469,266 Members | 1,803 Online
Bytes | Developer Community
New Post

Home Posts Topics Members FAQ

Post your question to a community of 469,266 developers. It's quick & easy.

The Industry choice

>From technical point of view, I could not understand the the reasoning
behind using Java in major companies. Sure that Python, is used in
some, but still Java is considered as a sure-job language.

After being a python programmer for long time, I consider it painful to
learn/use Java now (well, like many I will be forced to do that in my
job).

What makes such companies to choose Java over dynamic, productive
languages like Python? Are there any viable, technical reasons for
that?

Jul 18 '05
198 6726
Steve Holden <st***@holdenweb.com> wrote:
...
1: 25 u'Luther Blissett'
2: 21 u'Alex Martelli'
... And I *still* think we deserve to be told the Luther Blissett story ...

conspiratorial-ly y'rs - steve


Martin Elster's post back in Aug '02 already had just about all of the
necessary info...:

'''
I had just been reading about Luther Blissett, and this name is a sort
of umbrella name that is available for anyone that is interested, and
that has been used in different "underground" projects and pranks,
especially in Italy.
'''

Add other well-known facts, such as the fact that Bologna is or was a
hotbed of situationist pranks, Umberto Eco is a professor at Bologna
University, Umberto Eco's best-known historical novels from "The Name of
the Rose" onwards may be interestingly compared with Luther Blissett's
"Q", and so forth, and I'm sure you can design a perfectly adequate
conspiracy theory.
Alex
Jul 18 '05 #151
Bulba! wrote:
On 04 Jan 2005 19:25:12 -0800, Paul Rubin
<http://ph****@NOSPAM.invalid> wrote:

"Rob Emmons" <rm******@member.fsf.org> writes:
Me personally, I believe in free software, but always talk about open
source. My answer regarding forcing people to share -- I like the GPL
-- and I am perfectly happy to have anyone who does not like the GPL
not to use any GPLed software. I don't feel compelled to share.

I'd go further. It's not possible to force anyone to share, but the
GPL aims to remove software from a system that instead aims to force
people NOT to share.

Nope. IMHO, GPL attempts to achieve the vendor lock-in. For different
purposes than another well-known vendor, but it still does.

Well you are entitled to your opinion. But *my* opinion is that the GPL
attempts to ensure that if you re-use code by an author who so desires,
then redistribution of your code is only possible by making your own
extensions to it available on the same terms. This gives you a clear choice.

To put it another way, it allows an author to specify that their code
can't be hijacked for proprietary purposes *in distributed programs*. I
will specifically point out that there is *nothing* in the GPL that
requires you to reveal the source of program you write but do not
distribute, even when such programs incorporate tons of GPL'd code.
It's actually even worse: the only thing you can't share on a
well-known vendor's platform is the software written by that
well-known vendor -- you can choose to share or choose not to
share whatever you or other people write on this platform.
Well that's way over-simplified. And if you mean Microsoft, *say*(
Microsoft. And you certainly can't share GPL'd code on Windows without
doing so under the terms required by the GPL.
If GPL folks had their way, it would not be possible not to "share"
_anything_ you create. It is widely acknowledged that GPL
license has the "viral" aspect of extending itself on your
software - can you point to closed-source licenses that would
have this aspect? None of the licenses I've read except GPL has
this aspect. LGPL is still a different story, though.
The GPL folks are quite happy to have you "share" anything that *you*
create. Their simply-stated and elegantly-achieved intent is that you
don't "share" anything that *they* create except on the terms they have
required for their creations.

So, it seems to me, you are whining because the authors of GPL'd code
don't want you to release *their* code except under the GPL. What gives
*you* the right to dictate to them? How would you like it if Richard
Stallman insisted that you release your code under the GPL? Which, of
course, he doesn't.
As the MPAA knows, people do want to share, and
forcing them not to do so is impossible without turning the world into
a police state.

Socialism is unpopular for many reasons, and many of them are indeed to
do with maintaining the separation between individuals and thereby
retaining the ability to treat them as separate economic units. But we
aren't going to change that by insisting on particular software
licenses. Realize this is a very small part of a very large debate.
What's the cost of copying music files vs cost of combining
some programs together, even in the form of e.g. using an
external library?

Maybe if Python were GPL, then Bulba wouldn't use it,
but since it's not GPL, some people find themselves much less willing
to contribute to it than if it were GPL.

And that is their choice. They should realize, however, that some
licenses (including the more recent Python licenses) are cleared as
"GPL-compatible". I believe this means that if I receive software
licensed under a GPL-compatible license, I am at liberty to distribute
it under the GPL.

I suspect that this point is far too infrequently stressed.
Personally, I have precisely opposite impression: the OSS licensed
with BSD/MIT/Artistic/Python-like license gets contributed to a lot
simply because people like to use it and they are not afraid of
licensing issues.
This merely goes to show that different people can form different
impressions when discussing the same sets of facts, and therefore how
useless impressions are as the basis for rational discussion.
When people share:

_it is not because this or that license of software used by them says
so, but because they want to for reasons orthogonal to licensing
issues_.
Absolutely not. Some people want to share under very specific
conditions, hence the proliferation of licenses in the open source world.
(I myself contribute bug
reports and maybe small patches, but resist larger projects since
there are GPL'd things that I can do instead). So catering to the
wishes of Bulba and Microsoft may actually be impeding Python
development. Yes, there are some people selfless enough to do long
and difficult unpaid software tasks so that Bulba and Bill G can get
richer by stopping people from sharing it, but others of us only want
to do unpaid programming if we can make sure that the results stay
available for sharing.

Actually, I get the impression that GPL-ed software is written by
programmers for programmers, not really for end users.

Not at all. It's written to be redistributed under specific terms, and
anyone who doesn't like those terms has the option of redeveloping the
functionality for themselves.

You can't insist that people give you their intellectual property on
*your* terms. That would be like insisting that the music industry bring
down the price of their clearly-overpriced products, or that the
Baltimore Orioles stop the concession stands from charging $4.50 for a
one-dollar beer. If you want a vote in such situations then your feet
are the appropriate instrument. Walk away, and stop whining :-).
Insisting will do you no good.
GPL folks just insulate themselves in their ghetto from the rest
of the world. More and more of the successful OSS projects have
non-GPLed licenses: Apache, Postgres, Perl, Mozilla, Python. Do you
_really_ see few contributions made to those?

More and more? Can we see some numbers to support this "impression"?
--
It's a man's life in a Python Programming Association.


Since I'm taking issue with you, I will end by gently pointing out that
there's a substantial minority (? - my impression) of people who might
find your tag line (which I am sure is intended to be supportive of
Python and the c.l.py ethic, such as we might agree exists),
gender-biased and therefore just as unacceptable to them as the GPL
appears to be to you.

regards
Steve
--
Steve Holden http://www.holdenweb.com/
Python Web Programming http://pydish.holdenweb.com/
Holden Web LLC +1 703 861 4237 +1 800 494 3119
Jul 18 '05 #152
Roel Schroeven <rs****************@fastmail.fm> wrote:
Can you point to closed-source licenses that allow using the code *at
all*?


As I recall, for example, Microsoft Visual C++ came with sources for
various libraries; all that the (closed-source) license for those
libraries forbade you from doing was to further distribute the _sources_
themselves. You could do modifications big or small to those libraries
for whatever purposes, and redistribute the _compiled_ form of the code
as a DLL, or statically linked into your own programs, etc, etc.

Is this what you mean by "allow using the code *at all*"? I think it's
a pretty common arrangement when the code being sold under closed-source
terms is a set of libraries, or a development system part of whose value
is a set of accompanying libraries.
Alex
Jul 18 '05 #153
Bulba! wrote:
On Wed, 5 Jan 2005 11:19:56 +0100, al*****@yahoo.com (Alex Martelli)
wrote: [...] You see, I'm not disagreeing with you that your model applies
_where it applies_. I only disagree that it applies in face of
stronger forces. Now what kind of forces is dominant in
most frequent scenarios would have to be worked out in tedious
empirical research I think. Which I haven't done, because
learning some economics is just a hobby to me.
Yes, by all means let's just spout our opinions without any of that
inconvenient tedious empirical research which might invalidate them. [...]
Italian, at most German and French, and not Polish or Russian, closeness
to good international airports and other good transportation, closeness
to partner firms and potential customers' decision-makers, all appeared
to point to Warsaw, if I recall correctly. Mechanical engineers with
some programming experience or viceversa, good translators, and good
salespeople with connections in the mechanical industry, are not as
ultra-specialized as all that, after all.

Most sales offices in Warsaw do not employ esp. educated people in
my impression. OTOH, the carmaking facilities nowadays require
more much more know-how and specialized workforce than a sales
office does. Or at least that was my impression when I worked at
the construction machine manufacturer in Berlin.

Again you are forming impressions form rather limited evidence: I might
agree with you about the relative intelligence and education of
engineers over sales people, but that might be *my* bias showing.
Capital investments per worker in auto industries are reportedly
very high. Simple physical tasks are done largely by machines,
like this 100 million Deutschmark costing laser-cutting installation
that I've seen there, where a pile of iron bars is pulled in at one
end and the pile of ready components is spitted out of the other end
(unlike typical thermal cutting, laser has the advantage of not
destroying the metal structure adjacent to the cut, so the parts
of the machines subject to high-stress are oft produced this way).
The same is true of plasma-arc cutting for thicker steels, and I believe
it's still not possible to cut 3-inch stainless with a laser. But what's
your point?
Oh, and by the way that installation doesn't get used much.
Somebody at the office didn't check carefully enough the
energy prices before ordering it and later someone discovered
that off-site specialized cutting firms that take advantage of
energy available at low prices at special times in other countries
can get it produced cheaper. Moving it elsewhere or selling
is not an option, since it is a specially constructed, low, 50-meters
long hall that stands inside the huge manufacturing hall of the
company.
And you are using this example to try and argue that engineers are
better-educated than sales people? Who sold this installation? Who
bought it? How much, therefore, is education worth?
100 million DM (when 1 DM was worth some half of Euro
back then) down the drain. When the company was in rather
bad financial situation (later I've learned it was finally bought
out by Americans). Oh well. No big deal.

