[erlang-questions] A PropEr announcement

James Churchman <>
Fri Jun 17 13:36:24 CEST 2011


Dear Ulf

I agree almost entirely with the sentiment of your email. I think that it is quite easy to be in a position where you are funded by external bodies, in the case of proper being funded by a large university and also EU grants, and then afford to give the software away for free. I also think it makes it a tad rich to release it under something as restrictive as the gpl given that it's the profits of the commercial companies, by virtue of paying taxes, that have in effect funded the software, tho it is of course in the end entirely the decision of the original authors.

I however take a different take on the mater of open source and startups, especially in the case of developer facing tools. Taking the case of QuickCheck as the perfect text book example, it is crazy to turn up with a fantastic product that costs thousands of pounds & is totally closed to any contributions, present it to hundreds of the cleverest programers in the world and then expect them not to copy it when it is that good and also not something beyond the abilities of those people to build themselves. In the world, which is nothing but a realised version of game theory, this is crazy.

Also how long was this meant to continue? is it realistic to think in 10 entire years nobody would produce a copy, or 5 , or 3 ? There was an obvious huge demand and the barrier to building one far lower than that huge demand.

What it shows is that you want the benefit of these other people contributing to your product, essentially out-sourcing half the costs and innovation, and then sell some ancillary thing. If you create the tool you should know the most about it, and even if you don't people will approach you first. The world is a moving place, especially the world of IT. ( I would also argue that in many regards this is the basis Erlang Solutions operates on, Erlang is open source, and it sells its expertise in erlang whilst contributing somewhat back to erl/opt)

Also in the case of QC i would assume that 95 % of people who wanted to use it did not, entirely due to cost, and that 100 % of people who wished to contribute to it could not, and by the pure definition of the tool all 195 % of those people were ALL developers.

Therefore if you look at the biz model that by contrast Process One etc. etc. etc. follow they provide the software totally open, and in their case i think they were further clever picking up on an already popular open source project rather than an untested home grow affair, and this allows students that might be building the next Facebook or Google to use the software and pay for what they need in the future if / when they need to pay for things as well as when they have the cash to do so. I doubt ejabberd would have 1% of the users and even that process one would have a profitable biz if it was a closed up shop, tho that is a difficult one to back up.

I don't think this applies to all products, but it seems to apply to many developer tools in the software industry. Other products probably less so, as its harder to coordinate say the design of a fantastic GUI between hundreds of dispersed developers, and for many consumers they are not aware of the open alternatives, don't mind paying a few $ and are not comfortable compiling code etc.. but i think this is an ok situation as consumers don't want & would not pay for training, support contracts, custom extensions to be written etc.. for their copy of photoshop elements.

And you want to start with that model reasonably early, then hope to make it a successful business, not have the ticking time bomb of when something 0 cost, open source and in many cases far far superior come out & you have to change everything, as well as you becoming the awkward non-industry standard option.

In the case of property based testing, i would imaging that there is a far far far larger biz around supporting PropEr, if it becomes the industry standard teasing tool that every single erlang dev uses, than there is in selling the software outright, as writing good test can be tricky, so i hope that Quviq works this out and do prosper greatly from this opportunity :-)

James

On 17 Jun 2011, at 11:43, Ulf Wiger wrote:

