[prev in list] [next in list] [prev in thread] [next in thread] 

List:       gphoto-devel
Subject:    [gphoto-devel] Re: gPhoto2 licensing clarification (was Re: Introduction and Comments)
From:       Richard Stallman <rms () santafe ! edu>
Date:       2000-08-25 21:25:07
[Download RAW message or body]

    > This might perhaps influence their decision, but we cannot rely on it
    > to win them over.  The company may prefer keep the hardware interface
    > secret and distribute only a binary driver.  We have seen this happen
    > over and over with drivers for Linux and with proprietary X servers.

    but they lose the benefit of no-cost development and maintainence. 

This is a benefit, of course, but it is not a very large benefit by
their standards.  Writing the driver is probably a few months of work,
and maintaining it over time afterward for changes in gPhoto would
only require occasional work.  The expense is very small compared with
designing the camera itself, and not significant by the standards of
the money involved in such a project.

The fact that companies which DO cooperate are interested in hiring
people from our community to write drivers shows that getting the work
done gratis is not their major motivation.

    If they release only binary drivers,
    they will need to do the work internally, and release subsequent
    versions of their driver for all the platforms gPhoto2 supports
 
Compared with releasing drivers for IXLA and TWAIN and the other
access programs you mentioned, compiling one gPhoto driver on a few
platforms will be a fairly small job.  And they are not compelled to
support all the platforms where gPhoto works.  They could support the
most common one or two, and say for the rest "that platform is not
supported."  Only a few people would be unhappy, and they would accept
it as normal.

And they don't have to do the work in-house.  I could imagine that
some company might start developing and testing non-free gPhoto
drivers for camera manufacturers, and releasing them compiled for a
whole range of platforms.  Such a company could easily maintain a
server farm for the job.

    > We have one way to put pressure on them: by preventing gPhoto from
    > supporting their cameras.  The strength of this pressure will increase
    > year by year, along with the popularity of GNU/Linux.

    But if the proprietary algorithm is their source of revenue, there is no
    way they will open it up. They would be giving up their funding. 

These companies' source of revenue is sale of cameras.  But some may
think that their secret interface is tremendously important, and those
will be inclined to refuse to release it.

So we, the community, need to apply market pressure, by buying their
competitors' cameras instead.  With 20 million users, our potential
strength is substantial, and our numbers are increasing.  But will we
actually use our strength where it is needed?

Right now we mostly fail to use it.  The problem is that most of the
20 million users agree, more or less, with the Open Source Movement:
non-free software is annoying but acceptable.  And they tend to think
of the drivers as a small matter.  So they choose what hardware to buy
independent of this issue, and then use non-free drivers if necessary.
As a result, the market pressure we actually apply is far less than
what we ought to be capable of.

If gPhoto simply does not allow non-free drivers, then people who want
to use gPhoto will need to choose the cameras that support free
software.  Our strength will be fully mobilized for market pressure.
But if gPhoto does allow non-free drivers, we will lose this
opportunity to mobilize the community's strength.  We will depend on
users to care enoughF individually to take positive action, and only a
fraction of them will.

We could instead try to persuade users to reject the camers that
require non-free drivers even though they do "work".  We would have to
rouse people to feel stronger concern about the issue.  To make this
campaign effective, our actions have to fit our words.

If we say to a user, "We used to prohibit non-free drivers, but then
we decided to accept them instead; now we'd like you to reject them,"
then our actions do not fit our words.  The user is likely to respond
by thinking, "If you decided to accept them, why shouldn't I?"

I'm not making a blanket rejection of the lax policy option on
questions like this one.  I've chosen the lax policy from time to time
myself, one instance being when I decided to release GNU libc under
the LGPL.  (This is what permits non-free applications to be written
on GNU/Linux.)

But the place to yield in this way is where we are weak.  Where we are
strong, that is the place to stand firm, so that companies will
eventually yield instead.

My understanding is that gPhoto 1, employing a firm policy, already
supports a large number of cameras.  It appears we are not in the sort
of weak position where we would have to make concessions.  Our firm
policy seems to be mostly succeeding.

Ole, could you report on the situation for us?  Of the important camera
makers, which have cooperated, and which have refused?

[prev in list] [next in list] [prev in thread] [next in thread]