On Thu, 17 Mar 2005 10:48:42 -0500, Andrew Darby <[log in to unmask]> wrote:
> How would one distinguish between an "open source" project like Linux or
> Apache or Mozilla, and a LAMP/WAMP program developed by a library and
> then made freely available to modify, distribute, etc.
Short answer: The license. If there is no explicit license, full
copyright must be assumed under the law.
Long answer:
Free Software is Free-as-in-speech, and may be free-as-in-beer. A
useful way of thinking about this (due to the differing English
meanings) is the French "libre" = "Free" and "gratis" = "free".
Free Software is a notion introduced by Richard Stallman. His work
has been important in shaping the software landscape. However, he is
staunchly moral and draws distinctions and makes mandates which
limited the popular acceptance of Free Software. Open Source (led by
the Open Source Initiative and mascot'd by Eric Raymond) spun off as a
friendlier, gentler form of less-than-Iron-Curtain software.
It has indeed popularized a middle ground. Unfortunately, it has also
muddied the water. Some people consider "open source" to just mean
that source is available (as in Pirate Source). The Open Source
Initiative has a more expansive definition:
http://www.opensource.org/docs/definition.php and they're pretty
authoritative. They approve specific licenses as OSI certified
(http://opensource.org/docs/certification_mark.php) which is a quick
way to evaluate whether a particular license meets the requirements of
the OSI (http://opensource.org/docs/definition.html).
Open Source projects are only sometimes Libre, and rarely are gratis.
Free Software projects are always Libre, and usually gratis.
You do not necessarily have the right to modify or distribute source
of an "open source" project because of the muddied water above. The
license is what really matters. You do have those rights (and others)
in Free Software. Lowercase "open source" means code is available,
uppercase "Open Source" means the code is licensed under an
OSI-certified license.
You may already be familiar with this, but it sets the context of your
question. What really determines whether an Open Source project is
Free depends on the license the source is released under.
All source falls under copyright law; it is up to the copyright
holders to license it for use to others. Free Software grants an
exceptional number of allowances to the licensee, and usually binds
the licensee in to a similar contract for modified works (CC's
optional share-alike, GNU's copyleft).
Software which is licensed under a (L)GPL-compatible license is
generally called Free. Additionally, there are tons of licenses also
considered Free such as Apache, MIT, and BSD. The general distinction
between OSI certified licenses and Free licenses are due to moral
distinctions.
> I'm specifically thinking of East Carolina University's "Pirate Source"
> (http://systems.lib.ecu.edu/piratesource/) which we have reworked and
> rebranded (rather lamely) as Resources by Subject
> (http://www.ithacalibrary.com/subjects/), but perhaps MyLibrary would
> fall in this category, too.
Pirate Source appears to be (lowercase) open source since you can
download the source. We can't determine what rights we have because
there is no license included. Well on the page, it says we can
modify, but says nothing about redistribution. You are taking a legal
risk to redistribute that code without explicit licensing as it is
(necessarily) a copyrighted work and you have not been clearly granted
any rights to it. In general, it's good practice to include the
license -in- the code.
> Is it wrong or inaccurate to refer to the latter sort of program as open
> source? The source code is open, but it seems a quite different animal
> from the ever-evolving and distributed development of Mozilla, Linux, etc.
Determine what the license is for Pirate Source first. Your decision
may be made for you at that point, because these licenses often
require you to use the same license if you redistribute.
If you still have a choice under Pirate Source's license, here's a
short howto on selecting a (Free) license:
http://www.dina.kvl.dk/~abraham/rants/license.html
There are more:
http://www.google.com/search?hl=en&q=howto+choose+%22open+source%22+license&btnG=Google+Search
You should realize that you only accrue the benefits generally
associated with Open Source if you encourage community involvement,
including tools for communication, discussion of ideas, common
principles, and accepting code contributions back into the original
source.
This takes a lot of infrastructure; SourceForge.net can help.
Having these tools won't magically create a community, but -not-
having these tools can keep one from forming.
> Is there a different/better label for these library-developed
> applications, or am I splitting hairs?
Unfortunately, no. Without a license, software is not (uppercase)
Open Source (even if source is provided). There are many Free and
Open Source projects around which no (or a small) community forms.
Not everyone can expect an Apache success. The size of community does
not define whether something is Open Source. However, you can be sure
that (other things being equal) a closed system will have less
community than an open one.
I think that to the extent possible, library-developed applications
should be Open Source, because this allows collaboration at low cost
(and in the absence of economic incentives). Where a library stands
to benefit, perhaps Open Source (or commercial) is the better choice,
but consider your goals.
Sorry for verbosity; I hope it helps.
|