While I agree with ross in general about suggesting technical solutions
without suggesting how they are going to be maintained -- agree very
strongly -- and would further re-emphasize that it's improtant to
remember that ALL software installations are "living organisms"
(Ranganthan represent!), and need ongoing labor not just initial install
labor....
I don't agree with the conclusion that the _only_ way to do this is with
a "central organization" or "my organization which has shown
commitment through z"
I think it IS possible to run things sustainably with volunteer
decentralized not-formal-organization labor.
But my experience shows that it _isn't_ likely to work with ONE PERSON
volunteering. It IS more likely to work with an actual defined
collective, which feels collective responsibility for replacing
individual members when they leave and maintaining it's collective
persistence.
Is that foolproof? No. But it doens't make it foolproof to incorporate
and have a 'central organization' (still need labor, paid or unpaid), or
to have an existing organization that commits to it (can always change
their mind, or not fulfill their commitments even without actually
changing their mind). There are plusses and minuses to both.
I am a firm believer in code4lib's dentralized volunteer
community-not-organization nature. I may be becoming a minority, it
seems like everyone else wants code4lib to be Official? There are
plusses and minuses to both.
But either way, I don't think "officiality" is EITHER neccesary NOR
sufficient to ensure sustainability of tech projects (or anything else).
But i fully agree with rsinger that setting up a new tech project
_without_ thinking about ongoing sustainability is foolhardy, unless
it's just a toy you don't mind if it disappears when the originator
loses interest.
On 12/4/2012 11:08 AM, Ross Singer wrote:
Shaun, I think you missed my point.
Our Drupal (and per Tom's reply, Wordpress -- ...and I'm going to
take a stab in the dark and throw MediaWiki instance into the pile)
is, for all intents and purposes, unmaintained because we have no in
charge of maintaining it. Oregon State hosts it, but that's it.
Every year, every year, somebody proposes we ditch the diebold-o-tron
for "something else" (Drupal modules, mediawiki plugins, OCS, ... and
most recently Easy Chair), yet nobody has ever bothered to do
anything besides send an email of what we should use instead.
Because that requires work and commitment.
What I'm saying is, we don't have any central organization, and thus
we have no real sustainable way to implement locally hosted services.
The Drupal instance, the diebold-o-tron (and maybe Mediawiki) are
legacies from when several of us ran a shared server in a colocation
facility. We had skin in the game. And then our server got hacked
because Drupal was unpatched (which sucked) and we realized we
probably needed to take this a little more seriously.
The problem was, though, when we moved to OSU for our hosting, we
lost any power to do anything for ourselves and since we no longer
had to (nor could) maintain anything, all impetus to do so was lost.
To be clear, when we ran all these services on anvil, that wasn't
sustainable either! We simply don't have the the organization or
resources to effectively run this stuff by ourselves. That's why I'm
really not interested in hearing about some x we can run for y if
it's not backed up with "and my organization which has shown
commitment through z will take on the task of doing all the work on
this".
-Ross.
On Dec 4, 2012, at 10:41 AM, Shaun Ellis <sha...@princeton.edu>
wrote:
Tom, can you post the plugin to Code4Lib's github so we can have a
crack at it?
Ross, I'm not sure how many folks on this list were aware of the
Drupal upgrade troubles. Regardless, I don't think it's
constructive to put new ideas on halt until it gets done. Not
everyone's a Drupal developer, but they could contribute in other
ways.
-Shaun
On 12/4/12 10:27 AM, Tom Keays wrote:
On Tue, Dec 4, 2012 at 9:53 AM, Ross Singer
<rossfsin...@gmail.com> wrote:
Seriously, folks, if we can't even figure out how to upgrade
our Drupal instance to a version that was released this decade,
we shouldn't be discussing *new* implementations of *anything*
that we have to host ourselves.
Not being one to waste a perfectly good segue...
The Code4Lib Journal runs on WordPress. This was a decision made
by the editorial board at the time (2007) and by and large it was
a good one. Over time, one of the board members offered his
technical expertise to build a few custom plugins that would
streamline the workflow for publishing the journal. Out of the
"box", WordPress is designed to publish a string of individual
articles, but we wanted to publish issues in a more traditional
model, with all the issues published at one time and arranged in
the issue is a specific order. We could (and have done) all this
manually, but having the plugin has been a real boon for us.
The Issue Manager plugin that he wrote provided the mechanism
for: a) preventing articles from being published prematurely, b)
identifying and arranging a set of final (pending) articles into
an issue, and c) publishing that issue at the desired time.
That person is no longer on the Journal editorial board and
upkeep of the plugin has not been maintained since he left. We're
now several WordPress releases behind, mainly because we delayed
upgrading until we could test if doing so would break the
plugins. We have now tested, and it did. I won't bore you with
the details, but if we want to continue using the plugin to
manage our workflow, we need help.
Is there anybody out there with experience writing WordPress
plugins that would be willing to work with me to diagnose what
has changed in the WordPress codex that is causing the problems
and maybe help me understand how to prevent this from happening
again with future releases?
Thanks, Tom Keays / tomke...@gmail.com
-- Shaun D. Ellis Digital Library Interface Developer Firestone
Library, Princeton University voice: 609.258.1698 |
sha...@princeton.edu
