MapServer Foundation thoughts and reactions

[Ed McNierney]

Folks -

This morning I sent a few comments about the MapServer Foundation
off-list to Steve Lime, and (at my request) he forwarded them on to some
of the other folks involved.  After a comments by a few folks there were
requests that I post my messages to the broader community.  This post is
an attempt to do that in a consolidated way.  I apologize for being
wordy, but there's a lot to say.

I've been a member of the MapServer "community" for several years now.
The Foundation project is the first time I can ever recall there being a
conscious, ongoing, and deliberate attempt to exclude most of the
community from a discussion of significance about MapServer.  A small
number of people - some of whom are dedicated developers who've
contributed far more than I ever have - decided to enter into
discussions that included two commercial firms (DM Solutions and
Autodesk).  No one else got to participate, and the work was
deliberately kept secret.  Doesn't sound like much of an "open" project
to me.

A MapServer Foundation is a very, very good idea.  This MapServer
Foundation has gotten off to a very, very bad start.  I find myself in
the position of being quite reluctant to support this instance of a
concept I eagerly wish to support.

I think I should start by explaining why I think a MapServer Foundation
is a very good idea (as opposed to what others think, even though we
generally seem to agree).  MapServer has been well-served by the
technical and development community that supports it.  It has mainly
lacked many of the things that make a "program" a "product".  It needs
better documentation, easier setup and sample sites, product summaries
and literature, feature/benefit brochures and comparisons, benchmarking
tests, presentations, a coordinated trade show/conference plan, better
marketing, directories of consultants, reference sites, etc.  I don't
mean to denigrate any of the efforts made along any of these lines, but
I think we all know there are things you can currently get from
commercial vendors that aren't available with MapServer.  A Foundation
would be a great way to provide these things.  It wouldn't need to get
in the way of the development work, and could complement it by filling
in the blanks.

All of that takes money.  A MapServer Foundation needs funding to do
these things.  Fortunately, there are several subsets of the MapServer
community that are in a position to contribute funding.  There are
commercial users of MapServer (folks like me, GlobeXplorer, etc.),
commercial developers/consultancies like DM Solutions and others, and
the government and educational users who tend to not have much money to
spend but can usually contribute something.

To date, organizations interested in financial support for MapServer
have been limited to funding specific software development tasks.  The
pace of that development has been such that every time I raise an idea
about a project TopoZone could fund, it seems that someone else has
gotten there first.  I could have chipped in money for "future
development", but there was no place to put it - it didn't make sense to
just send Frank or Daniel or Steve a check and tell them to try to spend
it somehow.  And I would rather fund the "other stuff" than fund feature
development - there's more of a need for it.  A Foundation could fix
that, by providing a place that takes in revenue from members and
sponsors, and uses that revenue to fund projects - probably
non-development projects as I mentioned above, since those won't get
funded otherwise.  The OGC membership model is a relevant and simple
example of this sort of thing.

So what does the Foundation need to do that?  It needs to be open and
inclusive, eligible to all to participate as peers or as peers within
certain classes of membership.  It needs to be independent of any
particular sponsor, and it also needs to APPEAR to be independent.  It
needs to have a clear mission and it needs to simplify and clarify
things for its members and for its constituent base.  It needs to be
seen as the unswerving voice dedicated to the support of MapServer and
nothing else.

Today's announcement missed those goals by a wide mark.  Some of those
errors can be corrected, but some we'll have to live with forever.  And
most of them could have been avoided by the kind of open, inclusive
discussion we've always had in the MapServer community - until now.

Supporting the MapServer Foundation is a great PR and marketing
opportunity.  It appears that Autodesk and DM Solutions were extremely
aware of that, and made sure that they didn't have to share that
opportunity with anyone else.  Being a "founder" is very important, and
you've already seen Autodesk and DM Solutions take advantage of that
through their own press releases today.  No other company will *ever*
get that chance - the press doesn't really care about the next few
companies to sign on.  When I created TopoZone in 1999, it was
incredibly important to be the first topographic map site on the Web,
because the PR value was so great.  I suspect very few folks remember
who launched the second one....

Companies will be attracted to sponsor the Foundation because of that PR
value.  Unfortunately, that value's gone and nothing will get it back.
I'm certainly a potential financial supporter of the Foundation, but I'm
also running a business.  I can't simply give money away, but I can
spend it on things that give me PR and marketing value.  I could spend a
pretty substantial (for me) sum as an annual commitment to the
Foundation.  I am now a *lot* less inclined to provide that support to
this Foundation, because the value (in PR and marketing terms) is a
whole lot less than it would have been if I could have been invited to
the party.  I'm certainly welcome to sign on and take a seat right up
near the front - as long as it isn't in the front row.

