Summary

The SL6 desktop was developed and is now in use in the School. The primary os has switched to 64 bit which led to some problems with packages which were only available as 32 bit binaries (e.g. skype acroread) and which required a bit of effort to get working.

We are still officially only supporting firefox 3.x as shipped by Redhat with our config changes. We are providing up to date chrome/firefox quick release browers on a "best effort" basis. Currently the firefox rpms are rebuilt with our config changes but I suspect we will shortly be switching to repackaging the static tar releases because of dependancy issues with gnome/xulrunner.

We had an unexpected surge of software requests following shipment. These were mainly from people moving from sl5 to sl6 and tended to be rpms that shipped with sl5 so we presumably missed a number of applications whilst upgrading to F13 most were straightforward "add package to packagelist" but there were a few which were more complicated.

Because the desktop project was split from the RAT rpms part more time was available, and spent, on the user interface. e.g. bashedefenv got it's first clearout in years and we're shipping a full set of AV codecs

I had intended to ship the gconf component and preconfigure pidgin for our jabber server but ran out of time. These will get done in Operational/Personal Development time.

Observations

  • We should switch to all 64 bit applications ASAP and avoid new 32 bit only apps wherever possible.
  • We need to consider open (source) alternatives to skype.
  • There are increasingly persistent calls for Dropbox to be installed, though they tend to give up when pointed at my blog article.
  • We could control a number of useful things via gconf for those using gnome
  • We could probably provide a slicker "corporate" experience for the user: preconfiguring apps for use with our services more, customising some standard apps etc.
  • Not porting apps and waiting for the users to complain is more cost effective than autoporting...but probably not so good PR wise.
  • The best effort shipping of firefox browsers will have to switch to repackaging static binaries fairly soon

Recommendations

  • We should implement something like debians popularity-contest to try and predict which apps are actually used and drop things that are not.
  • We need to find something like dropbox that is both technically secure and legally complient.

Effort

  • The project took 9 weeks of effort. The estimate was 6, the overrun was partly because too much time was spent looking at better ways to manage firefox and providing an up to date "short release" version and partly because there was an unexpected surge of requests for missing applications.

-- IainRae - 31 Oct 2011

Topic revision: r1 - 31 Oct 2011 - 13:53:51 - IainRae
 
This site is powered by the TWiki collaboration platformCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback
This Wiki uses Cookies