Crazy talk: migrate to Google Apps?

Disclaimer: I'm talking here in my role as a SLUG member, not in my role as
a committee member; I reserve the right to have a different opinion when I
put my VP hat on.

Disclaimer: I'm a Google employee, so please add the appropriate amount of
salt.

Google launched a version of Groups for Google Apps domains today, which
finally allows for mailing list management on your own domain. For a long
time, this has been the one thing making Google Apps completely infeasible
for hosting SLUG's online presence[1]. Apps now offers mailing lists (on our
own domain), web hosting (Sites for very simple things. AppEngine for
everything else), calendaring (for an event feed)...

The biggest benefit I see to doing this is the reduced maintenance. SLUG at
present runs entirely on volunteer labour; and volunteer labour isn't quite
providing us with the amount of effort needed to do basic housekeeping on
our server. As one example - we were given a new VM over three months ago to
migrate our web/mail hosting to; but because the site hasn't been kept up to
date, the effort involved in migrating is large enough that it hasn't been
done yet[2]. Migrating to Apps would be less work than migrating to a new VM
- and would then require much less maintenance to keep us going.

The biggest downside I see to this is: Google Apps != FOSS. Sure, there's
capacity to run FOSS (on AppEngine) - and even to run FOSS that would run
just as well if it weren't hosted by Google; and Google provides some
support to FOSS; and if we chose to move away, Google make it fairly easy to
move away and even take our data... but it's not FOSS.

The second-biggest downside is that I haven't (yet) been able to figure out
how arbitrary email addresses can be subscribed to the list: as far as I can
tell, only email addresses with an associated Google Account can subscribe
(the email address gets a confirmation email with a link; you have to follow
the link to confirm on a form that you want to subscribe, and you have to
log in to a Google Account in order to get to the form).[3]

To save time and space and electrons, I'm now going to run through some
other points very briefly below.

Pro-Google: We're already not using pure-FOSS - the new VM we're migrating
to runs inside VMWare.
Con: But moving to Google is still moving further away from using FOSS - and
what kind of FOSS advocacy group are we if we don't use FOSS for our own
needs?
Pro: A manpower-starved FOSS advocacy group
Con: Really? You can't even manage to look after a couple of dozen static
pages and a mailing list or two?
Pro: History says no, not really.
Con: But wait - won't this cost money?
Pro: Yes, but only (assuming we get accounts for 7 committee members at a
time) $350/year
Con: So you're suggesting we migrate from a no-money-but-needs-more-time
solution that's known to work, to an almost-no-time-but-needs-more-money
solution that has unknown problems?
Pro: Sure, why not. Surely we can find that small amount of money - or we
could even cut it down to just, say, 3 web admins instead of all 7 of the
committee...
Con: Hang on - SLUG has always tried to avoid, as much as possible,
entanglements, even entanglements with FOSS-friendly companies. Doesn't this
entangle us? Plus, what's this nonsense about requiring people to create
accounts with Google just to join our mailing list? That's not very nice!
Pro: Urrr... But Google! Google is shiny! They do no evil, they promised!
And besides, they're shiny!

What do you think? Is a migration to Google Apps crazy talk? Is it something
we should consider? What matters more to us - using FOSS as far as is
possible, even if it severly strains our resources - or being pragmatic,
paying a small amount of money to save us a lot of time?

[1] It hasn't been the only reason not to go this way: there are still
plenty of reasons why this might not be appropriate, but at least it's
*feasible* now.
[2] We'd have to upgrade from Drupal 4.7 to at least version 5.20. This then
introduces other problems, such as the fact that the Flexinode module that
was used to create Event pages is no longer supported, so we'd have to come
up with an equivalent using CCK, or something similar... It's actually not a
lot of work, a single person should be able to knock it over in 2-3 days of
work
[3] The best workaround I've come up with so far - I don't believe you need
to authenticate to get the group's RSS feed; and we can set the group to
allow mail from any address and rely on manual moderation to keep out the
spam

-- 
SLUG - Sydney Linux User Group Mailing List - http://slug.org.au/
Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html