On Tue, 2004-02-24 at 05:51, Andrew Ho wrote:
> On Fri, 20 Feb 2004, denny adelman wrote:
> ...
> > I found myself wondering what alchemy would bring the open source
> > community into legitimate competition for such large projects.
> 
> Denny,
>   This is a great question that actually has already been answered. The
> answer is contained in the history of GNU/Linux, Apache, Mozilla, and
> other well-described open-source projects.

I think the most striking characteristic of those projects over the last
5 years or so is their gradual "professionalisation" - that is, more and
more of the core developers working on the projects are paid to do so on
a full-time or substantial part-time basis. That now seems to be a
defining attribute for such projects. The popular conception of open
source development, fostered by Eris S Raymond's essays, of thousands of
people all working a few hours a week in their spare time on a large
software project, no longer matches reality, it would seem. Sure,
thousands of people still contribute to, say, the Linux kernel, but
there is a full-time core of developers (including Torvalds) who are
essential to its continuing success. 

Can anyone come up with some counter-examples of very large, very
complex open source projects which don't have a core team which
comprises developers and/or managers who are paid to do that work.

Note that there are no value judgements here, and the insinuations of
traditional software companies that paid developers somehow produce
better code than volunteer developers are clearly false. It is more a
matter of pure practicality - in order to undertake really complex
projects, you need a team of people who can immerse themselves in the
project, 40, 50, 60 or 70 hours per week (40 hrs being the sustainable
level...).

> 
> > And if such a thing could happen, what reassurances could we give that
> > there would not be announcements like "the apparent failure of a new
> > $450m Open Source application at a XYZ Hospital."?
> 
>   The "reassurances" are the following:
>    1) it will not fail: it will succeed incrementally
>    2) it will not be a "$450m open source application at a XYZ hospital":
>         it will be a free application at increasing number of hospitals

My understanding is that the cost of the software licenses is only a
part of the cost of the implementation of large, complex hospital
information system - usually only a small part. And the implementation
is usually rather specific to each site. So there are still very
substantial costs involved in implementing hospital information systems,
reagrdless of the type of software used. I think we've had this
discussion before.

> > Oracle, Peoplesoft, Fujitsu know how to deal with that. Do we?
> 
> Of course we do. Who do you think works at Oracle, Peoplesoft, and
> Fujutsu?
> 
> What makes you think that "open source" means we should exclude Oracle,
> Peoplesoft, Fujitsu from participating? For example, both IBM and
> Microsoft are already open-source software providers.

Microsoft? Just curious: which of their products are licensed under an
OSI-approved open source license? Or do you mean that they provide
consulting services on open source projects built with their products?
If so, very interesting - can you give examples?

> There is no permanent division between "they" and "us". We are all trying
> to solve real world problems. Alliances and relationships will change over
> time. If Oracle employees know better how to deal with certain tasks, then
> they should do that part of the project.

Quite so, and that's why implementations of an open source hospital
system may still costs a substantial slice of that $450 million, because
those Oracle people generally don't work on a volunteer basis.

-- 

Tim C

PGP/GnuPG Key 1024D/EAF993D0 available from keyservers everywhere
or at http://members.optushome.com.au/tchur/pubkey.asc
Key fingerprint = 8C22 BF76 33BA B3B5 1D5B  EB37 7891 46A9 EAF9 93D0


Attachment: signature.asc
Description: This is a digitally signed message part

Reply via email to