I was utterly shocked. Having grown up in Soviet times I have
been used to seeing precious resources wasted by organizations
as if resources were growing on trees, but smth like this?! In a
shining ideal country of Germany?! Unthinkable.
Indeed not. Quite often the brown paper bag is a factor in purchases
like this. I wouldn't be at all surprised if somebody with a major input
to the decision-making process retired to a nice place in the country
shortly afterwards. You appear to be making the mistake of believing
that people will act in the larger interest, when sadly most individuals
tend to put their own interests first (some would go as far as to define
self-interest as the determinant of behavior).
The firm I was working for had a consensus decision-making process (even
I was involved) and managers (and other employees) and stockholders were
mostly the same people -- it wasn't all that large a firm at the time.
Nobody needed to practice risk avoidance.

Again, you may have had good luck. Where I worked (including
some places in Germany and UK) it was almost the only factor
that seemed to matter to people - they'd do ANYTHING not to
take a risky decision, to "pass the buck", not to stick their necks
out, not to declare doing some work that involved challenges.

Some people are like that. I chose a long time ago to try not to work
with them whenever I could avoid it and, while that may have had
negative economic consequences I an convinced it has improved my quality
of life immensely. Of course, I have no proof for such an assertion.
[and on, and on, and on ...]


regards
Steve
--
Steve Holden http://www.holdenweb.com/
Python Web Programming http://pydish.holdenweb.com/
Holden Web LLC +1 703 861 4237 +1 800 494 3119
Jul 18 '05 #154
Alex Martelli wrote:
Roel Schroeven <rs****************@fastmail.fm> wrote:

Can you point to closed-source licenses that allow using the code *at
all*?

As I recall, for example, Microsoft Visual C++ came with sources for
various libraries; all that the (closed-source) license for those
libraries forbade you from doing was to further distribute the _sources_
themselves. You could do modifications big or small to those libraries
for whatever purposes, and redistribute the _compiled_ form of the code
as a DLL, or statically linked into your own programs, etc, etc.

Is this what you mean by "allow using the code *at all*"? I think it's
a pretty common arrangement when the code being sold under closed-source
terms is a set of libraries, or a development system part of whose value
is a set of accompanying libraries.


OK, I've been bitten by my exageration. There are indeed special cases
such as some libraries.

I was thinking more of end-user packages: if you somehow could lay your
hands on the source code of Visual Studio itself, you're still not
allowed to do anything with it.

--
"Codito ergo sum"
Roel Schroeven
Jul 18 '05 #155
Roel Schroeven <rs****************@fastmail.fm> wrote:
...
Can you point to closed-source licenses that allow using the code *at
all*?
... Is this what you mean by "allow using the code *at all*"? I think it's
a pretty common arrangement when the code being sold under closed-source
terms is a set of libraries, or a development system part of whose value
is a set of accompanying libraries.


OK, I've been bitten by my exageration. There are indeed special cases
such as some libraries.

I was thinking more of end-user packages: if you somehow could lay your
hands on the source code of Visual Studio itself, you're still not
allowed to do anything with it.


Yes, apart from libraries and similar cases (frameworks etc), it's no
doubt rare for closed-source "end-user packages" to be sold with
licenses that include source and allow you to "do anything with it".

However, allowing customization (at least for internal use within the
customer organization), while rare, is far from unheard of. I used to
work for a software house which sold rich and complex packages of
software meant for 3D mechanical design. The packages came with tens of
thousands of lines of (closed-source) code, in a proprietary scripting
language implemented by the package itself, which users (typically
mechanical engineers) were _expected_ to tweak to customize the overall
product for their specific purposes -- such modified parts of the
"scripting source" of the overall product were routinely shared among
different customers, and occasionally sold from one to another.

The choice of which parts of code in the scripting language were made
thus customizable and sharable was quite deliberate: the application
contained much more code in that scripting language, but most of it was
only distributed in compiled form (or even with the compiled form
already turned into data in some library or executable) -- the parts
that were sold as sources were picked to be those which would be most
useful for customers to customize and share, yet not damage the business
model of the software house. (and yes, it WAS closed source software,
anyway -- customers were theoretically not permitted to give our source
or derived works thereof to others who _weren't_ customers, I think;
anyway the engine needed to run the scripts was not redistributable, so
that provisions, if there, was of modest value).

I wouldn't be surprised if such a mixed model was reasonably common
among "end-user packages" which include a substantial proprietary
scripting component, particularly if the end-users are expected to be
technically skilled (mechanical engineers aren't programmers, but they
will probably have some vague clue about it; a consumer product might be
different). Just a guess, but, wouldn't, say, Mathematica or some
similar closed-source product benefit from this kind of arrangement --
including, as part of the product being sold, some rich amount of
scripting code, freely customizable and sharable among customers
(perhaps with the license prohibiting giving it away to non-customers)?

I believe some (closed-source) games, including ones which use Python as
their scripting language, may also do something of the kind -- include
Python sources for "scenarios" with full license to tweak and
redistribute (making new scenarios which are derived works of ones sold
by the games' authors). Here the language is not proprietary but no
doubt the scripts use large amounts of calls to proprietary modules, so
again there is no damage to the game author's business model anyway.

Hmmm, come to think of it, doesn't Apple include very large amount of
working Applescript code with many of its closed-source applications?
So, although your general point is no doubt sound, almost by definition
of closed source, you may perhaps still need to tweak it further, beyond
libraries and frameworks, to also exclude that closed source with is a
"scripting", "configuration", or otherwise "ancillary" part of the
application.
One last reflection -- I believe there are or used to be some programs
written by people no doubt of very good will, distributed with all
sources and often with no profit motive at all, which are NOT open
source because they include in the license some restrictive clause, such
as "no military use", "no use by organizations which perform testing of
cosmetics on animals", or something of that kind. These would be
examples of closed-source software which DO allow ALMOST any kind of use
-- any EXCEPT the specific one the authors dislike so intensely.

While most people may not think of such programs as "closed source",
they most definitely ARE: the definition of open source is very strict
about this aspect.
Alex
Jul 18 '05 #156
Steve Holden wrote:
Bulba! wrote:

I was utterly shocked. Having grown up in Soviet times I have
been used to seeing precious resources wasted by organizations
as if resources were growing on trees, but smth like this?! In a
shining ideal country of Germany?! Unthinkable.

Indeed not. Quite often the brown paper bag is a factor in purchases
like this. I wouldn't be at all surprised if somebody with a major input
to the decision-making process retired to a nice place in the country
shortly afterwards. You appear to be making the mistake of believing
that people will act in the larger interest, when sadly most individuals
tend to put their own interests first (some would go as far as to define
self-interest as the determinant of behavior).


Indeed, it is almost expected that those in charge of any large
organization (whether government, corporation, trade union, industry
association, fan club, or whatever else) are likely to act in their
personal interests at the expense of the organization's interests.
This is why things like public-disclosure laws and oversight
committees exist. As they say, power corrupts. (Of course, this is
not at all limited to people in charge; it's just most notable there,
since those people can direct the efforts of the rest of the
organization for their personal gain, whereas a rank-and-file member
can typically only direct their own efforts.)

It's also noteworthy to consider that many times, waste happens not
because of corruption or self-interest, but simply because of errors
of judgement. Humans being as we are, it's inevitable that over time,
some "obvious" important details will escape our attention, and the
resulting imperfect information will result in poor decisions. This
is a simple fact of human nature, and (ob-Python ;) ) it's one of the
reasons that Python is designed as it is -- it makes a serious effort
to reduce the number of details that might escape detection.

(One should also consider that many business failures are a case of
simply having played the odds and lost. Many ventures depend on
outside events playing in a certain way; when by chance those events
happen, the decision-makers are called "bold and insightful", but if
things don't work out, they're called foolish or misguided. Often,
though, it was not foolishness but shrewd risk-taking -- if you take a
one-in-three chance of making a tenfold return on investment, then 66%
of the time you'll lose.... but if you hit those odds just once,
you'll come out way ahead.)

Jeff Shannon
Technician/Programmer
Credit International

Jul 18 '05 #157
On Thu, 06 Jan 2005 08:39:11 GMT, Roel Schroeven
<rs****************@fastmail.fm> wrote:
If GPL folks had their way, it would not be possible not to "share"
_anything_ you create.
That's generally the goal of the Free Software Foundation: they think
all users should have the freedom to modify and/or distribute your code.
You have the freedom of having to wash my car then. ;-)
Most people who use the GPL don't feel that way; they think that each
author should have the freedom to choice if and how he chooses to share
code.
Whatever they feel, GPL is designed in such a way that essentially
is an attempt to extend itself onto all the software in the world.
They just see the GPL as an appropriate way to share their code.
And why is that?

Suppose they want to ensure "no forking" or that "bugfixes
and enhancements of original software are given back".

OK, LGPL is fine for this goal. When you say "they see it
as appropriate way to share their code", there's nothing
in this statement that openly and honestly indicates what
is the goal of this method.
It is widely acknowledged that GPL license has the "viral" aspect of
extending itself on your software - can you point to closed-source
licenses that would have this aspect?

Can you point to closed-source licenses that allow using the code *at
all*?
Which code?

_Their_ code? I.e. written by them?

Back then in days of yore I was using C-tree DB library for some
time, that came with source - or actually in practical terms "as
source". My boss wrote quite a lot of nice C++ wrappings thanks
to that (that I didn't bother to use, because I could not be
bothered to learn C++ :-). They still sell it that way:

http://www.faircom.com/products/ctree/

I don't remember being forbidden to redistribute source
code of what _we_ wrote at the company. Just the redistribution
of C-tree source code itself was not allowed. Redistribution of
the compiled binary with the product was OK (and no royalties
were required).
With GPL, you have the choice: either you agree with its terms on
distributing the code and then you can use the code, or you don't agree
with it and you don't use (which is still no worse than closed source).
With all due respect, this is a universal statement that applies
to all the licenses - obviously, if you don't agree to the terms,
you can't use it.

However, it's not aboot agreement to conditions per se, this is
aboot the freedom of speech, err, this is aboot what those
conditions actually are.
Some people call that viral, but I think it's a distortion of the truth.


How so? If you combine source of your program with GPLed source,
your source no longer can have the license of your choosing - it
has to have GPL license, or at least contain the key GPL conditions.
Other licenses do not require you to disclose _your_ source. It's
a really, really weird world in which having to put "obnoxious
advertising clause" is considered as more of imposing yourself on
other people than requiring them to disclose your source.

This is "money with strings attached" approach. It's just ignored
for sake of Grand Goal that few people actually believe into, along
the lines "end justifies means".

OK, I rambled enough, this group should not degenerate into *.advocacy
trash dump. :-)

--
It's a man's life in a Python Programming Association.
Jul 18 '05 #158
On Thu, 06 Jan 2005 15:44:03 GMT, Roel Schroeven
<rs****************@fastmail.fm> wrote:
I was thinking more of end-user packages: if you somehow could lay your
hands on the source code of Visual Studio itself, you're still not
allowed to do anything with it.


