I'll reserve any comment on being "purchased" for when I have
something to sell. ;)
Here's what we currently have going on.
# The Software
The bulk of the content is obviously in Confluence. It's a very
powerful platform for editing and composing digital documents. It was
formerly running off mysql and my server was updated to use mariadb.
Confluence doesn't official support mariadb and I couldn't update for
a long time.
I recently migrated the database to postgresql and that seems to be
working fine. Confluence has native support for it - although they
will only be supporting > 9.2 in the next release. My server is more
than capable of handling the load (6 cores, 8GB ram < 30% CPU
utilization average), although keeping up with security updates (there
was just one a few days ago) is time/effort I have to expend. Overall,
it's not that much time to keep it running. My box does not run Apache
and instead is using nginx. Of course, Confluence is java based and
runs under its own jre.
# The GoogleGroup
This was setup for the purpose of notifications from the site or
github repo. It wasn't really meant for discourse. Although, because
it does use email, people will reply to stuff - just like I am now. :)
The Google Group is pretty much hidden unless those who read the site
seek it out. The stuff in the group could be integrated into the site
via frames though - I would assume.
# Disqus
I was only recently told about this by Micah Woods and it looks like
it's a good solution to the 10 user limit on Confluence. I had
contacted Atlassian about an open source license, but they pointed me
to their requirements doc and we can't meet the specs. I could try
again, but I wouldn't hold my breath.
Disqus seems like a nice option. It allows individual control of your
comment notifications, but pretty much everything you do through it is
completely public. You can click on a users's profile and see all
their comments through the Disqus system. Plus, the Confluence plug-in
is $50. Plus, the content is then under the control of a third party.
# Moving forward
I'm guessing that getting content out of Confluence and into something
else will be pretty painful (read time-consuming). Not impossible, but
will require more effort than I personally have. If the integrity of
what we have can be preserved, then I can simply offer up the db if
someone wants to parse - or knows of a converter. You'll need to be
familiar/comfortable with postgres to pull the data out - unless you
simply site scrapped
filemakerstandards.org with something like wget.
Maintaining cross-links would be a bit of a pain since internally
confluence does not use absolute urls.
Hosting is also an issue. My server is pretty much locked down tight
within ssh key based access only through a non root account.
Confluence is sitting behind a proxy because my other services are
using standard ports. If someone else does want to host and it seems
reasonable, then I can certainly help admin. I could also setup a host
somewhere and we move the site off my business box. This would require
someone else with sysadmin skills. Does anyone else want to volunteer?
Have good shell skills? It would also be an additional cost. Is this
something the community will bear?
We could easily say all comments should now go through Disqus and that
seems reasonable to me - aside from the downsides of not being under
our control. In fact, the comments weren't flowing at all until I
installed the add-on. Now we're having this discussion. :)
It seems like this is a bit more of a 'policy' issue than one of
technical detriment. Confluence is good at what it does. We just need
to outline how we want to handle things. The feedback I've heard from
people using the standards say they go to the site and print off pdfs
and provide them to their developers. I've heard this from a few
companies.
I've also been told that they like the fact that there are a few
'moderators' who curate the content.
Overall, I'm flexible in that it would be good to share the
responsibility/time/costs/efforts with others. The big question is
does the time/effort/requirements warrant the investment?