[Desktop_architects] Desktop Meeting in Japan, Jun. 1

[John Cherry]

> 
> > So far we have focused heavily on the aches and pains of ISVs. Certainly
> > this is an importand issue. However from conversations that I have with
> > people out in the field I hear about issues that have not really been
> > mentioned here.
> 
> can you share these with us?

Yes, I think we would all like to hear what these other issues might be.
For reference, the desktop architects chose to focus on...

- cross-desktop application development (i.e. Portland project)
 
- cross-distro application development (i.e. LSB and stable binary 
  and programming interfaces)

- driver availability (i.e printer drivers, wireless).  These are also
  the things that fall into the category of things that must "just
  work".

- power management, from the driver model in the kernel to the policy
  management in user space to power control interfaces in hardware
  (IC design and board design).

- audio and multimedia; discussions about standardizing media
  interfaces from an application perspective, but much debate over
  whether standard interfaces are needed/desired; can the various
  audio/mm communities come together on an approach for standardized
  media interfaces? 

- standard packaging/installation; requirements discussed with LSB

- ISV documentation portal; the concept was introduced on this list
  http://lists.osdl.org/pipermail/desktop_architects/2006-January/000704.html
  and a prototype was started...
  http://developer.osdl.org/bryce/DesktopArchitects/wiki/index.php/Main_Page 
  but little has been done to move this forward lately.  Perhaps
  it is a good time to revisit the documentation portal concept.

Other issues that have been discussed with OSDL member companies that
have not risen above the clouds with the desktop architects include...

- Enterprises will not deploy until an integrated mail/calendar client
  is available that (1) functions in a hetrogenious client environement,
  (2) has an easy, and possibly automated, way to migrate from existing
  mail/calendar clients, (3) integrates easily with mobile devices, and
  (4) will be well supported by the disto vendor.  One OSDL member that
  I recently spoke to would earnestly like to deploy with Evolution, but
  can't seem to get reasonable distro support and the development team
  has almost gone away.  Of course I recommended that this company get
  involved in development/support of Evolution if this is important to
  their deployment, but this is just an indicator of the gaps that
  are encountered in enterprise deployments.

- Enterprise desktop management (configuration management, software
  installations, software upgrades, integrated defect tracking, remote
  management, etc.).  Products exist for most/all of these
  capabilities, but those considering enterprise deployments need to
  be educated/convinced.

- Ongoing dilemnas: proprietary CODECs, driver delays due to non-open
  interfaces, DRM, web site dependencies on IE, etc...

Are we missing any other obvious inhibitors?  Sometimes it is the
elephant in the room that is the most difficult to see.

- John