And why would anybody want to waste their time reading the
source code of Visual Studio? ;-)

<duck>

No, honestly, after all most of the time what programmers learn
is just API. The very point of having libraries after all is not
having to learn the low-level mechs of this thing, but just
using them in a push-button manner!

My boss read the C-tree code. I was programming reports
and other "peripheral" stuff, so I never had to do it. I was
just using a subset of the C-tree functionality, and even that
was a very small subset actually. Now I'm sure that B-trees
used in there are a wonder of engineering - however, I simply
have other goals and not enough time to learn them to
appreciate that.

Personally, I think that for most people the _direct_ benefits
of access to source code are greatly exagerrated. I would
place much, much more emphasis on indirect, derived
benefits of availability of source code.

Yes, you CAN read the source code. But the point is, you
DON'T WANT TO.

Because economically speaking, division of labor applies,
and idealistically speaking, it's better to stand on the shoulders
of giants.

--
It's a man's life in a Python Programming Association.
Jul 18 '05 #159
Bulba! <bu***@bulba.com> writes:

[...]
That's remarkable, first time I see smth like this -
out of curiosity, could you say a word where was that?


Are you the same Bulba I know from alt.pl.comp.os.hacking?
Jul 18 '05 #160
al*****@yahoo.com (Alex Martelli) writes:
Yes, apart from libraries and similar cases (frameworks etc), it's no
doubt rare for closed-source "end-user packages" to be sold with
licenses that include source and allow you to "do anything with it".

However, allowing customization (at least for internal use within the
customer organization), while rare, is far from unheard of.


There's no obstacle to doing that with GPL'd software either.
Jul 18 '05 #161
Bulba! wrote:
On Thu, 06 Jan 2005 08:39:11 GMT, Roel Schroeven
<rs****************@fastmail.fm> wrote:
That's generally the goal of the Free Software Foundation: they think
all users should have the freedom to modify and/or distribute your code.


You have the freedom of having to wash my car then. ;-)


A more accurate analogy would be, "You're free to borrow my car, but
if you do, you must wash it and refill the gas tank before you return it."

Note that the so-called 'viral' nature of GPL code only applies to
*modifications you make* to the GPL software. The *only* way in which
your code can be 'infected' by the GPL is if you copy GPL source.

Given the standard usage of closed-source software, you never even
have access to the source. If you use GPL software in the same way
that you use closed-source software, then the GPL cannot 'infect'
anything you do.

The 'infective' nature of the GPL *only* comes when you make use of
the *extra* privelidges that open source grants. So yes, those extra
privelidges come with a price (which is that you share what you've
done); but if you don't want to pay that price, you have the choice of
not using those privelidges. This does not, in any way, prevent you
from using GPL'ed software as a user.

(Problems may come if someone licenses a library under the GPL; that's
what the LGPL was invented for. But the issue here is not that the
GPL is bad, it's that the author used the wrong form of it.)

Personally, I'm not a big fan of the GPL. I'm much more likely to use
BSD-ish licenses than [L]GPL. But it still bugs me to see the GPL
misrepresented as some plot to steal the effort of hardworking
programmers -- it is, instead, an attempt to *encourage* hardworking
programmers to share in a public commons, by ensuring that what's
donated to the commons remains in the commons.

Jeff Shannon
Technician/Programmer
Credit International

Jul 18 '05 #162
On Thu, 06 Jan 2005 12:20:35 +0100, Stefan Axelsson
<cr******@hotmail.com> wrote:
If GPL folks had their way, it would not be possible not to "share"
_anything_ you create. It is widely acknowledged that GPL
license has the "viral" aspect of extending itself on your
software - can you point to closed-source licenses that would
have this aspect? None of the licenses I've read except GPL has
this aspect.
Then you haven't read very many source code licenses, many (perhaps
most?) that state that if you've as much as looked at the code you're
not even allowed to write somethings similar twenty years down the line,
or anything that remotely resembles something similar.
I've read Novell license of internal development tools it provides
(which I reviewed for some purpose). This is I think relevant
part:

"Novell grants you a non-exclusive license to use the
Software free of charge if (a) you are [...] (b) you are
a contracted vendor (c) your use of the Software is for
the purpose of evaluating whether to purchase an ongoing
license to the Software.
[...]
You may not:
* permit other individuals to use the Software except
under the terms listed above;
* permit concurrent use of the Software;
* modify, translate, reverse engineer, decompile,
disassemble (except to the extent applicable laws
specifically prohibit such restriction), or create
derivative works based on the Software;
* copy the Software other than as specified above;
* rent, lease, grant a security interest in, or
otherwise transfer rights to the Software; or
* remove any proprietary notices or labels on the
Software.

Novell may have patents or pending patent applications,
trademarks, copyrights, or other intellectual property
rights covering the Software. You are not granted any
license to these patents, trademarks, copyrights, or other
intellectual property rights except as expressly provided
herein. Novell reserves all rights not expressly granted."

Other than that, the license had to do only with usual
stuff of disclaimers, different jurisdictions, export
controls, "US govt restricted rights", etc. Didn't find
there anything that forbids me to develop smth similar,
unless it's very well hidden in hooking into the technicalities
of specific intellecatual property laws.

I've also read similar IBM licenses -- granted, not very
carefully.

The thing that pissed off various bosses most strongly
in my impression was that the EULAs typically limited
or prohibited transferring the rights to use this thing,
e.g. to spin-offs or organizations cooperating with us.
It's bothersome in practice. None of that had anything
to do with source code or reverse-engineering or
developing similar products (if we wanted to develop
similar products, we would not be using this thing
in the first place, except for 'discovery' purposes).

I'm not saying licenses like you claim don't exist. Sure,
they may exist and they suck.

The point is, they have _limited impact_ and by
the very fact of their "exclusion" nature, this
aspect tends to repel users than attract them to
use this thing.
(Most do in fact
Now, Stallman might or might not want to achieve world domination, not
by sharks with lasers on their heads, but by aiming for all software to
be free software, but the GPL is actually a lot less ambitious than
that. All the GPL says is that: if you received a binary, the person who
provided you with it, must provide you with the source code that built
it. *All* the source code, not just what he happened to receive, on the
off chance that he's modified it.
Which I find again wrong: suppose this developer used GPL-ed
library A, developed patches B and C. He provided you with
the source code of publicly available library A and a patch
C, but he doesn't want to release patch B.

Now, I know that it would cost you the effort to recreate
patch B or maybe it wouldn't work without modifying
patch C. But that's an economic concern, which is orthogonal
to GPL, as GPL is (supposedly) about "free as in speech, not
free as in beer".
And as having the source code without
being able to modify it would be rather pointless, you're allowed to do
that too, it's a given. If you don't want to distribute binaries, that's
fine, and all of the GPL falls. The GPL doesn't *force* you to share
anything. It only says what must happen if you do.
True, not directly.

However, by attempting to create the context in which it's IMPRACTICAL
to use any non-GPLed software, it's attempting to achieve vendor
lock-in effect, and while in such situation you FORMALLY would not
have to give away, you PRACTICALLY would have to give away.

It's not there in license that you "have to give away whatever you
create". It's just attempting to create particular _economic_
context with those licenses, just like the licenses and practices
of a well-known vendor do, even though that is not spelled out
explicitly, obviously.
And I'm rather tired of the GPL's so called 'viral' nature. Look, if
you're using my code, you play by my rules, that's called copyright.
Of course. The point is, _what is specific content of those rules_.

Python's copyrighted, too:

http://www.python.org/2.3.2/license.html

And yes, we have to play by those rules, too. That obviously
doesn't mean that specific conditions are the same, and
those specific conditions are the crux of controversy.
If
you don't want to play by my rules, fine, don't use my code.
Do I have to point at who is also attempting to achieve this
particular context - "you don't like our licenses, don't use our
software"?
So far I'm
no better than Microsoft, or Sun (though that might change) or IBM for
that matter. With the GPL I'm actually not as bad as that, I'll even let
you look at the code, modify it, and distribute copies willy nilly
(though you I'm not forcing you to), in fact, I'll even *forbid* others
from taking that right away from you. If you use it, however, as a small
token of your appreciation, you'll have to also agree to not take the
same rights you had away from others.
I don't believe person A automatically has this sort of "right" of
reading/using whatever they want.
Finally, what *you* do with *your* code is of no concern to the GPL. As
long as you don't use *my* code you can do whatever you please. But, and
that's a big 'BUT', it really irks me when people release code under
e.g. the BSD (or as has happened to me in the past, public domain), and
then bitch and moan when I incorporate parts of it in *my* software and
release the whole under the GPL. As if that was somehow 'unfair'.
In a way, it is: you're taking liberties with their source and use it
in a way they did not _intend_ to. Just because this person doesn't
lock they house in hope other people will be considerate enough not
to rearrange the furniture in it just for fun...

Yes, their license _allows it_.

So what.
Look,
(and I'm obviously not saying this to the parent poster as he never
expressed any such sentiment, I'm just venting) that's what *you* wanted
when you released the code under that license. If you don't want me to
do that, then don't use those licenses, mkay.


Sure, if too many people do that, they will "lock" their house. And
only then I think we will be back on the way to the 17th century style
"secret science".

You're talking about taking the maximum liberties of agreements
and situations to achieve ideological goals. I claim that maybe it
would be nicer to base this thing on good-willed consensus rather
than paranoia and legal technicalities.


--
It's a man's life in a Python Programming Association.
Jul 18 '05 #163
Jeff Shannon <je**@ccvcorp.com> writes:
Note that the so-called 'viral' nature of GPL code only applies to
*modifications you make* to the GPL software.
Well, only under an unusually broad notion of "modification". The GPL
applies to any program incorporating GPL'd components, e.g. if I
distribute a Python compiler that incorporates some component from
GCC, then my entire Python compiler must be GPL'd even though the GCC
component is isolated inside the Python compiler and I wrote the rest
of the Python compiler myself. If I don't like this, I have an
obvious recourse, which is don't use GCC components in my Python
compiler.

