Re: [tlug] Bill Gates and the GPL , let the flames begin

[Curt Sampson <cjs@example.com>]

On 2008-04-29 09:11 +0100 (Tue), Godwin Stewart wrote:

> If you've just spent ages writing millions of lines of code, what's
> the big deal with writing another couple of thousand to have a library
> functionally equivalent to readline that doesn't force you to release
> the whole enchilada?

Hooo, boy! I'm having a bad day, but at least now that I've read this I
get a good explanation out of my bad day.

I came in to the office on this beautiful holiday day to spend
some time making improvements to our (Starling's) open source QAM
build/test/deploy framework, in the hope that I could make builds faster
for all of the millions of users out there. (Well, ok, seven, but still,
the point is that I'm spending my time on something that anybody can
use, and will go to directly benefit certain people I know.)

But before I did that, I thought I'd have a quick look at my
long-neglected irb tab-completion. It appears to be using editline,
perhaps in some sort of readline compatability mode, rather than
readline, as would seem reasonable on a BSD system. (It saves installing
the readline package, so that I have some hope of having to deal with
only one configuration file (.editrc) rather than two (.editrc and
.inputrc).)

But it was not to be; after a full hour of furious debugging, and
finding bugs beyond belief in the editline library (it considers any
unparsable line in .editrc--and maybe some parsable ones, too--to
successfully terminate the parsing of that file) I finally gave up in
frustration. This put me in such a bad mood that I spent the rest of the
day starting flamewars on mailing lists instead of coding.

So the open source community just lost several hours of development
(one directly attributable to readline not being usable in BSD, and a
few others that you might count for less because I'm moody), and will
doubtless lose many more for similar reasons, not even yet counting the
time spent on actual editline development.

So here's how you tell if you really believe in the GPL. Ask yourself,
do you honestly thing that this is a good thing, and that not only me,
but my customers and others who were about to get free, open-source
software, should not have it because they agreed to use any non-GPL
software?

On 2008-04-29 09:24 -0500 (Tue), Daniel A. Ramaley wrote:

> I don't really see what the difference is between a 10 line program
> and a multi-million line program.

Well, this one fortunately is relatively easy to find. out. Write a
ten line program, write a ten thousand line program, and look at the
difference. It's something like a thousand times that.

> If you don't like the licensing terms on the readline library, find a
> different library to use. Or just write a library yourself.

Did that. With the beneficial effects on the GPL community as described
above.


On 2008-04-29 11:45 -0400 (Tue), Joe Larabell wrote:

> If I spent three weeks writing a million-line program (I work really
> fast ;-) and I decide to share my work with others, I don't see any
> problem with requiring whomever uses my work to share whatever they
> were able to produce with the three weeks they saved not having to
> re-implement my functions.

I do.

Say I could sign up to some sort of global license to all software that
I could use any software for free if I would only spend as much time
contributing free software as I'd saved by not writing that software
myself. I'd be paying for a lot more software if this were the case,
because in my entire life I could never rewrite even a tenth of all of
the software I use.

Probably I'd abandon programming entirely and find another career, if
the above were my only option.

Which starts to make me wonder, actually, what categories various folks
are in, and how those correspond to the various opinions on this. Here
are some, let me know where you fall, if you feel like it, as you
express your opinions.

A) Does not write software
B) Writes software only when paid by others
C) Writes software when not paid by others, and
    C1) makes primary living from other than writing software
    C2) makes primary living from writing software
D) Writes software when not paid by others and gives it away, and
    D1) makes primary living from other than writing software
    D2) makes primary living from writing software
E) Pays others to write software to give away, and
    E1) funds this from other than writing software
    E2) funds this from writing software

(I happen to be in category E2, for the record.)

On 2008-04-30 04:11 +0900 (Wed), Stephen J. Turnbull wrote:

> What we are arguing about here is the *difference* between the value
> of the program in use to users (whether end users or programmers of
> derivatives) and the value of the time used to produce the original
> program to its developer.  Who gets to keep that surplus?  The goal
> of the free software movement is to ensure that those who *did not
> produce the software* have an overriding right to it.  Hm ....

Fascinating! This is about the most interesting comment on the GPL I've
heard in years.

(Is it to early to jump up and down yelling, "Yes! Yes! They really
*are* communists!"? :-))

>  > Isn't that what the GPL is basically about: enforced sharing?
> 
> "Enforced sharing"?  Another one for etc/OXYMORONS!  (This is
> *precisely* what Curt finds hard to swallow, I suspect.)

Yup.

On 2008-04-29 14:06 +0900 (Tue), Stephen J. Turnbull wrote:

> Curt Sampson writes:
> 
>  > I don't think that just "play nicely" is quite the right way of
>  > putting it.

> Hey, IP is the legal version of the dog in the manger. It *doesn't*
> hurt me to let other people use my software, I lose nothing (except
> revenue). I think that when you've got an infinite amount of sand,
> putting up a fence around the sandbox is *not nice*.

Nah, I'm not going to buy the infinite amount of sand analogy. First of
all, it's not even free to copy free software; someone's got to spend
time to do it. It may be taking less and less time now, but only because
people have spent time writing new software to make the copying take
less work. Perhaps, for just this, an analogy involving an infinite
amount of sand "over there," where you still have to get it to where you
want to use it, would apply.

But more importantly, we are talking about spending time and effort
writing new software (or expanding existing software, or even locating
bugs and filing bug reports, which is a surprisingly time-consuming [and
of course valuable] activity).

So where does, "I used readline in my the giant office productivity app
I was silly enough to write" fit in?

> The world is not a nice place.

I agree with this, in the sense that you're using it. So where's the
analogy that works for this? I might even say (pending further thought)
that I think BSD-style licensing makes the world a nicer place than
GPL-style licensing does. Even some reasonably enthusiastic GPLers
(who use minimal non-GPL software) might actually benefit more from
closed-source programs using more GPL code than they benefit from what's
been GPL'd due to the viral structure of the license.

> Anyway, that's a bad example.  I know at least one lawyer who wanted
> to fight that exact case, and one who has published a book in which he
> says that you'd win if you took it to court.

If I had the money, I'd certainly be tempted. Though I'd imagine that
not may folks want to go down in history, even in the short term, as a
Judas.

cjs
-- 
Curt Sampson       <cjs@example.com>        +81 90 7737 2974   
Mobile sites and software consulting: http://www.starling-software.com