> 
> It has been brought to my attention that some have interpreted my posts in this thread as gunning for Kostis. I am not. I sincerely hope that no one has been on the wrong side of the law here (and even more that there will be no attempt at legal action). Knowing Kostis, I am pretty certain that he had very carefully considered what he was doing.
> 
> I think it is extremely important to have a discussion about not just what is strictly legal, but also how we as a community act to show respect for other people's right to innovate and make money on their innovation.
> 
> OTOH, Kostis definitely could have written a quickcheck-like tool with a different interface, staying absolutely clear of even the slightest risk of being accused of infringement. Safer for Kostis, but actually worse for the community. I believe it is better, ultimately for all parties involved, that the APIs are almost fully compatible. I hope that Quviq will find a way to profit from this too, not least since they have done a tremendous amount of good for the Erlang community.
> 
> Uncritically applauding acts of essentially cloning other people's IPR may have unwanted consequences.
> 
> Let's consider for a while that one of the main driving forces behind the current patent system is the Pharmaceutical industry. Why? Much because it is extremely expensive to develop and approve a new drug, but very cheap to copy it. 
> 
> It's clear that OSS is pushing the point at which one can charge for a software component higher and higher. Today, it seems, if you don't have several man-years of effort invested in your software, you might as well not even bother, since as soon as you launch, chances are that someone in the OSS community might decide to build a free version just as a fun exercise.
> 
> I can't help thinking that this sort of activity is mostly self-gratifying, and doesn't take into account the fact that the Open Source community has yet to figure out how to be self-sustaining. Much of the money that funds Open Source (or at least pays the rent for those who contribute to it) comes from largely proprietary companies.
> 
> What OSS has so far demonstrated is that giving things away for free can be extremely beneficial for innovation and quality. On the topic of ensuring steady revenue and managing to pay people's salaries at the end of each month, it has some way to go still.
> 
> BR,
> Ulf W
> 
> On 16 Jun 2011, at 09:28, Ulf Wiger wrote:
> 
>> 
>> On 16 Jun 2011, at 09:01, Joel Reymont wrote:
>> 
>>> 
>>> On Jun 16, 2011, at 10:43 AM, Ulf Wiger wrote:
>>> 
>>>> As far as I can tell, PropEr is an unauthorised reverse-engineering effort of the later commercial versions (the early version did not have shrinking or statem support, for example);
>>> 
>>> How do you tell it's a reverse-engineering effort as opposed to a clean-room re-implementation?
>> 
>> First of all, my intent was not to pass judgement (IANAL), only to point out that there are other legal subtleties regarding PropEr. But since the point of expressing an opinion is mainly to stimulate discussion, I dug into the legal text in order to try to answer your question.
>> 
>> The relevant EU legal text [1] says, under Article 4 - restricted acts:
>> 
>> "(a) the permanent or temporary reproduction of a computer program by any means and in any form, in part or in whole. Insofar as loading, displaying, running, transmision or storage of the computer program necessitate such reproduction, such acts shall be subject to authorization by the rightholder"
>> 
>> This seems to imply that if it happened as Kostis said - that they wrote PropEr based on available tutorials, articles and documentation, never having been in possession of a copy of the software, this will not constitute a violation of said directive.
>> 
>>> Is QuviQ QuickCheck rocket science that requires decompiling bytecode?
>> 
>> It doesn't have to be. It only has to fall under the laws regarding copyright protection.
>> 
>> "3. A computer program shall be protected if it is original in the sense that it is the author's own intellectual creation. No other criteria shall be applied to determine its eligibility for protection."
>> (Article 1)
>> 
>> But the more difficult part is perhaps what parts are protected in the first place:
>> 
>> "2. Protection in accordance with this Directive shall apply to the expression in any form of a computer program. Ideas and principles which underlie any element of a computer program, including those which underlie its interfaces, are not protected by copyright under this Directive."
>> 
>> A possible interpretation of this is that releasing any implementation that almost perfectly emulates the API of a protected program *does* violate the directive, much as a reproduction of a painting would. What is hard to reconcile is whether it is still infringement if the copy was made without ever having been in possession of the original. :)
>> 
>> Presumably, this would also mean that the PropEr authors never used QuickCheck Mini - also a protected program.
>> 
>> BR,
>> Ulf
>> 
>> [1] "Council Directive 91/250/EEC of 14 May 1991 on the legal protection of computer programs"
>> http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX:31991L0250:EN:HTML
>> 
>> Ulf Wiger, CTO, Erlang Solutions, Ltd.
>> http://erlang-solutions.com
>> 
>> 
>> 
>> _______________________________________________
>> erlang-questions mailing list
>> 
>> http://erlang.org/mailman/listinfo/erlang-questions
> 
> Ulf Wiger, CTO, Erlang Solutions, Ltd.
> http://erlang-solutions.com
> 
> 
> 
> _______________________________________________
> erlang-questions mailing list
> 
> http://erlang.org/mailman/listinfo/erlang-questions

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://erlang.org/pipermail/erlang-questions/attachments/20110617/5048cc03/attachment.html>


More information about the erlang-questions mailing list