[OSGeo-Discuss] "Free"

[Tim Bowden]

On Tue, 2007-03-06 at 00:43 -0700, Zachary L. Stauber wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> ross s wrote:
> > Just to add a bit more spice to the discussion.  I think the root
> > problem here is a definition amoung open source purists.  Jeff Thurston
> > has added some interesting points to his blog (below).
> > 
> 
> I think the root of the problem is that there is such a thing as an "open source
> purist" at all.  Why does "open source" have to mean "closed mind?"  On the
> other end, there is an equally troublesome "Microsoft Evangelist" (this is a
> real title on a business card, by the way), but they aren't the problem at hand.

Open source doesn't mean closed mind at all.  If you are referring to
the habit of FOSS purists being pedantic about what is FOSS and what is
not, then I think that's a good thing.  It's not uncommon for people
outside the FOSS development world to think that free as in beer is the
same as free as in speach.  For those of us who have experienced first
hand what the difference is, I think it's important to keep hammering
away at the perception they are the same. `The problem of course, is the
dual meaning of the word free, as has often been pointed out in the
past.

Don't get me wrong, I'm not proposing a "Stallman" approach here.  If
someone wants to use proprietary software, fine.  If someone wants to
offer proprietary software (free as in beer or otherwise), again no
problem.  You just won't get me jumping for joy about it.  It's not
about the quality of the software.  There is plenty of great software
and junk software in both the FOSS and proprietary development
ecosystems.  What I don't like, is handing the keys to my business to
the vendor of my mission critical software, which is what so often
happens when proprietary software is at the heart of a business.  If
being in that position makes me an "open source purist" then so be it,
but I don't see how my insistence that *I* use only open source can be
the heart of the problem for *someone else*.

>  Limiting oneself out of team spirit is senseless.  Obviously we don't want
> something that WAS free or open source to be commercialized, but that is an
> extremely uncommon, and we don't want any commercial software taking up room at
> a FOSS4G conference if it isn't (or couldn't be) related to open source in some
> demonstrable way, but I trust nobody is going to allow that to happen.

Yes, we do want open source commercialised.  I make my living of selling
open source solutions.  If that's not commercial open source, then I
don't know what is.  Do we want *proprietary* software built on open
source?  Well that's entirely up to the prerogative of the original
developer, and it *is* quite common.  Anything with an MIT/BSD style
license can be used this way.  Do the improvements get passed back down
the line?  Sometimes, sometimes not.  That's a value decision that's up
to the developers to make.

> 
> Outside of a university setting, one HAS to integrate with commercial software.
>  Most businesses in the US are Windows-only.  My employer used to mandate IIS,
> SQL Server, and Oracle Spatial exclusively.  To even try to change it (and there
> are plenty of us on the inside that wanted to), the open source community had to
> meet us halfway and make things like MS4W and windows installs of
> PostgreSQL/PostGIS.  Now a couple years later, parts of it have switched from
> Oracle Spatial to PostgreSQL/PostGIS.  Why?  Because PostgreSQL/PostGIS actually
> has better Windows support (Oracle 10.1.0 wouldn't even install under Windows
> without massive errors and security issues).  More important, you can write
> plugins for PostgrSQL/PostGIS with MS Visual Studio, whereas Oracle has
> intentionally decided to stay away from Microsoft programming languages, leaving
> it with just Java (which isn't even an open standard, unlike C#).  We can't
> change from IIS to Apache, so if MS4W didn't exist, we'd be using ArcIMS instead
> of MapServer.  If the PostGIS people and Jeff McKenna had written off Windows
> users because we happened to not think Microsoft and ESRI were the devil, we'd
> have to write off open source, but luckily they are an open-minded bunch, and I
> think that kind of open mentality deserves to be rewarded and further promoted.
> 
> 	-Zack

It is often necessary to integrate with *proprietary* software in many
different ways; Platforms, data formats, server apps, client apps etc.
The question is often asked, should open source support windows.  The
answers are as varied as the developers who write open source.

At one end, we have the ESR type approach- "No, we don't support Windows
— get a better operating system" [1] which does have merit.  If I'm
developing software to scratch a personal itch [2], why should *I* worry
about *your* operating system.  You've got the source code.  If you
really want it, go for it, so long as you respect the licence I choose. 

At the other end, we have the mozilla approach.  Firefox is targeted
first and foremost at windows, because it's the dominant platform.  The
problems it causes linux distributions are well known and only recently
addressed [3].  Which is the correct approach?  Probably no single
approach.  It all depends on the motivations of the developers, and
that's a wide open field.  I don't think you can say any one project
should or shouldn't support windows or any other relevant application.

In the end the decision by business to use open source or proprietary is
a value decision.  How well do the different options solve the business
problem at hand.  That's not a clear cut issue either.  If a business is
windows centric, then the value proposition of using open source for one
application may be very different for another business that is used to
open source.  The cost will be higher if there are no open source skills
in house; think about installation (./configure, make, make install is a
shock for windows admins), legal issues (licensing is a big one and
takes a while to get your head around if you haven't seen an OSI license
before), bug reporting, support etc.  It's all a bit different from what
many are used to.  If you want to push open source in a proprietary
centric environment, you have to show the value proposition in respect
to the business problem.  Yes, having open source apps running on
windows may make it easier in some cases and it is an issue that gets a
regular run in open source circles, but it is not always the correct
answer.

Regards,
Tim Bowden

[1] http://gpsd.berlios.de 
[2] http://gpsd.berlios.de/history.html
[3] http://www.linux.com/article.pl?sid=06/12/08/1655211