The notion here is that the GCC components are attractive enough that
being able to use them provides an incentive to GPL my Python
compiler, which I might not do otherwise.
(Problems may come if someone licenses a library under the GPL; that's
what the LGPL was invented for. But the issue here is not that the
GPL is bad, it's that the author used the wrong form of it.)
The "problem" is not a problem except that in the case of some
libraries, simply being able to use a library module is often not
enough incentive to GPL a large application if the library module's
functionality is available some other way (including by
reimplemntation). If the library does something really unique and
difficult, there's more reason to GPL it instead of LGPL'ing it.
The 'infective' nature of the GPL *only* comes when you make use of
the *extra* privelidges that open source grants. So yes, those extra
privelidges come with a price (which is that you share what you've
done); but if you don't want to pay that price, you have the choice of
not using those privelidges. This does not, in any way, prevent you
from using GPL'ed software as a user.


Well put.
Jul 18 '05 #164
In article <1g***************************@yahoo.com>,
Alex Martelli <al*****@yahoo.com> wrote:
Jul 18 '05 #165
On Thu, 06 Jan 2005 09:27:49 -0500, Steve Holden <st***@holdenweb.com>
wrote:

I'd go further. It's not possible to force anyone to share, but the
GPL aims to remove software from a system that instead aims to force
people NOT to share.
Nope. IMHO, GPL attempts to achieve the vendor lock-in. For different
purposes than another well-known vendor, but it still does.
Well you are entitled to your opinion. But *my* opinion is that the GPL
attempts to ensure that if you re-use code by an author who so desires,
then redistribution of your code is only possible by making your own
extensions to it available on the same terms. This gives you a clear choice.
I agree with you. However, I don't see how your statement contradicts
mine.
To put it another way, it allows an author to specify that their code
can't be hijacked for proprietary purposes *in distributed programs*.
How can the source code that is _guaranteed to stay as public
availability_ be _hijacked_?

If it's hijacked, it's not available anymore.

Making derived work proprietary in no way implies that the base
work is publicly unavailable anymore.
I
will specifically point out that there is *nothing* in the GPL that
requires you to reveal the source of program you write but do not
distribute, even when such programs incorporate tons of GPL'd code.
Again, I don't see why that negates my thesis of vendor lock-in:
whatever software that uses GPLed code crosses inter-organizational
or inter-personal border, it has to be released with source.
It's actually even worse: the only thing you can't share on a
well-known vendor's platform is the software written by that
well-known vendor -- you can choose to share or choose not to
share whatever you or other people write on this platform. Well that's way over-simplified. And if you mean Microsoft, *say*(
Microsoft.
Oh can't you take a little joke, why do we have to be so serious..

If my allusion was not funny, well, sorry.
The GPL folks are quite happy to have you "share" anything that *you*
create.
Oh absolutely, and I would be happy with them washing my car
for free. ;-)
Their simply-stated and elegantly-achieved intent is that you
don't "share" anything that *they* create except on the terms they have
required for their creations.
But their base work is available anyway, regardless of whatever
I do or don't do.
So, it seems to me, you are whining because the authors of GPL'd code
don't want you to release *their* code except under the GPL.
If that was limited to _primary_ effects, that would be
understandable. Which is why I'm rather fine with LGPL for
instance.

However, an openly stated goal is an indirect effect: achieving
the goal of "all the software in the world being free" (as in
their definition of freedom).

Which means that indirect, _economic_ result they hope to
achieve is precisely creating a practical context when this author
would have hard time to release his work under license other
than GPL.

Why do they call "library GPL" a "lesser" GPL, Steve, and
do not really like it? Is it not for the sake of this goal?

Watch this carefully: if what you claim was ALL they
care for, there would be no big difference for them between
LGPL and GPL. And yet for them it is quite a big deal.
What gives
*you* the right to dictate to them?
Conversely, what gives them the right to dictate the authors
of derived works of what they do with THEIR part of work?
How would you like it if Richard
Stallman insisted that you release your code under the GPL? Which, of
course, he doesn't.
Oh but he does - just indirectly. He's attempting to create such
context. GPL is a sort of wolf in a sheep's skin, while Stallman
pretends it's not really a wolf, and then preaches how wonderful
it will be when we will sit with millions of such sheep at the
table and vote what's for lunch.
As the MPAA knows, people do want to share, and
forcing them not to do so is impossible without turning the world into
a police state.

Socialism is unpopular for many reasons, and many of them are indeed to
do with maintaining the separation between individuals and thereby
retaining the ability to treat them as separate economic units. But we
aren't going to change that by insisting on particular software
licenses. Realize this is a very small part of a very large debate.
Absolutely. I have discussed intellectual property rights issues with
friends to great lengths, not just regarding the software.
And that is their choice. They should realize, however, that some
licenses (including the more recent Python licenses) are cleared as
"GPL-compatible". I believe this means that if I receive software
licensed under a GPL-compatible license, I am at liberty to distribute
it under the GPL. I suspect that this point is far too infrequently stressed.
I really don't find it very important: where the main battle
is, and where some vendors achieve domination and some
fail are precisely indirect economic effects of what they
do.

<snip> Actually, I get the impression that GPL-ed software is written by
programmers for programmers, not really for end users.

Not at all. It's written to be redistributed under specific terms, and
anyone who doesn't like those terms has the option of redeveloping the
functionality for themselves.


But they won't. And most of the time they never do. That is the very
point.

It's a subtle game: what you are _allowed_ to do intertwines with
practical situations and what you would _will choose_ to do
given how many factors influence your decisions.
You can't insist that people give you their intellectual property on
*your* terms.
God forbid! This certainly not what I meant, ever, and if
anybody suggests that, I have this rabbit right here that I
will release to get them. :-)
That would be like insisting that the music industry bring
down the price of their clearly-overpriced products, or that the
Baltimore Orioles stop the concession stands from charging $4.50 for a
one-dollar beer. If you want a vote in such situations then your feet
are the appropriate instrument. Walk away, and stop whining :-).
Insisting will do you no good.
Absolutely.

However, what you present is very partial picture: there's much
more to it.
It's a man's life in a Python Programming Association.

Since I'm taking issue with you, I will end by gently pointing out that
there's a substantial minority (? - my impression) of people who might
find your tag line (which I am sure is intended to be supportive of
Python and the c.l.py ethic, such as we might agree exists),
gender-biased and therefore just as unacceptable to them as the GPL
appears to be to you.


You haven't seen the episode of "Owl Stretching Time" by MP I see.
:-) No worries, you just need a little re-education. ;-)


--
It's a man's life in a Python Programming Association.
Jul 18 '05 #166
On Thu, 06 Jan 2005 09:42:42 -0500, Steve Holden <st***@holdenweb.com>
wrote:
You see, I'm not disagreeing with you that your model applies
_where it applies_. I only disagree that it applies in face of
stronger forces. Now what kind of forces is dominant in
most frequent scenarios would have to be worked out in tedious
empirical research I think. Which I haven't done, because
learning some economics is just a hobby to me.
Yes, by all means let's just spout our opinions without any of that
inconvenient tedious empirical research which might invalidate them.
Err, but what did I do that got you visibly pissed off?

I certainly did not mean offending anyone. If I did smth
that did, I apologize, but honestly I didn't mean that. I
just expressed my opinion and cited some small bits of evidence,
which I think I'm entitled to.
Capital investments per worker in auto industries are reportedly
very high. Simple physical tasks are done largely by machines,
like this 100 million Deutschmark costing laser-cutting installation
that I've seen there, where a pile of iron bars is pulled in at one
end and the pile of ready components is spitted out of the other end
(unlike typical thermal cutting, laser has the advantage of not
destroying the metal structure adjacent to the cut, so the parts
of the machines subject to high-stress are oft produced this way). The same is true of plasma-arc cutting for thicker steels, and I believe
it's still not possible to cut 3-inch stainless with a laser. But what's
your point?
<shrug> I was just explaining the issue for someone who could
wonder "why bother with cutting that with laser". The components
of those machines, even bigger ones, typically were not as thick
as 3 inches.
Oh, and by the way that installation doesn't get used much.
Somebody at the office didn't check carefully enough the
energy prices before ordering it and later someone discovered
that off-site specialized cutting firms that take advantage of
energy available at low prices at special times in other countries
can get it produced cheaper. Moving it elsewhere or selling
is not an option, since it is a specially constructed, low, 50-meters
long hall that stands inside the huge manufacturing hall of the
company. And you are using this example to try and argue that engineers are
better-educated than sales people?
Nope. The context was that behavior of companies tends to
be highly rational, optimized and not wasting resources. My
naturally individual experience was that it was oft not the case,
and that was the example.

Which was my point when explaining the clustering that
demonstrably happened: if the behavior of decisionmakers
is highly informed, rational and not really induced much by
risk avoidance as Alex claims, then the clusters are created
by "natural economic forces".

However, if the process is not that rational, then maybe
clusters are the correlation of "cover your ass" aspect
in managers' behavior all wanting to get their branch office
in yesterday in Tokyo, today in Beijing, and during
speculative craze in Russia in Moscow "because everybody
is doing that". Which observations of Paul Krugman on
"defective investors" seem to support.

Now, I'm very strongly opposed to saying that all that
somehow invalidates economics, including economics
of software, as _science_.

All I'm saying is that maybe this particular model is not
what some people think it is. This is the problem with
economics, people tend to get hot under the collar about
it for some reason and it's oft hard to debate that calmly.
Which personally I find a pity, because e.g. economics
of software is such an interesting subject..
Who sold this installation? Who
bought it?
I have no idea, as I were not a manager there and it
didn't really pertain my work.
I was utterly shocked. Having grown up in Soviet times I have
been used to seeing precious resources wasted by organizations
as if resources were growing on trees, but smth like this?! In a
shining ideal country of Germany?! Unthinkable. Indeed not. Quite often the brown paper bag is a factor in purchases
like this. I wouldn't be at all surprised if somebody with a major input
to the decision-making process retired to a nice place in the country
shortly afterwards. You appear to be making the mistake of believing
that people will act in the larger interest, when sadly most individuals
tend to put their own interests first (some would go as far as to define
self-interest as the determinant of behavior).
But there is a subtler point here: most likely it was NOT in the
short-term personal interest to make this mistake (as I believe
corruption was not the case in this decision)!

After all, whoever responsible was still running the considerable risk
of getting fired. It is an example that precisely negates either
collective or individual, long-term or short-term, interest was
primary factor in this decision.
The firm I was working for had a consensus decision-making process (even
I was involved) and managers (and other employees) and stockholders were
mostly the same people -- it wasn't all that large a firm at the time.
Nobody needed to practice risk avoidance.

Again, you may have had good luck. Where I worked (including
some places in Germany and UK) it was almost the only factor
that seemed to matter to people - they'd do ANYTHING not to
take a risky decision, to "pass the buck", not to stick their necks
out, not to declare doing some work that involved challenges.

