[Date Prev] [Date Next] [Thread Prev] [Thread Next] Indexes: Main | Date | Thread | Author

RE: [ba-ohs-talk] Ideagraph 0.5 release (+source)

>Danny, I understand this issue well. Perhaps I don't understand it well
>enough to analyze it in the context(s) it should be taken in, but I think
>that you, along with most other "hackers" who participate here, know fully
>well that there are more than enough open source libraries 'out there'
>that, whatever we do, it won't be all that hard for others to do
>it as well.    (01)

Well yes, indeed. For a lot of applications (my own included) it could even
work out easier for a developer to simply look at what it does and rebuild
from scratch. At a conceptual level, chances are someone's already written
up something that could be considered equivalent. Genuinely new bits of
lower-level programming are few and far between, and there are so many
potential opportunities in gluing existing functional modules in different
ways, it's rather hard to think of anywhere where traditional ideas of
intellectual property are much use.    (02)

Apart from cases where there is a particular niche involved (like
GhostScript and printers), I suspect the only kinds of real security for an
application/company in the markets come from big business (possibly
monopolistic) strategies or a slightly Leninist kind of continual
innovation. The (outstanding) "Cathedral and Bazaar" material [1] approaches
this whole situation well, but (as far as I've seen) though the Bazaar
approach works well for a lot of projects, and explains the success of o-s,
some of the more traditionally-styled *good* applications are considerably
less likely to come about without some element of "Cathedral" (it will be
interesting to see how Mitch Kapor's project fares).    (03)

I suppose I'm vaguely optimistic that there might be a (cue cliche) third
way, that of "Lone Gunman Programming"* (TM). Most of the really productive
hackers don't seem that bothered about having a ready-to-eat end product,
rather focus on getting over the technical hurdles. The LGP (cute acronym
anagram, all right reserved) approach would be to cherry-pick the Bazaar
stalls, and recombine this in a Cathedral style to produce something at
least oven-ready. The question remains how such a setup can be symbiotic
rather than parasitic.    (04)

In the case of my own project, fact is that whatever I might prefer to do
right now, I've committed myself to there being some closed (commercial)
aspect. This constraint is effectively new to me, and I still haven't really
figured out the implications. The hope that this constraint might mean the
bills get paid, well, that turns it into an interesting experiment...    (05)

* the public perception is that there's just one developer/group of
developers on the grassy knoll, whereas in reality a huge conspiracy of
open-source agents have done most of the work (this analogy falls down
somewhere around the Jack Ruby point)    (06)

>With that consideration (perhaps alone) in mind, I tend to shy away from
>notions of particular "not open source" licensing schemes (e.g. the Apple
>Public Source License).  Johnathan Cheyer gave a particularly lucid
>articulation about why he chose GPL and there may actually be some
>merit in
>using that license if it is your desire to be "open source" but, at the
>same time, prevent commercial ripoffs of your intellectual
>property.  Commercial ripoffs have never been my concern (though they do
>lurk at the edges of reality for me), so I stay away from GPL and now
>prefer the Jabber Open Source License because it goes beyond BSD
>in that it
>requires that you put back your improvements, bug fixes and so forth (but
>not necessarily your extensions), and it terminates your rights if
>you make
>any patent claims on users of the software.    (07)

A couple of new avenues to explore, thanks.    (08)

>Couple of EUROs for the day.    (09)

Buon Natalie,
Danny.    (010)

[1] http://www.tuxedo.org/~esr/writings/cathedral-bazaar/    (011)