KOSH


Translate
KOSH recently received an email from the Blue Wizard - nope I haven't lost my marbles (although I am playing Baldur's Gate). Mr Wizard sent in a mail which actually mirrors the questions that many have about KOSH. It is nothing special, but does bring many strands together. Thus I though basing this report on it may help to further the cause.

Here goes...

>The problems that I see are several-fold:
>
>(1) Needs a cohesive vision plan and/or a roadmap
>
> Linux works precisely because there is already a roadmap
> --- namely the earlier commercial UNIX OSes, plus that a
> sufficient number of people have used it (e.g. from the
> college days). So for example you could not go "wrong"
> with implementing a clone of 'cat' or 'awk' (which is 'gawk').
> KOSH would need to have something "tangible" that everyone
> would understand and aim for. This can usually be done via
> a combination of prototyping and description. This is
> difficult due to the fact that it is hard to communicate
> over Internet how things should be done.

This is very true. However, Linux works precisely BECAUSE it is working from a previous roadmap. KOSH is a clarion call to all those who are tired of walking the same old trails over and over again. The plan for KOSH isn't that mysterious or hidden. KOSH goes back a good few years, but has those roots in informal talks, concept projects and general experience of and disatisfaction with the products and projects being developed at that time.

The other point to understand is that KOSH is aiming at creating an open ended virtual environment rather than a tighty controlled and created operating system. The whole purpose of adopting an object sea approach is that third parties can themselves build the objects that populate the sea. We intend to provide the necessary structure and services to allow this environment to exist, and free up the developers to create for themselves the world that they want. The plan for KOSH is coming together nicely, albeit in a time constrained manner, as many of us have other lives. This is one reason why the option and opportunity for a more commercial based entity is allowed within the framework of KOSH, again though in an innovative manner.

>(2) Politicization of community
>
> It is difficult to avoid have politics, especially in the
> election of speakers and such. I personally have seen
> plenty of it, having lived in the Washington, D.C. myself
> .

KOSH stands for innovation in organisation and operation, as much as it stands for innovation in technology. The above is very true, but we have to move beyond that, both as individuals, and as a community if we are to progress. The community has a huge amount to contribute but creating a mechanism to allow that to interface directly to any research and development would cripple KOSH. Similarly, KOSH can be divided into applications users and hackers (in the proper sense of the word). One group wants an invisible OS that gives them the apps they want at the performance they want and allows them to configure to their hearts content. The other group wants to write its own objects, pull the object sea apart and create, modify and advance as they see fit. The suggested structures cater to both.

More importantly, KOSH belongs to all. If ppl want to start playing politics then they will damage the very heart of the community and the community, if it cares will take action. Otherwise we have nothing left but anarchy or dictatorship. KOSH aims at a benign, minimalist dictatorship embedded in the middle of an autonomous meritocracy - much like an object sea ;-) Will it work? I guess we'll find out soon enough, and if it doesn't then we can always reinvent ourselves. We don't suggest that we know everything or are infallible, just that we all want a better way, and both trust and participation are the key to success.

>(3) Fleecy proposes some form of "work for virtual credit" system.
>
> The problem with this is that how would someone's work be
> evaluated fairly (credit-wise)? In traditional commercial
> works, the market basically decides the value of that software.
> Fleecy seems to have discarded the marketing mechanism, and
> instead proposes that the work be "paid" by the hour at a
> flat rate. The problem lies in the metric of the work itself,
> namely how does a peer fairly determine how much work has been
> done for a given assignment? Once again, Linux works because
> it is free, relying instead on notion of gift culture, as
> described in "The Cathedral and The Bazaar" paper.

In a perfect world, we could all work for free, getting money to live and being able to work on what we want for the benefit of all, a la Star Trek. Unfortunately (or fortunately, depending on your point of view), we do not live in such a world.

I personally am a firm believer in ppl being rewarded for work that they put into a project. This is especially true in such a skill heavy area as platform design and development. We all know many ppl who have given many hours of service for no pay to maintain communities in which we have previously been involved. Many are willing to do this again. However, for KOSH to succeed as an organisation, we need to reward those who put effort into the community, in whatever manner. That is why it has been divided into 4, and that is why the rewards for developers are a fixed but substantial amount. It gives those who have put hours into the project say in the future direction, as much as those who invest money and loyalty as third party developers and users.

Flat rate is better than graduated rates for precisely the reason you mention - namely who decides if one piece of work is worth more than another. I don't want to start doing that and I am pretty sure that no one I know wants to do it either.

Measuring hours is better because no matter how much we can do in an hour, we all have only 24 of them a day to spend. Of course there are safeguards in this. For the majority of contracts, a number of hours will be specified, thus eliminating someone claiming millions of hours for something that took five minutes. For those without estimates, we ultimately be falling back on a unique set of concepts called trust and honour.

Also, nothing is fixed. Whilst hours will be logged, it is upto the first Council of KOSH to decide what to do and how to convert logged hours into shares or VCs. Again, I don't pretend to have all the answers and this may change over time, but as you intelligently point out in your first questions, KOSH needs a direction, and an open source system will not give it that - it will give it great code and adaptability, but direction needs centralised focus, or it is left to the very politicing, fragmenting and drift that you point to as being a danger.