Some people are like that. I chose a long time ago to try not to work
with them whenever I could avoid it and, while that may have had
negative economic consequences I an convinced it has improved my quality
of life immensely. Of course, I have no proof for such an assertion.


Which in economic terms could mean that your "utility function"
is "modeled" in this particular way (well, strictly speaking utility
functions regard simple consumption), while most people tend
to misunderstand it as the idea that supposedly is "vulgar
consumption of as much stuff as possible is what makes people
happy" and they feel rightly repulsed from such an idiotic idea.
The trouble is, well-informed people do not to argue that, while
people tend to think economists actually do...

--
It's a man's life in a Python Programming Association.
Jul 18 '05 #167
Bulba! <bu***@bulba.com> writes:
Making derived work proprietary in no way implies that the base
work is publicly unavailable anymore.


Since you want to be able to incorporate GPL code in your proprietary
products, and say there's no problem since the base work is still
available from the same places it was available from before, fairness
would say you shouldn't mind that people incorporate code from your
products into THEIR products, since your version is still available
from you.

Really, you're just a freeloader looking for handouts.
Jul 18 '05 #168
On Thu, 06 Jan 2005 10:38:53 -0800, Jeff Shannon <je**@ccvcorp.com>
wrote:
It's also noteworthy to consider that many times, waste happens not
because of corruption or self-interest, but simply because of errors
of judgement.
Precisely.

That is one of the main points I was trying to get across in
discussion with Alex. I have no reason to make conspiracy
theories that bribes were involved in that bad decision
at that German company. No, it _seemed_ like it was simple
mistake, because somebody responsible for this did not
research the issue in enough depth. And note that it
was definitely not in his personal interest, whoever that
was, a person or group of persons, as he/they risked getting
fired for that.
Humans being as we are, it's inevitable that over time,
some "obvious" important details will escape our attention, and the
resulting imperfect information will result in poor decisions. This
is a simple fact of human nature, and (ob-Python ;) ) it's one of the
reasons that Python is designed as it is -- it makes a serious effort
to reduce the number of details that might escape detection.
I suspect it is one of the reasons why many people switch from
learning Perl to learning Python (at least I did - I simply
could not remember after a month or two what the hell
I concocted in this place in Perl program, it felt as if this
were somebody else's code).
(One should also consider that many business failures are a case of
simply having played the odds and lost. Many ventures depend on
outside events playing in a certain way; when by chance those events
happen, the decision-makers are called "bold and insightful", but if
things don't work out, they're called foolish or misguided.


Again, I agree with you - all that I can add is that it this is what
may be a rational element in the otherwise irrational decisionmaking
process - managers may get hit if it goes wrong, but not rewarded
if it goes right. Consider this:

"Moreover the fact that a CEO can command does not mean that other
employees will obey. Instructions can be given, but they need to be
obeyed enthusiastically by others for them to mean anything. CEOs have
tools to win the enthusiasm of their subordinates: the rhetoric of
shared accomplishment of action and vision; the carrots of promotions,
salary increases, and bonuses; the sticks of demotion and dismissal.
But even with these tools, managing a large bureaucratic organization
is a difficult task. And changing its direction away from that of mere
business-as-usual requires great skill and luck."

http://www.j-bradford-delong.net/Eco...porations.html


--
It's a man's life in a Python Programming Association.
Jul 18 '05 #169
In article <10*************@corp.supernews.com>,
Jeff Shannon <je**@ccvcorp.com> wrote:

Note that the so-called 'viral' nature of GPL code only applies to
*modifications you make* to the GPL software. The *only* way in which
your code can be 'infected' by the GPL is if you copy GPL source.


That's not true -- consider linking to a GPL library.
--
Aahz (aa**@pythoncraft.com) <*> http://www.pythoncraft.com/

"19. A language that doesn't affect the way you think about programming,
is not worth knowing." --Alan Perlis
Jul 18 '05 #170

"Bulba!" <bu***@bulba.com> wrote in message
news:no********************************@4ax.com...
Which I find again wrong: suppose this developer used GPL-ed
library A, developed patches B and C. He provided you with
the source code of publicly available library A and a patch
C, but he doesn't want to release patch B.


Then he does not have to. As I understand the GPL, as long as he does not
release (distribute) the patch in any form (in particular, binary), then
the GPL has no effect.

What strikes me as funny about GPL sniping is that many programmers,
including I am sure some of the snipers, sign Terms of Employment contracts
far more restrictive of their freedom than one could possibly accuse the
GPL of being. But I have seen little or no discussion of this (at least on
clp). In fact, I wonder if the GPL might be a substitute target.

Terry J. Reedy

Jul 18 '05 #171
On 06 Jan 2005 15:38:53 -0800, Paul Rubin
<http://ph****@NOSPAM.invalid> wrote:
Making derived work proprietary in no way implies that the base
work is publicly unavailable anymore.
Since you want to be able to incorporate GPL code in your proprietary
products,
Nope. That is not what I'm arguing. Really, I think you have
jumped to conclusion about that: I merely pointed out that
I don't like what I perceive as end effect of what GPL license
writers are attempting to achieve: vendor lock-in.

I think I stated that clearly.
and say there's no problem since the base work is still
available from the same places it was available from before, fairness
would say you shouldn't mind that people incorporate code from your
products into THEIR products, since your version is still available
from you.
I merely pointed out that previous poster's argument about
"hijacking" the OSS product: that it's just not possible
as long as this person is not in legal position to make _base_
work proprietary.

I think I stated clearly: base work is still available regardless
of whatever derived work creators do or don't do - esp. that
they tend to release only binaries, thus making it impossible
to create further derived works!

From the viewpoint of looking at availability of source code A,
it's completely irrelevant if those guys are fishmongers or
make derived work A' and redistribute only binary of A'. Not
a single line of publicly available source code appeared or
disappeared as the result of whatever they do. Amounts of
binaries - yes, that is affected. But not the source code.
Really, you're just a freeloader looking for handouts.


Rest assured this is not my motivation, esp. that I
attempt not to use the GPL-ed software whenever
I reasonably can (e.g. it's rather hard to abstain from
using gcc sometimes, as you oft have a hard time
compiling this damn thing with anything else -- see,
the beginnings of vendor lock-in appear). And I thought
I stated clearly that basically I have no problem with
LGPL - which requires distributing modifications
of the _base_ work, but not _your_ code.

Oh, and "freeloading" argument really doesn't make much
sense: since that software is available to EVERYONE,
its availability / unavailability in the economic sense it
is merely reducing / increasing costs equally for everyone
(ceteris paribus, assuming both vendor A and B find
that OSS product equally useful). I like to think of
OSS as a "tide that rises all boats".

--
It's a man's life in a Python Programming Association.
Jul 18 '05 #172
On 06 Jan 2005 14:16:13 -0800, Paul Rubin
<http://ph****@NOSPAM.invalid> wrote:
Yes, apart from libraries and similar cases (frameworks etc), it's no
doubt rare for closed-source "end-user packages" to be sold with
licenses that include source and allow you to "do anything with it".

However, allowing customization (at least for internal use within the
customer organization), while rare, is far from unheard of.
There's no obstacle to doing that with GPL'd software either.


Which is absolutely true, but generally it's not the main _expected
effect_ that license writers aim to achieve..


--
It's a man's life in a Python Programming Association.
Jul 18 '05 #173
"Terry Reedy" <tj*****@udel.edu> writes:
Which I find again wrong: suppose this developer used GPL-ed
library A, developed patches B and C. He provided you with
the source code of publicly available library A and a patch
C, but he doesn't want to release patch B.


Then he does not have to. As I understand the GPL, as long as he does not
release (distribute) the patch in any form (in particular, binary), then
the GPL has no effect.


I think the hypothesis is that the developer distributed the patched
library. The GPL then requires distributing source for all the
patches.
Jul 18 '05 #174

"Steve Holden" <st***@holdenweb.com> wrote in message
news:rZbDd.68969$Jk5.62111@lakeread01...
Absolutely not. Some people want to share under very specific conditions,
hence the proliferation of licenses in the open source world.


Indeed, Prof. Lessig at Standford University, I believe, recently designed
the Creative Commons license with a checklist of options with the intention
of allowing and thereby encouraging people to share under the specific
conditions they are willing to share under.
http://creativecommons.org/licenses/

Terry J. Reedy

Jul 18 '05 #175
In article <tt********************************@4ax.com>,
Bulba! <bu***@bulba.com> wrote:

Nope. That is not what I'm arguing. Really, I think you have
jumped to conclusion about that: I merely pointed out that
I don't like what I perceive as end effect of what GPL license
writers are attempting to achieve: vendor lock-in.

I think I stated that clearly.


And my counter-argument is that I believe your perception is wrong. If
I agreed with your focus on lock-in, I'd say that what the GPL is trying
to lock in is a specific legal philosophy by subverting the existing
system.
--
Aahz (aa**@pythoncraft.com) <*> http://www.pythoncraft.com/

"19. A language that doesn't affect the way you think about programming,
is not worth knowing." --Alan Perlis
Jul 18 '05 #176
On Thu, 06 Jan 2005 14:27:55 -0800, Jeff Shannon <je**@ccvcorp.com>
wrote:
That's generally the goal of the Free Software Foundation: they think
all users should have the freedom to modify and/or distribute your code.
You have the freedom of having to wash my car then. ;-)

A more accurate analogy would be, "You're free to borrow my car, but
if you do, you must wash it and refill the gas tank before you return it."
That analogy is an equivalent of LGPL. Not GPL. GPL in analogy would
require you to donate whatever you created thanks to studying the
construction of that car in the meantime. And yes, again, sure
that is what the borrower agreed to do, sure that NOT using it
at all does not result in such obligations, and again - that is not
the point.
Note that the so-called 'viral' nature of GPL code only applies to
*modifications you make* to the GPL software. The *only* way in which
your code can be 'infected' by the GPL is if you copy GPL source.
Given the standard usage of closed-source software, you never even
have access to the source.
Which means that standard closed-source software does not impose
itself on your software - except those detestable cases where
Stefan Axelsson pointed out, it prevents you from developing
smth similar if you agreed to such a pre-condition in EULA of this
thing.

It would appear that such a software attempts to economically
lock the user in using the only binary in the world that implements
this functionality; and FSF attempts to economically, indirectly
lock the developers - and indirectly users - in "being pressured
to give away" on the rationale of sharing being so good for
everyone.

