercial developments, without any further permission or alternative
>>> license from ourselves - the existing license already permits this.
>>> Further, we've just developed a new feature at the website that allows
>>> developers to download the symbol metadata (ie, grammar, category, and so
>>> on), to incorporate more intelligence into their apps, rather than being
>>> limited to the symbol name itself. Consequently, our symbols have a
>>> well defined naming convention. MindExpress, for example, is able to
>>> the symbol name, and recognising it as a verb from the accompanying
>>> metadata, can conjugate the name within their app.
>>> That info has always been available via our API, but many developers wish
>>> to incorporate the info within the app they distribute, so as not to
>>> upon wifi connectivity for their app to work.
>>> At the mo, that data is available to specific straight street 'partners'
>>> while we were finalising the data format. However, now that's been
>>> finalised, we'll soon be making that data available to all subscribers.