Introducing the Python Business Forum
Andy Robinson
andy at reportlab.com
Fri May 10 21:04:39 EDT 2002
We would like to take this opportunity to introduce you to the Python
Business Forum. Recent discussion among the leaders of a number of
businesses which develop using the Python programming language has found
a business case for closer collaboration. A more formal organisation
was indicated. Thus was born the Python Business Forum: a trade
organisation
set up to coordinate official formal business of the Python Business
Community.
The PBF has a webpage, currently: http://pbf.nuxeo.org/
where you can find:
a) the minutes of the formation of the PBF
b) the board minutes of the PBF
and
c) (the good part) the Bylaws of the PBF.
In addition, a partial list of Projects already planned by the PBF will
be arriving as soon as each Special Interest Group has selected its
own Chairman. Membership pre-requisites, and instructions on how to
join are found there also.
In the hopes that it will encourage membership in the PBF, we hereby
present the 'preemptive FAQ'. This is simply a the list of questions
and answers that we expect people will ask us. The FAQ currently
lives here:
We will, of course, update it with answers as we get real questions,
frequently asked or not.
Laura Creighton
Secretary, PBF
---------------------------------------------------------------------------
Q1: What are some projects currently in the scope of the PBF?
A1.a -- Creating build and test farms and maintaining a stable
business release of Python
A problem that faces many of us is the certification of the Python
distribution on different platforms, and the certification of our products
with the Python distribution of different platforms. Through the PBF we
could share the hardware and personnel needed to maintain a range of
platforms that would be very expensive for each one of us to keep in
operation.
The current 6 month release cycle that Guido van Rossum and the Python
development team prefers is far too quick for most businesses to keep up
with. It puts undue strain on a company to re-certify your products every
6 months and your customers get unhappy if you make them upgrade too
frequently. It also means that there are multiple candidates for
"best version to install" at one single time. Note that some businesses
don't find it a problem, but at least 75% of those we have spoken with do.
The consensus is that an 18 month release cycle would be better, for
business.
But this release cycle is far too sluggish for language developers.
Their needs are for quick release cycles with freqent changes and
constant testing. This attracts the bright, young developers and
avoids the problem of 'force-feeding the Python' in order
to make a release date, then idling for months. But the more
releases one has, the harder it is to make sure that crucial bug
fixes are ported to an increasing number of release candidates.
The current process requires a hero, who needs things so badly
himself that he is willing to do whatever it takes to get a
bug-fix release of whatever version out the door. We can improve
on this process.
But, it falls on us, the Business Community, to make sure that
we have something that is stable and which sends a single, uniform
signal to the market. This means that we need to backpatch bugfixes
and significant performance improvements from the development release
of Python to the business release and certify it on a wide range of
platforms. We may also want to backport new features that do not break
any code. For the core language, this is a sensitive issue, which should
require a strong consensus before anything is done, while for libraries
it is much less sensitive though of greater importance to some of us -
'service packs' are quite possible.
GvR has agreed to flagging such a release and sending such market signals.
He would designate 2.x.y as "the release that wears-a-Tie, as endorsed
by the Python Business Forum members....". In return, he wants
us to take on the burden of backporting any critical fixes; and is
keen on the idea of a build farm where pythonlabs' code can be tested
against a wide variety of real world apps and extensions, as well
as the Python test suite.
We have already located a superb resource with over 30 Unix machines to
use as the build-and-test farm for Python, which we dub the "Snake Farm".
A Separate Snake Farm document will be released early next week (week 20
of 2002).
-------------
Status: SiG established. Chairman Laura Creighton
A1.b -- Creating a network of agents
There is a business case for using other companies that rely on Python
as agents for your products and services. Most of our companies are small,
with few resources to reach an international market. By co-operating we
can all extend our markets without heavy investment. The fact that we
are all using Python should make us all able to explain why the underlying
technology makes the products we are selling better. We also have the
base knowledge to provide additional services and consultancy to our
local customers. This limits the need for transfer of know-how to the
product itself.
-------------
Status:
A1.c -- Creating a Consulting and Referral Network of Resources
Some of us have the problem of having more business than we can handle,
while others have idle resources, or would like to use some of our
resources for generating short term revenue in conjunction with working
on development that produces revenue in a longer perspective. The PBF
would provide a forum where requests and availability could be announced,
with reasonable confidentiality.
-------------
Status:
A1.d -- Escrow and Contract Services
The PBF has a bank account and a professional accountant.
