Back to the basics: improving user experience
stroller at stellar.eclipse.co.uk
Sun Oct 19 15:21:17 CEST 2008
On 19 Oct 2008, at 13:46, Dale Maggee wrote:
> arne anka wrote:
>>> ==Pim device==
>> imho that's exactly the kind of task openmoko did _not_ ask for.
> I would respectfully disagree - Openmoko asked about "Improving user
> experience", and users are saying they want to experience PIM
>> and it's doable by community!
> Agreed, it could be done by the community, but I don't see anyone
> it, and I'm not smart enough, nor do I have the time at the moment.
It does not matter whether YOU could do it or not. it's a matter of
where Openmoko's resources are best spent.
If you can't write a PIM app, then you CERTAINLY can't write kernel
drivers - THAT is where Openmoko's resources should be mist focussed,
IMO. As others have stated, there is some hardware-level stuff that
only Openmoko has NDA for. And without working hardware drivers to
ensure that phonecalls work flawlessly (and wifi, and bluetooth), a
PIM is irrelevant.
>> there are a lot of posts lately completely ignoring the point of
>> and "no eyecandy"
> I haven't seen anybody ask for pretty-looking PIM applications, people
> seem to be asking for *reliable* PIM applications. I'd call
> and robustness "basic".
Basic reliability and robustness resides in a program with which you
can enter a number and make a call. Once that prototype exists it is
much easier for the community to extend it to PIM functionality.
Openmoko can then move on to wifi drivers, Glamo hardware acceleration
and pairing of bluetooth headsets.
>> pim frinst is at it's best part of a middle tier, but rather of a
>> particular distribution --
> This kind of comes into the "Should FSO merge be sped up?" debate,
> as I
> believe the framework has PIM stuff built into it.
AIUI (and I would be delighted to be corrected if I'm wrong), the FSO
stuff is intended to provide functions which will allow you to make
simple DBUS calls such as "get number $var from PIM manager" and
"make call to number $var". Once these are complete, writing your own
applications becomes easy. True the first of these example calls
requests you integrate the functionality in your own app, but the
latter makes problems with dealing with the dialler & the GSM chips &
whatever go away. It is FAR more important to provide the community
with these tools than it is to provide any kind of application that
utilises them (beyond a command-line version which gets numbers from a
text-file and operates as a test example). Once these calls are
available there will be dozens of PIM managers posted to this list and
being written by enthusiastic Python programmers.
More information about the community