Now, this may not be what they INTENDED - but from my
viewpoint it seems like that is the RESULT.

Not only I see this as unfair, but also as counter-effective
in getting more of people into both using and developing
software with publicly available source code ("free software"
guys tend to get hot under the collar when they hear
"open source").

Trying to bully people into this thing _for their own good_
tends to have the effect opposite to the intended. You
achieve more by showing positive attitude rather than via
conspiracy theories and paranoia. It is also this mostly
friendly and relaxed aspect of Python community that I
like so much.

Again, agreeing / not agreeing and resulting use / walking
away _are not the issue_. I don't see MS defended like "well
if you don't like whatever MS does, just don't use their
software, so shut up about whatever licenses MS actually
has and about whatever it does!".
If you use GPL software in the same way
that you use closed-source software, then the GPL cannot 'infect'
anything you do.
True, but that abstracts from source code issues, doesn't it?
The 'infective' nature of the GPL *only* comes when you make use of
the *extra* privelidges that open source grants.
Those extra privileges are the only way of _building_ large
software systems, isn't it? So I would define it as "normal".

Yes, closed-source is sort of "crippled" in this regard.
So yes, those extra
privelidges come with a price (which is that you share what you've
done); but if you don't want to pay that price, you have the choice of
not using those privelidges. This does not, in any way, prevent you
from using GPL'ed software as a user.
Obviously; but while what you wrote is true, it is not the crux of
the problem.
(Problems may come if someone licenses a library under the GPL; that's
what the LGPL was invented for. But the issue here is not that the
GPL is bad, it's that the author used the wrong form of it.) Personally, I'm not a big fan of the GPL. I'm much more likely to use
BSD-ish licenses than [L]GPL.
Personally, I think that LGPL in abstract sense does make sense: if
you use and modify this thing, you should return back the _modified_
part - and this should apply regardless whether you keep it in private
or not, release in binary or in source to anybody.

However, it's definitely safer and since it also FEELS more benign,
BSD-like licenses are probably more productive in terms of motivating
people to cooperate, so I agree with you on that point.
But it still bugs me to see the GPL
misrepresented as some plot to steal the effort of hardworking
programmers -- it is, instead, an attempt to *encourage* hardworking
programmers to share in a public commons, by ensuring that what's
donated to the commons remains in the commons.


OK, a quick sanity check: does Python not remain "in the commons"
because some people use it in closed-source applications? I would
say that ALL of software released under GPL, LGPL or other free /
open source licenses remains in the commons.

All of the controversies seem to be about the derived works, don't
they?

--
It's a man's life in a Python Programming Association.
Jul 18 '05 #177
On Thu, 06 Jan 2005 22:42:01 +0100, Peter Dembinski <pd***@illx.org>
wrote:
[...]
That's remarkable, first time I see smth like this -
out of curiosity, could you say a word where was that?
Are you the same Bulba I know from alt.pl.comp.os.hacking?


No, but I like to see I have an evil twin out there. :-)


--
It's a man's life in a Python Programming Association.
Jul 18 '05 #178
aa**@pythoncraft.com (Aahz) writes:
I don't like what I perceive as end effect of what GPL license
writers are attempting to achieve: vendor lock-in.
And my counter-argument is that I believe your perception is wrong. If
I agreed with your focus on lock-in, I'd say that what the GPL is trying
to lock in is a specific legal philosophy by subverting the existing
system.


Bulba's perception is wrong in a different way too. The GPL tries to
encourage the enlargement of a public resource, namely a codebase
shareable by a worldwide community of users and developers, a cyber
equivalent of a public park whose facilities anyone can visit and use
but nobody can take away and exclude others from using. The idea of
referring to such a shared community resource as a "vendor" is just
bizarre.
Jul 18 '05 #179
Bulba! wrote:
.... And note that it
was definitely not in his personal interest, whoever that
was, a person or group of persons, as he/they risked getting
fired for that.


This doesn't necessarily follow. The decision-maker in question may
have received a fat bonus for having found such a technically
excellent manufacturing process, and then moved into a different
position (or left the corporation altogether) before construction was
complete and the power-cost issue was noticed. That person may even
have *known* about the power-cost issue, and forged ahead anyhow due
to the likelihood of such a personal bonus, with the intention of no
longer being in a bag-holding position once the problem became general
knowledge.

Of course, this discussion highlights the biggest problem with
economics, or with any of the other "social sciences" -- there's
simply too many open variables to consider. One can't control for all
of them in experiments (what few experiments are practical in social
sciences, anyhow), and they make any anecdotal evidence hazy enough to
be suspect.

Jeff Shannon
Technician/Programmer
Credit International

Jul 18 '05 #180
Cameron Laird wrote:
In article <1g***************************@yahoo.com>,
Alex Martelli <al*****@yahoo.com> wrote:
.
.
.
One last reflection -- I believe there are or used to be some programs
written by people no doubt of very good will, distributed with all
sources and often with no profit motive at all, which are NOT open
source because they include in the license some restrictive clause, such
as "no military use", "no use by organizations which perform testing of
cosmetics on animals", or something of that kind. These would be
examples of closed-source software which DO allow ALMOST any kind of use
-- any EXCEPT the specific one the authors dislike so intensely.

While most people may not think of such programs as "closed source",
they most definitely ARE: the definition of open source is very strict
about this aspect.
Alex

With my mathematical background, I'm consistent about calling
these "non-open" rather than "closed". I don't insist others
adopt my nomenclature ...


I'm with Cameron on this one.

--
Robert Kern
rk***@ucsd.edu

"In the fields of hell where the grass grows high
Are the graves of dreams allowed to die."
-- Richard Harter
Jul 18 '05 #181
Bulba! wrote:
On Thu, 06 Jan 2005 09:42:42 -0500, Steve Holden <st***@holdenweb.com>
wrote:
You see, I'm not disagreeing with you that your model applies
_where it applies_. I only disagree that it applies in face of
stronger forces. Now what kind of forces is dominant in
most frequent scenarios would have to be worked out in tedious
empirical research I think. Which I haven't done, because
learning some economics is just a hobby to me.
Yes, by all means let's just spout our opinions without any of that
inconvenient tedious empirical research which might invalidate them.

Err, but what did I do that got you visibly pissed off?

I certainly did not mean offending anyone. If I did smth
that did, I apologize, but honestly I didn't mean that. I
just expressed my opinion and cited some small bits of evidence,
which I think I'm entitled to.

I am probably not as pissed off as you might think, but I'm beginning to
find (what I regard as) your naivety a little irritating. However,
nothing I have said was intended to question your integrity. [...]Oh, and by the way that installation doesn't get used much.
Somebody at the office didn't check carefully enough the
energy prices before ordering it and later someone discovered
that off-site specialized cutting firms that take advantage of
energy available at low prices at special times in other countries
can get it produced cheaper. Moving it elsewhere or selling
is not an option, since it is a specially constructed, low, 50-meters
long hall that stands inside the huge manufacturing hall of the
company.
And you are using this example to try and argue that engineers are
better-educated than sales people?

Nope. The context was that behavior of companies tends to
be highly rational, optimized and not wasting resources. My
naturally individual experience was that it was oft not the case,
and that was the example.

Well, anyone who *does* believe that companies are managed in a way
that's any more rational than the way individual humans manage their own
lives is deluding themselves. But perhaps I'm a little clearer now about
*your* thinking on this.
Which was my point when explaining the clustering that
demonstrably happened: if the behavior of decisionmakers
is highly informed, rational and not really induced much by
risk avoidance as Alex claims, then the clusters are created
by "natural economic forces".
The agent of economic forces, however, are always individual humans. It
just isn't possible to factor them out.
However, if the process is not that rational, then maybe
clusters are the correlation of "cover your ass" aspect
in managers' behavior all wanting to get their branch office
in yesterday in Tokyo, today in Beijing, and during
speculative craze in Russia in Moscow "because everybody
is doing that". Which observations of Paul Krugman on
"defective investors" seem to support.

Now, I'm very strongly opposed to saying that all that
somehow invalidates economics, including economics
of software, as _science_.

All I'm saying is that maybe this particular model is not
what some people think it is. This is the problem with
economics, people tend to get hot under the collar about
it for some reason and it's oft hard to debate that calmly.
Which personally I find a pity, because e.g. economics
of software is such an interesting subject..
I can live with that, I guess.
Who sold this installation? Who
bought it?

I have no idea, as I were not a manager there and it
didn't really pertain my work.

I was merely trying to point out that the sales people may well have
worked a flanker on the "better educated" engineers who might have
bought the plant.I was utterly shocked. Having grown up in Soviet times I have
been used to seeing precious resources wasted by organizations
as if resources were growing on trees, but smth like this?! In a
shining ideal country of Germany?! Unthinkable.

Indeed not. Quite often the brown paper bag is a factor in purchases
like this. I wouldn't be at all surprised if somebody with a major input
to the decision-making process retired to a nice place in the country
shortly afterwards. You appear to be making the mistake of believing
that people will act in the larger interest, when sadly most individuals
tend to put their own interests first (some would go as far as to define
self-interest as the determinant of behavior).

But there is a subtler point here: most likely it was NOT in the
short-term personal interest to make this mistake (as I believe
corruption was not the case in this decision)!

After all, whoever responsible was still running the considerable risk
of getting fired. It is an example that precisely negates either
collective or individual, long-term or short-term, interest was
primary factor in this decision.

Well, you would know better than me, but believe me when I say that such
decisions being made for the economic benefit of a single individual are
far from rare.
The firm I was working for had a consensus decision-making process (even
I was involved) and managers (and other employees) and stockholders were
mostly the same people -- it wasn't all that large a firm at the time.
Nobody needed to practice risk avoidance.
Again, you may have had good luck. Where I worked (including
some places in Germany and UK) it was almost the only factor
that seemed to matter to people - they'd do ANYTHING not to
take a risky decision, to "pass the buck", not to stick their necks
out, not to declare doing some work that involved challenges.


Some people are like that. I chose a long time ago to try not to work
with them whenever I could avoid it and, while that may have had
negative economic consequences I an convinced it has improved my quality
of life immensely. Of course, I have no proof for such an assertion.

Which in economic terms could mean that your "utility function"
is "modeled" in this particular way (well, strictly speaking utility
functions regard simple consumption), while most people tend
to misunderstand it as the idea that supposedly is "vulgar
consumption of as much stuff as possible is what makes people
happy" and they feel rightly repulsed from such an idiotic idea.
The trouble is, well-informed people do not to argue that, while
people tend to think economists actually do...

