On Wednesday 2004-04-14 20:00, Julia Thompson wrote:
> Article at
> http://firstmonday.org/issues/issue9_4/levesque/index.html
>
> I'm interested in anything folks here can add about open source
> software, either from a user's point of view or a developer's point of
> view.  Or if someone has a refutation for any of the points raised in
> the article, I'd be interested in reading that.

I think Levesque makes some good points and is basically correct.  (I'm a bit 
of a sociology snob, so I don't think the essay rises to accademic quality, 
but that does not make it any less correct.)

Levesque complains that FOSS, especially that intended for desktop 
consumption, has poor usability (except for the technocratic elite), poor 
documentation, obscure access to technical support (no matter how cheap or 
effective), tends to substitute technical support for documentation, and 
suffers from "not invented here" syndrome (particularly with regard to 
proprietary software).  

I think that she is quite correct.  All of her complaints can be traced back 
to volunteerism.  (And those projects such as Mozilla Firefox or OpenOffice 
that she might excuse have histories rooted in externalization of cost 
centers thus have been less influenced by volunteerism. [See:
http://lintellect.org/index.php?p=2].)

Volunteerism strongly implies that a "typical" FOSS project lacks discipline.  
When working on proprietary or in-house MIS software projects the developers 
are subject to managerial discipline.  Management decides on the project 
goals, decides on features, user interface issues, and devotes resources to 
documentation.  Developers (more or less) work toward the official project 
goal, confine themselves to features specified in the requirements documents, 
negotiate with the usability experts (which they don't like) and take time to 
answer questions from the technical writers (which they also usually don't 
like).  Developers who get too out of line get fired.

In theory it might be possible to get the industrial design and engineering 
folks to contribute to a FOSS project.  You might even get technical writers 
to contribute.  Presumably they would do so for many of the same reasons as 
the hackers contribute code.  (Except that NOBODY thinks technical writing is 
fun.  Also, most marketing and usability research CANNOT be done for free.)

Even if a project COULD get reasonable documentation, marketing, and usability 
contributions, it would make no difference in the results produced in a 
classic bazzar style project developed by cowboy hackers.  The developers 
would take the contributions of the marketeers and usability folks as 
(annoying) guidelines.  That is, they would mostly ignore them.  They would 
spend too little time with the technical writers, and release applications 
without waiting for updated documentation (since documentation necessarilly 
lags development).

Only with managerial discipline could a FOSS project benefit from contributed 
documentation, marketing, and usability resources.  Thus, we only expect to 
see an emphasis on non-programming aspects of software development in loss 
leader and externalized cost center projects.  Academic projects could also 
emphasise non-progamming software development components but one expects that 
departmental structures and grantsmanship would de-emphasize para-programming 
in academic software development projects. 

In this light it is worth noting that Red Hat has basically withdrawn from the 
end-user desktop market.  Perhaps we should simply accept that Linux (and 
most other FOSS) is for servers.  FOSS on the desktop is for techno-geeks and 
those too poor to purchase proprietary software with higher quality 
para-programmatic features.
---------------------------------------------------
PLUG-discuss mailing list - PLUG-discuss@lists.plug.phoenix.az.us
To subscribe, unsubscribe, or to change  you mail settings:
http://lists.PLUG.phoenix.az.us/mailman/listinfo/plug-discuss