When dealing with small businesses or with governments there is a
significant risk of non- or late payment; likewise small consultants
find difficulty in receiving work because their potential clients
worry that the work may be unsatisfactory. The PBF will provide an
escrow service so that vendors can insist a deposit is paid before
starting work, to be paid out under agreed conditions.
-------------
Status:
A1.e -- Providing the group of companies a better media profile
Most of us are small companies with little attractiveness to journalists.
Python and the Python community also lack the factors that make media
want to talk about us. A commercial organisation that stands behind an
OpenSource language would be something new and interesting to report on.
By managing this initial news value well, we can then build a lasting
media interest for Python as well as our companies and products. Once the
productivity advantages of Python are made clear to the general press
(as opposed to the computer press), they will need success stories to
exemplify why Python works so well.
-------------
Status:
A1.f -- Organising business-oriented conferences
There is considerable interest from various government bodies as well as
from large corporations in leveraging the advantages that using OpenSource
bring. However, we are usually too small as individual companies to get
the big organisations to listen to us. By arranging conferences where
we mix the presentation of our individual products with presentations on
the principles of how OpenSource works and why Python is a great
development language, we can attract decision-makers from these
organisations, thereby giving us the chance of conducting business with
them directly, as well as creating the base-knowledge of our existance,
a prerequisite for getting any big contracts at all.
Examples of groups that we would like to target with such conferences are
various offices of the EU Commission, Central Government Agencies across
Europe and the largest corporations in industry, finance and transport.
-------------
Status:
A1.g -- Training
We hope to kick off a series of reputable Python training courses
around Europe.
-------------
Status:
Want more? Join the PBF and form your own SIG.
-------------
Q2. -- Why Now?
A2. Because the time is right. The Forum's founder members have long
experience in the software industry and are determined to create a
professional organisation which grows the commercial market for
Python software.
Q3. -- Come Again? I didn't understand the last part.
A3. Ok, we will now attempt to answer both in the dialect of English
used by business, and annotated in the dialect used by hackers.
Here's the hacker for Q2.
< The Timbot suggested that if the business community was unhappy
it could jolly well do something other than complain. The
first thing we did was organise. >
Q4. -- This looks like a European Organisation to me. Correct?
The initial focus is European, for efficiency because the founding
companies are located there, but this is a full-fledged international
organization. Local regional sub-networks of the PBF are encouraged.
< Trying to get tons of work done when your members are spread
over the globe is a real pain. We've done it. Every time
you need consensus, somebody is bound to be asleep. Unless
you have tremendous discipline, your schedules slip. We
know how to run international projects with members in various
time zones, but the core team does not want to commit itself
this way right now. >
Q5. -- Why isn't the PSF doing this?
The PSF is a developers group, which is concerned about the integrity
of the Python Language, which is an entirely different function. It is
likely that many companies will be members of both organizations.
< We all want better Pythons, but we are bringing a completely
different skill-set to the mix. We want different things, and
know how to get them. We don't want to discuss nerdy things
here, but rather things that bore nerds to tears. Trust me
on this one -- if the phrase 'marginal product' causes you
to think 'badly made junk' and not 'the rate at which output
increases as the quantity of that input increases, all other
inputs held constant', then you probably do not wish to be
the representative that your company selects for the PBF.
Conversely, we do not want to be involved with the technical
decisions about the design of Python. We don't care. You give
us a Python, and we will make a Python-in-a-Tie out of it, and
that's all. Hacker-PBF members will have to leave their
propeller-hats at the door before arriving for our meeting,
because if the talk gets technical, we will declare such
chatter out-of-order and resume with the issue at hand. >
Q6. -- Isn't this a bit formal?
No. It's _extremely_ formal.
Q7. -- Why did you form the PBF before announcing it?
We are results-driven, commercial enterprises. We consider the forming of
the society as a formality. The things of real interest will be happening
in the SIGs.
< If we had done that, the debate about whether to do this
would _still_ be going on!>
Q8. Won't this hurt ActiveState and other companies reselling Python?
No. The only thing that will change is that they will have a Python which
has been tested on more platforms to start out with. Indeed, they will be
free to spend more time in the creative efforts which results in product
differentiation, and less on that necessary but tedious work of testing
and bug-fixing.
Q9. Won't this make more work for Guido and the PythonLabs team?
No.
< We only got the aggrement with PythonLabs to go ahead with this because
we promised them _less_ work, not _more_. >
More information about the Python-list
mailing list