Well, even assuming I were to allow that such an intellectual thing as a
utility function were to apply to my behavior in this case (which would
be allowing economists more credence than I believe most of them
deserve, but what the heck), all we are really taking about is that "it
takes all sorts to make a world".

I'm as fond of what money can buy as the next guy, but there are still
limits on what I will do to acquire it.

So really a utility function is a very dry and abstract way of looking
at human behavior, and is simply incapable (at the present stage of the
game) of modeling human behavioral complexity.


--
It's a man's life in a Python Programming Association.


regards
Steve
--
Steve Holden http://www.holdenweb.com/
Python Web Programming http://pydish.holdenweb.com/
Holden Web LLC +1 703 861 4237 +1 800 494 3119
Jul 18 '05 #182
Bulba! wrote:
I've read Novell license of internal development tools it provides
(which I reviewed for some purpose). This is I think relevant
part: I'm not saying licenses like you claim don't exist. Sure,
they may exist and they suck.

The point is, they have _limited impact_ and by
the very fact of their "exclusion" nature, this
aspect tends to repel users than attract them to
use this thing.


Well, as I didn't catch that we were discussing development tools but
even so, I thought the license you quoted rather strengthened my point
in the section on 'reverse engineering?. (Then again I've seen other's
where that was not the case, was it MS (?) quite a few years back that
didn't allow you to develop an OS using their tool chain? And that was a
binary license.) If you look at actual end user program licenses (such
as Solaris/wxWorks etc) the picture is much clearer.

In any case, it's somewhat beside the point as even the license you
chose to quote took away precisely all those rights that the GPL grants
you. You weren't allowed to (in any way shape or form) to communicate
anything you knew about that source to anyone else.

And I don't buy into your argument that a use of some source code would
go against the 'intended' use by making it GPL. How the hell am I
supposed to know what the author 'intends' if he doesn't tell me?
Especially if, of all the available licenses, choses one that explicitly
permits the sort of use that I envision? I mean, it's not as if many BSD
fans complain that MS took much TCP/IP related code and put it in
Windows (ftp etc). Without sharing anything back with the community. If
that's OK, how am I supposed to know that my use isn't?

It's just the same situation as when I lock my house. That's intended as
a means of communication with the rest of society. It communicates my
intent that I'm not home (or don't want to be disturbed) and that I
don't want them in the house, and it does so more strongly than mere
social convention does. There are after all several members of community
that needs this reminder; notably children. It's emphatically not to
keep thieves out (much) as they won't be much hampered by the ordinary lock.

If you don't want me to use your code in the way I envision then tell
me. The license is the *perfect* place to do so.

Stefan,
--
Stefan Axelsson (email at http://www.cs.chalmers.se/~sax)
Jul 18 '05 #183
Paul Rubin <http://ph****@NOSPAM.invalid> wrote:
al*****@yahoo.com (Alex Martelli) writes:
Yes, apart from libraries and similar cases (frameworks etc), it's no
doubt rare for closed-source "end-user packages" to be sold with
licenses that include source and allow you to "do anything with it".

However, allowing customization (at least for internal use within the
customer organization), while rare, is far from unheard of.


There's no obstacle to doing that with GPL'd software either.


Absolutely -- as long as you don't redistribute, np. But Roel was not
asking about GPL, he was asking about closed-source licenses -- whether
any of them "allow using the [[source]] code *at all*" -- and the answer
is that, yes, quite a few do, in specifically constrained ways.
Alex
Jul 18 '05 #184
Robert Kern <rk***@ucsd.edu> wrote:
...
While most people may not think of such programs as "closed source",
they most definitely ARE: the definition of open source is very strict
about this aspect.
... With my mathematical background, I'm consistent about calling
these "non-open" rather than "closed". I don't insist others
adopt my nomenclature ...


I'm with Cameron on this one.


There is no "official" definition of closed-source as there is of
open-source, but I'm with the Wikipedia:

http://en.wikipedia.org/wiki/Closed_source

"any program whose licensing terms do not qualify as open source".

I'm not disputing it would be useful to draw many distinctions within
the universe of programs with non-opensource licenses, just pointing out
that such distinctions are not currently reflected in a popular
definition. Since it's a wiki, it may be worthwhile editing it to add
some materials to start influencing popular usage and perception, maybe.
Alex
Jul 18 '05 #185
Bulba! <bu***@bulba.com> wrote:
Suppose they want to ensure "no forking" or that "bugfixes
and enhancements of original software are given back".

OK, LGPL is fine for this goal. When you say "they see it


Neither LGPL nor GPL can ``ensure "no forking"'', nor can any other
open-source license.
Alex
Jul 18 '05 #186
Jeff Shannon <je**@ccvcorp.com> wrote:
Note that the so-called 'viral' nature of GPL code only applies to
*modifications you make* to the GPL software. The *only* way in which
your code can be 'infected' by the GPL is if you copy GPL source. ... (Problems may come if someone licenses a library under the GPL; that's
what the LGPL was invented for. But the issue here is not that the
GPL is bad, it's that the author used the wrong form of it.)


Stallman now says that you should use GPL, not Lesser GPL.

http://www.gnu.org/licenses/why-not-lgpl.html

Specifically, he wants library authors to use GPL to impose the viral
nature of GPL on other programs just USING the library -- the very
opposite of what you say about "only applies ... if you copy"!

Quoting RMS from that URL (about Readline, a GPL library):
'''
Releasing it under the GPL and limiting its use to free programs gives
our community a real boost. At least one application program is free
software today specifically because that was necessary for using
Readline.
'''

Until some judge passes some judgment, the intent and effect of GPL must
remain a matter of opinion. But RMS's opinion is probably more
meaningful than mine or yours -- certainly regarding intent, given his
role in designing that license. If he's badly erred, and one day a
judge endorses your opinion and says that a program which copies no GPL
source cannot be infected by GPL, ah well -- then I guess GPL is badly
designed as to putting its intents into practice. But until there is
some strong basis to think otherwise, I believe it's prudent to assume
RMS is probably right, and your statement therefore badly wrong.
Alex
Jul 18 '05 #187
Alex Martelli wrote:
Robert Kern <rk***@ucsd.edu> wrote:
...
While most people may not think of such programs as "closed source",
they most definitely ARE: the definition of open source is very strict
about this aspect.
...
With my mathematical background, I'm consistent about calling
these "non-open" rather than "closed". I don't insist others
adopt my nomenclature ...
I'm with Cameron on this one.

There is no "official" definition of closed-source as there is of
open-source, but I'm with the Wikipedia:

http://en.wikipedia.org/wiki/Closed_source

"any program whose licensing terms do not qualify as open source".


A definition with a nice big "This article may need to be reworded to
conform to a neutral point of view" warning at the top. ;-)
I'm not disputing it would be useful to draw many distinctions within
the universe of programs with non-opensource licenses, just pointing out
that such distinctions are not currently reflected in a popular
definition. Since it's a wiki, it may be worthwhile editing it to add
some materials to start influencing popular usage and perception, maybe.


There seems to be such an edit on the way:

http://en.wikipedia.org/wiki/Talk:Closed_source

--
Robert Kern
rk***@ucsd.edu

"In the fields of hell where the grass grows high
Are the graves of dreams allowed to die."
-- Richard Harter
Jul 18 '05 #188
Robert Kern <rk***@ucsd.edu> writes:
http://en.wikipedia.org/wiki/Closed_source
"any program whose licensing terms do not qualify as open source".


A definition with a nice big "This article may need to be reworded to
conform to a neutral point of view" warning at the top. ;-)
...
There seems to be such an edit on the way:
http://en.wikipedia.org/wiki/Talk:Closed_source


After they're done defining closed source, maybe they can work on
"compact source" (a bounded closed-source program, i.e. one that,
unlike Windows, doesn't try to take over every computer in the world).
Note also from the Heine-Borel theorem that every closed source
program can be covered by some finite collection of open source
programs.
Jul 18 '05 #189
Paul Rubin <http://ph****@NOSPAM.invalid> wrote:
Note also from the Heine-Borel theorem that every closed source
program can be covered by some finite collection of open source
programs.


Every _compact_ one, surely? Quoting by heart from old memories, but,
isn't Heine-Borel about (being able reduce any open covering of X to a
finite subcovering) <-> (X is compact) ...?
Alex
Jul 18 '05 #190
Alex Martelli wrote:
Until some judge passes some judgment, the intent and effect of GPL must
remain a matter of opinion. But RMS's opinion is probably more
meaningful than mine or yours -- certainly regarding intent, given his
role in designing that license.
But it may not have practical effect in an actual dispute. I believe
that in some jurisdictions[1], the important piece of information in
interpreting a license is the common understanding of what the license
means between the disputing parties. The author of the license, if he is
not one of the disputing parties, has no say in what the license means
for that dispute.

[1] IANAL; TINLA. This is from my memory of what I read in Lawrence
Rosen's book[2] that I don't have in front of me right now. See his
book, chapter 12, I think, for more details.

[2] http://www.rosenlaw.com/oslbook.htm
If he's badly erred, and one day a
judge endorses your opinion and says that a program which copies no GPL
source cannot be infected by GPL, ah well -- then I guess GPL is badly
designed as to putting its intents into practice. But until there is
some strong basis to think otherwise, I believe it's prudent to assume
RMS is probably right, and your statement therefore badly wrong.


I've always found, especially in light of what I wrote above, the best
thing to do is to ask the author himself what he wants. If he subscribes
to an unreasonable interpretation of the license, it's better that you
found out quickly and avoid getting sued even though you might end up
winning. You also avoid inadvertantly stepping on anyone's toes and
garnering ill-will even if you never go to court.

--
Robert Kern
rk***@ucsd.edu

"In the fields of hell where the grass grows high
Are the graves of dreams allowed to die."
-- Richard Harter
Jul 18 '05 #191
al*****@yahoo.com (Alex Martelli) writes:
Note also from the Heine-Borel theorem that every closed source
program can be covered by some finite collection of open source
programs.


Every _compact_ one, surely? Quoting by heart from old memories, but,
isn't Heine-Borel about (being able reduce any open covering of X to a
finite subcovering) <-> (X is compact) ...?


Yeah, whoops, that's what I meant; your old memories are clearer than
mine. Actually sometimes the definitions and theorems interchange.