I don't say that because I'm personally miffed at being excluded - I'm
just TopoZone.  I say that because we'll never know how many firms and
how much financial support could have been raised if someone had tried
to solicit input and support in an open, inclusive way.  There are lots
of us out here.  I've been told that it's "incredibly important" that
the Foundation be seen as vendor-neutral and that it not be at the mercy
of a single funder's contributions.  Sounds good, but don't tell me that
now - those are both reasons to solicit a larger number of contributing
founding members rather than selling the whole package to Autodesk.
It's not easy to undo that; the Foundation is clearly already seen as an
Autodesk initiative by the press (in part because Autodesk has tried to
make that point clear) and not many firms are interested in throwing
money at Autodesk - they've got more of it than I do.

My second huge concern is the branding/product lineup for the
Foundation.  I woke up this morning to two MapServers where we had one
before.  One of them has the impressive-sounding name "MapServer
Enterprise" while the other is currently named after a large pussycat
but may or may not be open to the possibility of being named after a
different mammal.  There's no doubt in the potential customer's mind
which one is the grown-up, field-tested, production-ready, scalable,
capable system.  Unfortunately, they're thinking of the wrong one.

Branding really matters.  It's very important.  Tyler Mitchell says so,
too, on the new MapServer site.  Autodesk has zillions of people who
know that very, very well.  They just bought a great brand and MapServer
suddenly managed to take a back seat to itself, something I would have
thought anatomically impossible.  They've managed to appropriate a
well-respected brand name and take center stage with it.  Autodesk's
press release takes advantage of that ambiguity by introducing Steve
Lime as the "creator of MapServer" without saying which one they're
talking about!  Speaking of press releases, in an effort like this it is
common for all founding members to see and sign off on each other's
press releases in advance, something which appears (from some developer
comments) to not have happened here.  This is PR 101 stuff - if you
don't try to keep what you're doing a secret, you might get helpful
advice.

The same is true, by the way, about the questions raised on Autodesk's
patent policy.  This should NOT be an open question *after* the
announcement - Autodesk's patent portfolio and their defense of it are
well-known.  It should have been one of the first questions raised and
answered.  Once the Foundation's plans were made public it only took a
few hours to bring it to everyone's attention - remember the benefits of
open development?

The "MapServer Enterprise" product just got inserted into the MapServer
family by decree.  Customers know very well that when they see two
similar products side-by-side, usually due to a merger or acquisition,
they sit back and wait to see which one gets killed off.  This usually
has the effect of discouraging adoption of BOTH products, because
customers don't know which one to implement and don't want to make the
wrong choice.  Believe me, I've been a CTO standing up in front of
customers in that situation more than once - they don't believe you can
serve two masters, and they're right.

Does the Apache Foundation offer two Web servers?  Apache Enterprise and
Apache Other?

Can't kill off MapServer, you say?  Perhaps not in a technical sense,
but if there's a MapServer Foundation and a MapServer Enterprise, who's
going to notice if that other thingy doesn't get the same amount of
attention?  Perhaps the platypus is indeed a good choice, as it may
belong with the nearly-extinct monotremes.  You can't kill the MapServer
code, but you can certainly kill the brand.  Please don't confuse the
two.

Why was the Foundation "announced" when it apparently doesn't actually
exist?  It seems like today's announcement was designed primarily to
maximize the PR value to DM Solutions and Autodesk - after all, the
press got briefed about it before the rest of us did.  As far as I can
tell, there isn't any foundation, but when we get one it's going to be
great and open to all, because DM Solutions and UMN and Autodesk have
all assured each other that it will be.  Each time I hear that "now's
the time to participate", I cringe because I'm being told that by the
exclusive group who deliberately prevented all of us from participating
until they decided they had gotten what they needed out of it and it's
now OK to let the rest of us inside.  The time to participate was last
week, or last month, before anything got announced and before we were
all handed the Foundation.  If the Foundation is really a genuinely open
opportunity for us, then tell us that the inclusion of Autodesk's
product isn't non-negotiable.  Do the rest of us get to insert
MapServer-branded products whenever we want to?

All of these problems were preventable.  All it would have taken was an
open discussion of the proposal.  You get a lot of people spouting off,
and then you find out who's really interested.  You find out how many
commercial sponsors you can get and at what level of support.  You
create what appears to the public as a truly open consortium that's
worth watching, instead of one that triggers discussions about Autodesk.
You demonstrate right from the start that you have a broad base of
commercial support, with commercial firms from the USA, Canada, Europe,
South America, Australia, etc.  What was the perceived benefit of
keeping the process secret and exclusive?  Did someone threaten to pick
up their marbles and go home?  You can often be surprised at how many
folks are willing to contribute their own marbles when something like
that happens - but you never know until you ask.

The MapServer community really needs a Foundation to support it and to
keep the product healthy and growing.  There are many examples of the
creation of such consortia to draw from, both inside of and outside of
the Open Source community.  It doesn't appear those examples were
considered.  We really need a MapServer Foundation - I'm not at all sure
that we need this one.

	- Ed

Ed McNierney
President and Chief Mapmaker
TopoZone.com / Maps a la carte, Inc.
73 Princeton Street, Suite 305
North Chelmsford, MA  01863
Phone: +1 (978) 251-4242
Fax: +1 (978) 251-1396
ed at topozone.com