>(4) Problem with valuation of virtual credits
>
> Who will fairly evaluate the value of the virtual credits
> upon redemption? Do the credits vest? It seems that this
> system is banging against the real world economy. Also, it
> seems to me that the mixing of various types of credits (Red,
> Green, whatever) seems to be inevitable (since one would like
> to develop third party stuff as well as the KOSH parts itself),
> which would complicate the redemption. Also, there is another
> problem with the virtual credit system which I will discuss
> next.

The problems that this system tries to address are ;

1 - to advance we need capital investment of sorts
2 - the ownership of KOSH needs to remain forever in the hands of the community
3 - one part of that community should never be stronger than any other
4 - money does indeed, at the moment, make the world go around

The value of virtual credits will be done by the elected Council of KOSH. No one holding credits is forced to convert them at that rate if they don't want to. Yes, the credits vest in line with the annual level of US inflation. There is no mixing of shares. No single commercial or non commercial user can own any Green or Blue shares - they are an artifice to give control and dividend to those abstract groups. Individuals can hold red shares but then only for work done on KOSH at the invitation of the Council of KOSH, and these cannot be sold to anyone but back to KOSH, which means to others who do work on KOSH. The only vaguely normal shares are the Yellow shares, which have to be that way in order to encourage standard investors.

Obviously the system is an artifice, but one designed to fight against the pressures and preserve the ideals and spirit of KOSH. Feel free to suggest an alternative, but it must be one that also achieves the same goals. Adopting the Linux model will produce a Linux community, and that is NOT what KOSH is about.

>(5) Creeping corporatization
>
> I must emphasize at the outset that this is not to be confused
> with the corporationship, or even with the corporatism itself.
> I use this phrase to emphasize that as it stands, it will
> inevitably begin to acquire various trappings of a corporate
> as it grows. Take for example the USA law mandating the report
> of income and such. Normally, this would require corresponding
> tax forms from companies showing how much income et al. that the
> employees get. For KOSH, this will eventually require
> paperworks, and that means adding new group (or staff, whatever)
> to keep track of these things. Things like that will inevitably
> push KOSH toward corporationship, which is not a good idea in
> my humble opinion.

This may be true, and it may turn out that we need some of these features. For instance, a professional 24*7*365 user support network, or developers to work on a particular feature (although this may end up being contracted). KOSH will need to exist as a company, much the same way Red Hat does, if it is to be the single source of KOSH for the non Hacker members of the community.

The key word you have defined though is "creeping". If it is a concern, they we have to make sure that it doesn't creep up on us, but that any and all steps taken towards corporatism and done because they have to be done, and because they give us an advantage that cannot be gained in any other way.

>(6) Excessive wealth
>
> I understand the need for reserve of wealth at the outset,
> but what if the sales were huge? Where would the money go?
> Once again there is a good chance of that being politicized.

Again very true. If we make a huge amount of money then what do we do? A quarter is a reward to those who gave their time to KOSH freely. A quarter is for the standard investor and half of it is for use in improving support, documentation, providing events, loans to developers, marketing and advertising, paying for KOSH systems to go into schools in poorer countries. Of course, the other option is to reduce the price......... ;-) >(7) Ownership
>
> This is an essential part of the work. Commercial companies
> writing software normally 'own' it. Linux relies on the GNU
> Public License (commonly abbreviated to GPL), which
> basically declares that no one can take ownership; it belongs
> instead to "GNU" (a phantom entity that constantly accumulates
> such works ;-). Fleecy, however, instead envisions KOSH as
> belonging to a certain community (the 'K' in KOSH conveys
> this notion). The problem with this is that it is very
> ill-defined. Exactly what constitutes a community in legal
> sense? For example, does someone, having downloaded KOSH,
> become a member of the 'community'?


Our definition is either someone who works on developing and maintaining KOSH and/or someone who buys a KOSH system. This way, it is in the hands of those who have at least invested some effort, money or time, and who continue to do so. However, each entity only gets one vote, to prevent a company for instance from buying a million copies.

Ownership is the most dangerous part of it. If we don't have ownership and go GNU then we risk losing focus, a core plan and falling prey to fragmentation. If we do have ownership then we have to build elaborate safeguards and protections to ensure that that ownership can never be taken away from the community. The benefits of ownership, as far as I am concerned outweight the risks but it is a community decision.

>I basically have too many questions with this, and I feel that
>by going with Public Licensing many issues would be solved.
>
>I personally feel that Fleecy wrote this KOSH proposal more on an
>emotional level rather than on a logical level, especially so soon
>after his layoff.


Funny you should say that - I have had so many ppl telling me that it should be done as a proper company, it should be done GPL, that I am splitting the Amiga community, that I am saving the Amiga community - seems there are many opinions, and that is good, but KOSH will move ahead and the community that forms around it is the community that forms around it, for whatever reason....and if we get it wrong, then we will evolve.

As to the layoff bit, it is a valid concern, but anyone who knows me, or has followed the JMS, the ICOA, my time at AInc or OASYS will know that these are ideas that have been developing for years. This is no tantrum or spat. It is a final realisation that we have to grow up and take control of our own lives and destinies or we will forever be whining moaners......that's not for me, and judging by the mails, it is not for a lot of ppl either.

Great mail, Mr Wizard, and none of this is meant to be personal. It is just my beliefs, and of course, any others in the community are as valid.

This is who we are.