Re: Debian goes big business?
Paul your right in the Non-traditional part:
Read below the commented text for my remarks:
> > > Please feel free to start up a company to sell Debian and generate
> > > some income from this. Any surplus can easily be reinvested in making
> > > Debian and it's cause even more a reality. Morals only can't buy nor
> > > sell anything. I *really* hope someone would stand up and found a
> > > Debian company. So if you think you can...
> >
> > I don't think Debian is usable to found a company on that. No company
> > can actually control Debian, impose release dates and such needed
> > things (for a company). Even if it's feasible, no company ever SHOULD
> > have such rights, for Debian to keep it's spirit.
> >
> You are thinking in the wrong traditional terms. It's not about
> controlling Debian or imposing anything upon anyone. A company based
> on Debian would need a different business strategy. Just take into
> account what a company like Cygnus is doing for free software!
> A company basing it's business on Debian Linux should ideally be
> composed of Debian people and would mainly care in making Debian known
> as a viable product on the wider market. Generated income could be
> used to give full time jobs to Debian developers who could then fully
> concentrate on Debian for a living. This could probably help increase
> the release frequency and would provide a financial framework for
> Debian. At the moment it's really a pity that mainly third parties
> are generating income mainly for themselves and i believe we would
> considerably benefit if a company would do the same specifically for
> Debian. Wouldn't you like to be paid working for Debian?
> I'm no business person and don't have any clue about this sort of
> thing but i sure would like to see somebody with the appropriate
> knowledge found a company benefitting Debian like described above.
> And i surely would like to work for such a company getting paid for
> what i enjoy to do.
> Paul Seelig <pseelig@goofy.zdv.uni-mainz.de>
I just had to weigh in on the discussion thread. I'll give a little
background so as to attempt to add a little credibility to me comments
;-) I have been a DOS & Windows Developer for quite a long time. I
have been an employee and a contractor. I, like many of you, have done
development for non-profit organizations and for fun (hoping that
someone might like my hack). I have taught software classes for end
users and advanced users. I have also run my own small business
(independent software development). Recently, I spent two years with a
startup company that brought a Clinical Patient Record software system
to market. During this stint I worked in Internal Business Systems
supporting and integrating the Finance, Marketing, Sales, Support
Quality Assurance and Training Departments. Enough already, you know my
type, here come the comments: (I do not know the current orginization of
the Gnu/Debian Project.
Business Entity:
This must be a service based non-profit corporation with no intent to
sell a product. Marketing, Distribution, Support and Development must
be the core focus in the order listed. I would compare it to a
volunteer fire department. Get the word out, professional from the
observers point of view and rapid with response for support.
Marketing:
The major question here is target market. Is this an OS for
network/system admin types
or for general consumption (including "push button windows types").
Could their be a
chance for both? An alliance with someone to put a noce wrapper on
Debian. (I noticed
the intent to work with SEUL, good move.) This would have to be
resolved first and
avoid the too many directions therefore now progress problem.
Next this area has 3 sub areas; 1) Champion the product and get it
visible - send it to
journals and magazines both inside and OUTSIDE the Linux community.
This serves to
generate users and credibility in the market. RedHat and Caldera
don't just wait for
PCWeek and LAN Times to call them. 2) Strategic Alliances -
although commercial
packages are against many peoples altruistic intentions, in reality
business and private
users still want a manual and "real" CDs. One does not have to
become commercial to get
companies to QA on their distribution and mention them in their
literature. Sub Area 1
gives weight to this effort by creating a larger user base with
which to sweeten the
alliance. An additional carrot is to provide the company with
technical assistance in
making their product stable on Debian. 3) Fundraising - this is
different than
strategic alliences since it must be done by looking for a variety
of sources. I would
suggest a large amount of money might be available through the
government. What are\
their computer security people using for opperating systems and the
like. They
have a large amount of money for grants and would probably support
the effort since
their unclassified code is publicly owned already. They could also
hire Linux folks
for civilian positions in their computer labs and allow them to work
on Debian on the
clock. Again, software company alliances could provide some cash,
start with the folks
who are already distributing Linux Versions (Adobe, WordPerfect,
Netscape/AOL). Private
donors and computing societies would be a next. Notice no mention
of revenue from CD
distribution. As you may have guessed this would be the first full
time position
funded.
Distribution:
I am not suggesting the CD in the mail barrage like AOL. ;-) I am
talking about a
central point of contact for an interested business entity to get
the most current
stable release with software from all alliance partners and
paperback documentation
(this could be a pretty version of the debian HOWTOs and tutorial
with a nice cover -
see RedHat and SuSE User's Manuals). The book currently on the
market could be
included It is fair for a non-profit to require a donation from the
requestor that
will cover printing/book purchase and CD cost. This must be run
with a very lean
inventory and a quick turn around from a publisher for new CDs and
books. A
professional distribution package for professional inquiries will
pay off by reducing
the "loose network of Linux hackers" image that debian has. Even if
that defines the
development effort it does not have to look like it. BTW as you all
know, corporate
development is often less focused and effective than the Debian
project.