I do remember something about Tikhonov's Theorem that says that no
matter how often bounded closed source programs multiply, the product
is still closed source. So, for example, Adobe Acrobat is still
closed source even if you can download it from Adobe's web site
infinitely often. But that theorem doesn't apply to noncompact
(i.e. unbounded) closed source programs. So ordering Microsoft to
release parts of Windows as open source was one of the potential
remedies discussed in the penalty phase of the US Justice Dept's
antitrust suit. Unfortunately, the DoJ lawyers were not good
topologists so they didn't give enough consideration to this
possibility.
Jul 18 '05 #192
Bulba! <bu***@bulba.com> writes:
From the viewpoint of looking at availability of source code A,
it's completely irrelevant if those guys are fishmongers or
make derived work A' and redistribute only binary of A'. Not
a single line of publicly available source code appeared or
disappeared as the result of whatever they do. Amounts of
binaries - yes, that is affected. But not the source code.


From the viewpoint of the availability of Adobe Photoshop, it's
completely irrelevant if I run off a few hundred thousand copies on CD
in a warehouse by the waterfront and then sell them out of the back of
a truck at the flea market. Not a single shrink-wrapped retail copy
of Photoshop disappeared from any stores as the result. But Adobe
will still send the US Marshals to raid my operation with guns and
stuff. I don't think you would approve of my illicit Photoshop
replication either.

So why should the situation be any different with GPL code? If
someone wants to copy it, they need to do so according to the license.
Jul 18 '05 #193
On 6 Jan 2005 19:01:46 -0500, aa**@pythoncraft.com (Aahz) wrote:
Note that the so-called 'viral' nature of GPL code only applies to
*modifications you make* to the GPL software. The *only* way in which
your code can be 'infected' by the GPL is if you copy GPL source.
That's not true -- consider linking to a GPL library.


Will someone please explain to me in simple terms what's
the difference between linking to LGPLed library and linking
to GPLed library - obviously in terms of consequences of
what happens to _your_ source code?

Because if there isn't any, why bother with distinguishing
between the two?

Oh, and by the way - since Python bytecode can be relatively
easily decompiled to source, could it interpreted to "really"
count as source code and not binary? What are the consequences
of releasing code _written in Python_ as GPLed?

Licenses are frigging cans of worms..


--
It's a man's life in a Python Programming Association.
Jul 18 '05 #194
Paul Rubin wrote:
Jeff Shannon <je**@ccvcorp.com> writes:
Note that the so-called 'viral' nature of GPL code only applies to
*modifications you make* to the GPL software.

Well, only under an unusually broad notion of "modification".


True enough. It can be difficult, in software development, to define
a distiction between a situation where two software products are
distinct but cooperative, and a situation where one software product
is derivative of another. Stallman has chosen a particular definition
for use in the GPL; one may debate the value of using this definition
over any other possible definition, but the line had to be drawn
*somewhere*. (And given Stallman's philosophies, it shouldn't be too
surprising that he's drawn it about as broadly as he reasonably could.)
(Problems may come if someone licenses a library under the GPL; that's
what the LGPL was invented for. But the issue here is not that the
GPL is bad, it's that the author used the wrong form of it.)

The "problem" is not a problem except that in the case of some
libraries, simply being able to use a library module is often not
enough incentive to GPL a large application if the library module's
functionality is available some other way (including by
reimplemntation). If the library does something really unique and
difficult, there's more reason to GPL it instead of LGPL'ing it.


To my mind, the intent of the GPL is "use it, but if you change it or
make a derivative, share the changes". With libraries, though, you
*can't* use it without hitting the FSF-specified definition of a
derivative. The LGPL exists to make it clear that, for libraries, the
common meaning of "using" and "changing" are different than they are
for applications.

Of course, there's nothing that stops people from insisting that, if
you *use* their libraries, anything you use them for must be
free-as-in-speech (which is the effect of using the GPL instead of the
LGPL); it's the author's choice what restrictions should be put on the
software. But the usage-restrictions on a library under GPL are more
severe than they are on an application under GPL. The unfortunate
thing, in my opinion, is that a fair number of library authors don't
think about that when they GPL their code.

Jeff Shannon
Technician/Programmer
Credit International

Jul 18 '05 #195
Bulba! wrote:
On 6 Jan 2005 19:01:46 -0500, aa**@pythoncraft.com (Aahz) wrote:

Note that the so-called 'viral' nature of GPL code only applies to
*modifications you make* to the GPL software. The *only* way in which
your code can be 'infected' by the GPL is if you copy GPL source.


That's not true -- consider linking to a GPL library.

Will someone please explain to me in simple terms what's
the difference between linking to LGPLed library and linking
to GPLed library - obviously in terms of consequences of
what happens to _your_ source code?

Because if there isn't any, why bother with distinguishing
between the two?


Releasing a product in which your code is linked together with GPL'ed
code requires that your code also be GPL'ed. The GPL goes to some
lengths to define what exactly "linked together" means.

Releasing a product in which your code is linked together with LGPL'ed
code does *not* require that your code also be (L)GPL'ed. Changes to
the core library must still be released under (L)GPL, but application
code which merely *uses* the library does not. (I've forgotten, now,
exactly how LGPL defines this distinction...)

Jeff Shannon
Technician/Programmer
Credit International

Jul 18 '05 #196
Alex Martelli wrote:
Jeff Shannon <je**@ccvcorp.com> wrote:
Note that the so-called 'viral' nature of GPL code only applies to
*modifications you make* to the GPL software. The *only* way in which
your code can be 'infected' by the GPL is if you copy GPL source.


...
(Problems may come if someone licenses a library under the GPL; that's
what the LGPL was invented for. But the issue here is not that the
GPL is bad, it's that the author used the wrong form of it.)

Stallman now says that you should use GPL, not Lesser GPL.

http://www.gnu.org/licenses/why-not-lgpl.html

Specifically, he wants library authors to use GPL to impose the viral
nature of GPL on other programs just USING the library -- the very
opposite of what you say about "only applies ... if you copy"!

Ah, I haven't kept up on Stallman's current opinions, and was speaking
from the understanding I had of GPL/LGPL as of a number of years ago
(before that article was written).

By "copy", above, I meant "use GPL source in your product". The GPL
defines what it means to use source in a rather inclusive way. That
inclusiveness means that the standard usage of libraries falls under
their definition of "using source". This distinction in the normal
terms of "usage" is what impelled the FSF to create the LGPL in the
first place...

So, I think what I said still (mostly) stands, as long as you look at
it in terms of whether object code is copied into your executable. ;)
It's still true that one can use (in a consumer sense) GPL software
for whatever purpose one wishes, and the restrictions only kick in
when one includes GPL code in another product. Indeed, I should have
used the word "include" rather than "copy"...

(It's hardly surprising that Stallman wants to use whatever leverage
he can get to encourage FSF-style free software...)

Jeff Shannon
Technician/Programmer
Credit International

Jul 18 '05 #197
On Fri, 07 Jan 2005 12:06:42 -0800, Jeff Shannon <je**@ccvcorp.com>
wrote:
Bulba! wrote:
On 6 Jan 2005 19:01:46 -0500, aa**@pythoncraft.com (Aahz) wrote:

Note that the so-called 'viral' nature of GPL code only applies to
*modifications you make* to the GPL software. The *only* way in which
your code can be 'infected' by the GPL is if you copy GPL source.
That's not true -- consider linking to a GPL library.

Will someone please explain to me in simple terms what's
the difference between linking to LGPLed library and linking
to GPLed library - obviously in terms of consequences of
what happens to _your_ source code?

Because if there isn't any, why bother with distinguishing
between the two?


Releasing a product in which your code is linked together with GPL'ed
code requires that your code also be GPL'ed. The GPL goes to some
lengths to define what exactly "linked together" means.


That looks like a typo. The LGPL goes to great length to how you can
link to LGPL software without using either the LGPL or GPL. The GPL
(linked to by fsf.org) merely states:

2. You may modify your copy or copies of the Program or any
portion of it, thus forming a work based on the Program, and
copy and distribute such modifications or work under the terms
of Section 1 above, provided that you also meet all of these
conditions:

Note that the conditions are all those of any program released under
the GPL. Whatever "forming a work based on the Program" means is
whatever you and the copyright owner agree to, or whatever copyright
law considers a derived work in areas you wish to release your code
into. I would suggest consulting a lawyer before getting close to the
line, but you can expect that any legally enforceable restrictions
claimed by FSF and/or RMS to be legally binding on all software
released under the (L)GPL that the FSF owns the copyright of (they
encourage programmers to sign over copyright to the FSF itself).

Releasing a product in which your code is linked together with LGPL'ed
code does *not* require that your code also be (L)GPL'ed. Changes to
the core library must still be released under (L)GPL, but application
code which merely *uses* the library does not. (I've forgotten, now,
exactly how LGPL defines this distinction...)

Jeff Shannon
Technician/Programmer
Credit International


Scott Robinson

Jul 18 '05 #198
Bulba! wrote:
Oh, and by the way - since Python bytecode can be relatively
easily decompiled to source, could it interpreted to "really"
count as source code and not binary? What are the consequences
of releasing code _written in Python_ as GPLed?


Well, to your first question, in a word 'no', it wouldn't count as
source code. To quote the GPL section 3:

"The source code for a work means the preferred form of the work for
making modifications to it. For an executable work, complete source code
means all the source code for all modules it contains, plus any
associated interface definition files, plus the scripts used to control
compilation and installation of the executable."

As the preferred form for making changes to Python programs would be
Python source, that's what counts. This is also what forbids obfuscated
code. If you were to *write* Python bytecode, as a form of assembly,
then of course that's another matter.

I've released Python source as GPL and as far as I'm concerned it ought
to work, even though that's not explicitly covered. As the only way
you're going to receive my program is by receiving the source then
you'll end up having it and everything's basically OK. If someone tries
to make a binary from that and distribute that without also making the
source available then the GPL obviously comes into effect, and the game
is up. I haven't sought legal (or FSF) input on this matter though, it's
just my understanding. You can be fairly confident that the GPL is iron
clad though, it would have been dragged through every court in the land
by now if it wasn't.

I've also followed the LGPL/GPL library debate, and while I have
opinions on that as well, this is getting long in the tooth already.

Stefan,
--
Stefan Axelsson (email at http://www.cs.chalmers.se/~sax)
Jul 18 '05 #199

This discussion thread is closed

Replies have been disabled for this discussion.

Similar topics

12 posts views Thread by tompa1969 | last post: by
385 posts views Thread by Xah Lee | last post: by
15 posts views Thread by Xah Lee | last post: by
reply views Thread by suresh191 | last post: by
By using this site, you agree to our Privacy Policy and Terms of Use.