[Marketing] quick assessment of the Service Provider Directory

[jo at frot.org]

dear all,

I had a quick look at the Service Provider Directory's source 
the other weekend, and tweaked it to show per-project links 
on the front at http://www.osgeo.org/search_profile ,
here are some notes i dug out about the experience.

Issues that have been discussed are:
- making the SPD show up in search results better
  ("support for mapserver" should show it in the top 10.)
  - more projects should link to the SPD
  - projects don't link to their "own" pages because the URLs
    are longwinded and unmemorable 
    (e.g.
    http://www.osgeo.org/search_profile?SET=1&MUL_TECH%5B%5D=00013 )
  - it would be *much* nicer if we could have 
    http://www.osgeo.org/spd/project/openlayers
    or at least 
    http://www.osgeo.org/spd?project=openlayers
  - getting the SPD widely linked is key to getting it used.

- prioritising according to code contributions
  - http://wiki.osgeo.org/index.php/SPD_Prioritization
  - Either PSC chairs responsible for updating listings
  - OR easier, they provide a list to SPD maintainer(s)
  - companies that pay committers appear at top of list

- php backend 
  - the backend doesn't make much use of what drupal provides 
    in terms of building blocks. 
  - it has possible SQL injection issues which should be fixed
  - there has been talk of throwing it away and rebuilding
  - but time and money has gone into this already and there's 
    not *much* extension by way of functionality to add
  - main 'missing' is admin-like edit interface but a dedicated
    maintainer / tweaking permissions would address this 

- UI
  - it's not so pretty, is it
  - this should be considered before any backend changes 
  - would a shiny map UI really help find data? shrug.
  - the 3-select-box combinatorial search is, well, idiosyncratic
  - sketches on a postcard appreciated

That's about all i can recall having to say. I would be happy enough
to spend another hour or so hacking the php around if minor changes
were deemed "enough". Without UI/interaction suggestions i don't think
there's much point going down the "reimplement to use more of drupal" 
path. It would be easy to get carried away with profile pages for
service providers, links to blogs, job feed, etc but then i think this
would become a quite different app requiring a big thing and maybe not
belonging on the main osgeo site.

cheers,


jo

--