Re: planner database access to non-Linux people via Ruby on Rails

[Kurt Maute <Kurt Maute us>]

On Mon, 2008-02-18 at 15:01 -0800, Chris Dawson wrote:
> I stopped by #planner earlier to ask some questions and was directed
> to this list.  I wrote a simple Ruby on Rails application that builds
> a graphical view of the planner database.  It is lacking in almost any
> useful feature aside from this.  Kurt, apparently you've been working
> on something similar?  If you think this is useful, or want
> assistance, let me know.
> 
> http://webiphany.com/Planner/planner.html

Hi Chris - this looks real cool.
My efforts have also produced rather modest results so far, but I think
there's a great deal of potential in developing a web tool for Planner
in RoR.

> The interesting thing is that you can drag items; I want to add an
> AJAX call the will update the database when a task is dragged, and
> perhaps allow someone to double-click on a task and edit the
> properties.  The biggest benefit is that this allows anyone (meaning
> people not running Linux) the ability to view the planner database.  
> 
> The code I wrote is here.
> 
> http://webiphany.com/Planner/planner.tgz

I was starting my approach to a web tool from a different perspective.

First, I looked at RoR and noticed how it very nicely handles many
different databases with very little effort to configure, and the whole
migrations thing makes upgrades easy to handle.  Maybe it would be cool
to see if RoR could manage the database updates for Planner.

So far I'm experimenting to see if I can get RoR to import a Planner xml
file.  At this point I know I can.  I created migrations for a Project
table and a tasks table, and have a page where you can choose a file and
launch an import.

I'm successfully importing the Project level info and the task info from
the xml file (including the parent_id's for the tasks).  Its really not
all that tough, it just takes some time to grind thru a few iterations
to get it right.

One of the main challenges I see is with keeping the database info in
synch with local xml file data.  If we're going to allow folks to update
the database via the web, we need some way of merging changes the next
time the PM wants to synch with the database.  I think one major step in
this direction would be to create a unique id for each task, so each
could be correctly identified (since there's no current field in the xml
file that can't be changed for each task).

So the first step I was looking at was to achieve a complete xml import
function.  A close second is a nicely formatted read only display.

After that, I'd start working on the ability to make modifications via
web and merge with the local xml file...  starting first with simple
stuff that project participants would need to update - like %complete.


-- 
Kurt Maute <Kurt Maute us>