Logo and catch phrase development, Debian needs a definite image to
set it appart.
There needs to be additions to the distribution agreement that if a
Logo other than the
company/publisher that is providing the CDs is displayed (e.g. SuSEs
logo) then the
Debian logo is displayed. This must be consistent with the Debian
distributed package.
This will again increase visibility and reputation. Initially, this
could be handled by
the marketing rep(s).
Support:
I will have to say that if you like computers and like searching,
linux has great
support (especially Debian). However, real end users and non-hacker
corporate users
want a central location just like the commercial distributions.
Mark Spencer and
Charles P. Wright at Linux Support Services project have the right
idea. If you haven't
done so visit them at http://support.marko.net/. This is a
professional approach with
volunteer technicians. Users want confirmation of a request,
trouble tickets assigned to
technicians and results. The list servers/newsgroups are great but
lack the polish that
will be competative. I just hope RedHat and Caldera don't figure
out that they could get
setup a similar system and get users from their newsgroups to answer
for free. ;-)
However, the Debian project may want to partner with them and
migrate existing functions
to their format. I do not believe telephone support is necesary
based on the excellent
existing support. One arguemnt might be that a single operator be
available for folks
who have lost their ability to boot or get to their ISP/LAN.
Development:
There are some issues here that I will address bluntly without
getting wanting into
arguements about how difficult the implementation would be. I will
also add the
disclaimer that I am not currently in a posiotion to begin to learn
the enviroinment
to write code for the project. However, as this message implies, I
am very interested
in contributing in ways that do not include development. The major
issues that I see
that will continue to make Linux and Debian a tough sell to
non-technical users:
Installation:
SuSE has Yast and RedHat has their cool stuff. I personally
prefer Debian even with
dselect due to its flexibility. Debian would do well to get to
Yast ease of use while
keeping the flexibility of dselect. I know APT is comming, I am
anxiosly awaiting the
front end.
Software Packaging:
This area reminds me of DOS & Windows in 1993. Developers made
zip files
available with brief instructions to unzip, make a directory, copy
files to it and
copy the libraries to C:\Windows\System. Forget library version
control! Then came
the self-extracting zip file. Then installation scripts, still no
attempt at version
control. Software packaging software like the Wise Installation
System became
available for the developer. Finally, someone litened to users
and added library
version checking. Microsoft even helped some until the first COM
objects release.
You VB developers know what I mean. ;-) Version control in
Windows X is still not
solved, but it is much better than 1993. Software distrubuted
with instructions on
how to untar and then compile is not going to help compete against
windows.
So far my Debian boxes have been very stable, however, software
installation has
caused system lockups and complete re-installs. (In some areas I
refuse to be much
more that and end user due to time constraints and knowledge of
the internals)
Software companies looking to Get into Linux need to hire people
to package for the
major distributions, hence, if Debian were one then someone would
have a paying job
and Debian development time and the community would benefit.
Someone needs to go
fishing for a new job. OR The Debian project could charge the
companies for packaging
and QAing the package install on Debian and provide the Debian
Logo as certification
for the companies. This is very common, check the Windows Logo
program. Initially
volunteer developers could be paid as contractors, then hire a
couple to ensure quick
turn around.
Migration Services:
Again conulting work. Contact companies with great Windows
products and charge to
help them port to Linux and Certify them on Debian. You don't
think RedHat is doing
this do you. :-O
Continue Giving:
Although I have spent a lot of time talking about going to where
the money is. We all
know that true innovation often comes from someone getting an Idea
and sharing it.
This aspect of Linux development and Debian development MUST NOT
CHANGE. The OS must
be maintained and goals set and reached while all of this other
stuff is going on.
Innovative tools and applications must continue to be developed.
In fact, if someone
working on an Xeditor/wordprocessor wants an idea for a new
feature, drop me a line.
I have identified a need that lacks a developer.
Sorry about the length of this note. Many people in corporate america
are saying Linux still
lacks the organization to make a big push. Why must commercial
distributions be the only
game in town with organization? This note could make a good start to a
business plan for a
non-profit organization. I am however, busy tring to get into medical
school. I hope I have not stepped on any toes of people who are on the
project. I am just commenting on my market research over the last 4
months and my use of several products in the Linux community.
I would also like to thank all of you for the support that you give
newbies. I hope to rise to a level that I too can begin to answer a few
and give back.
I have stepped down from the soap box. Have a nice day.
--
Danny R. Gray
Research Technician
Department of Pathology
UNC-CH School of Medicine
Reply to: