[pkg-fso-maint] fsousaged does not start

[Steffen Möller]

gregor herrmann wrote:
> On Sat, 16 Jan 2010 01:04:12 +0100, Steffen Moeller wrote:
[...]
>> To me it all seems like we need someone to organise us all a bit. Why is there
>> no Neo Dunk Tank? No reader of this list would object this, would one?
> 
> Hm, not sure if anyone would put money into the FR; but paying bright
> guys like enrico and lindi to replace fso+shr with something that
> actually works sounds like a nice idea in principle.

Hm, is anybody here claiming that for instance the N900 with Maemo truly 
works? Alone for the short battery life I presume the smartphones to all 
need another generation to give us our "charge once a week" feeling 
back. Once we get our long lasting childhood diseases out of our way, 
then we should still be surprisingly competitive. The surprises about 
Neo having a short battery life was only because it was so early on the 
marker. The latest phones are no better. But I would not necessarily 
promote the Neo only as a phone.

> So if you have a concept in mind: go ahead :)

 From my perception, the problems are less about individual programs but 
about their interplay, i.e. they are not broken "all at once" but just 
something always breaks: hal, tslib, debootstrap, zhone, fso, dbus, 
gpsd, keyboard/screenlock, install.sh .... something is too often not 
behaving as expected. Once we are at a stage that we can rely on the 
sid-testing barrier (are we already?) this should already somewhat 
improve, i.e. we would prevent one application to move forward until 
other uploads have adapted.

Once it is all more stable, the FreeRunner should be barely beatable for 
anyone interested in "phonifying" some legacy hardware. There should be 
landlords monitoring the temperatures/pressures/.... of some remote 
heating, alarm systems, OpenStreetMapping/GeoCaching/Navigating device, 
... and it is a much neater gaming machine than the old Nintendos. But 
as long as the SD does not boot every Xths time I start the phone, or 
when the machine just hangs and becomes un-switch-off-able, then this 
means some work, to someone.

I am hoping for someone to identify weak spots and to inform the 
community and the respective developers about those. That person should 
work towards stability him/herself, and possibly identify working 
constellations and prepares such snapshots as reference images. And 
there is the Debian installer work all pending, still. This should all 
increase the fun for the volunteer developers: the Neo Dunk Tank would 
have the responsibility for all the context, while the volunteer 
developers would be allowed (not forced) to concentrate on singular issues.

Who would spend the money I cannot tell. The SPI? Would the others like 
Nokia still profit from a better Debian? Neonode maybe? Vodaphone? 
Should we form a legal society to collect such earmarked money ourselves 
and have that society employ someone? Certainly Debian would profit from 
it since also fixes to packages or infrastructure relevant to other 
architectures than ARM can be expected. Would anybody know whom to ask? 
How much dedicated man power would we need to have some advantage over 
the current situation?

Many greetings

Steffen