Re: Mathworks-hosted GPL'd software

[Jaroslav Hajek]

On Tue, Feb 16, 2010 at 4:44 PM, Judd Storrs <address@hidden> wrote:
> On Tue, Feb 16, 2010 at 5:37 AM, Jaroslav Hajek <address@hidden> wrote:
>> It's a little long for my personal taste. Can something be dropped?
>
> Please. I just started writing and sent what I came up with when I was
> tired of editing. An intermediate version was much longer :) I wasn't
> sure where to cut further so I thought to get some feedback.
>
>> "paying Matlab customers" -> "MATLAB users" or "MathWorks customers"
>
>>
>>> For this
>>> reason we are establishing a secondary repository for user-contributed
>>> software and invite you to submit your files there also.
>>
>> "secondary" -> "alternative" ?
>
> Yeah, secondary emphasizes derivative. I was trying to emphasize that
> contributors could participate in both sites. Another word that come
> to mind is "independent".
>
>> Numerical systems/numerical environments. Maybe we can just speak of
>> Octave?
>
> Technically, the ban affects users of Octave, Scilab, Numpy, R, IDL,
> GDL, Freemat, SAS, SPSS, PSPP, S+, C, C++, Fortran, FORTRAN, etc ad
> nausium.

That's why I'd like to stick with Octave.

>
>> Also, I think it's not fair to say that MathWorks is trying to
>> undermine the community of Matlab users. They are just trying to limit
>> the service to Matlab users,  which even Matlab users may not be happy
>> with.
>
> It's not unfair. They are trying to limit our association with their
> customers and balkanize the community. The losers are the customers
> and us.
>
>>> Briefly, we
>>> believe that new restrictions mandated by the Mathworks on licensing
>>> of files hosted in File Exchange may limit your ability to share and
>>> derive benefit from your work
>>
>> As we previously agreed, the author's ability to share his work is of
>> course not affected. We shouldn't say this.
>
> I disagree. The inability to choose the GPL and use File Exchange
> simultaneously is a limitation on the author's ability to share.
>

But only to share through FX, not to share his work in general.


>> I was sort of assuming that these conclusions (resulting from the BSD
>> license enforcement) will be mostly obvious to developers. After all,
>> they either chose BSD themselves or agreed to a change, so they
>> probably know what's going on.
>
> Never assume anything. I would assume a GPL hostile audience.
>

Authors hostile to GPL and Octave won't help us, so we can as well
just ignore them.

>> Although Tim Davis' example is instructive (and worth following), it
>> also makes the whole text longer. I would at least start a new
>> paragraph with the words "For Example", so that the reader may skip it
>> more easily.
>
> Well, my opinion is that a new paragraph would help readers skip to it
> not away from it. If they don't to read it they can continue to the
> next paragraph.
>
> Also, something is bothering me at the moment. I mentioned Dr. Davis
> by name. Ultimately, we should ask his permission first before sending
> a letter with his name. I was just experimenting with trying to make a
> case for being distrusting of Mathworks while staying factual. I don't
> think the audience is going to be as receptive to "GNU intrusion" as
> you do. i.e. I doubt they will understand what we mean when we use
> short-hands such as "free software" So be careful about this.
>

I think it's generally known what a "free software license" means. If
someone doesn't know, there's Google and Wikipedia.

>>> Secondly, the Mathworks has updated the MatlabCentral Terms of Service
>>> to ban all use of the website not related specifically to using
>>> Mathworks products. It is not clear what the intent this change
>>> signals about the future of the File Exchange within the Matlab user
>>> community. At the very minimum it appears that Mathworks intends to
>>> isolate Matlab users and contributors from the knowledge, feedback,
>>> and experience of users of Matlab competitors. But the Mathworks may
>>> also intend to limit access to your contribution to customers--the new
>>> Terms of Service may go so far as precluding users of other systems
>>> from viewing, studying and adapting your software.
>>>
>>
>> Too much speculation, I say. Let us not make guesses what MathWorks
>> may or may not do in the future.
>> I think we should focus this on the simple primary concern: the new
>> restriction in the ToS you have discovered. Whether or not it is
>> actually legally effective, it was surely put there on purpose and
>> therefore we are rightfully concerned about getting into trouble by
>> violating it.
>>
>> I say, chances are some authors will not sympathize with us and alert
>> MathWorks about the action; some authors are even MathWorks employees.
>
> Newsflash, I bet some Mathworks employees subscribe to
> address@hidden There is absolutely no secrecy. Did you think
> Mathworks would not find out about a secondary archive?
>

I think they will not care about this unless they get complaints from
the authors themselves. That is why I think it's best to avoid
attacking MathWorks at all, just provide the fact (the restriction)
and why we think it matters to us.


>> I think we should reduce all speculations and campaign-like statements
>> to a minimum, if not zero, because they could be eventually used
>> against us.
>
> I though you called us to action for a coordinated campaign. My bad.
>

No, I don't think I wanted to start a campaign.

>> I assume most (if not all) of the developers are
>> intelligent enough and will easily understand what are we trying to do
>> and why, based on the bare minimum of information we provide. The
>> point is to alert them about the ToS restriction, because many of them
>> are surely not aware of it. One thing I usually dislike about
>> commercial software companies is that they tend to speak to users like
>> they were idiots.
>
> I'm very sorry, but I *am* an idiot. I'm not a genius operating at
> your level. You're going to have to spell it out to me if you want me
> to follow your reasoning.
>

Nonsense. You know very well that you are no idiot and you don't need
it spelt out. Unless you show me a counterexample, I'll consider the
existence of idiots amongst the target audience to be purely
hypothetical :)

>> A purely technical matter is that I would like to stress that the
>> authors may simply send their package by email in response. Many of
>> them will surely not want to maintain their code at two places and
>> will remain loyal to the MathWorks site, so we should minimize the
>> actions required on their part.
>
> Also as a purely technical matter, we don't actually need anything
> from the authors at all. The BSD license allows us to just take and
> host the code.
>

... and possibly violate the ToS by doing that. There is no need for
you to participate if you think this effort is pointless.

-- 
RNDr. Jaroslav Hajek, PhD
computing expert & GNU Octave developer
Aeronautical Research and Test Institute (VZLU)
Prague, Czech Republic
url: www.highegg.matfyz.cz