[Community] Growing the Community

Dr. H. Nikolaus Schaller hns at goldelico.com
Tue Jun 18 13:39:01 CEST 2013


Am 18.06.2013 um 13:01 schrieb Bob Ham:

> On 2013-06-18 10:40, Dr. H. Nikolaus Schaller wrote:
>>> Am Sat, 15 Jun 2013 09:39:11 +0000
>>> schrieb Bob Ham <rah at settrans.net>:
>>> 
>>>> Here is the acid test I think any phone needs to pass in order to be
>>>> successful: you must be able to take the phone out of the box, stick a
>>>> SIM card in, charge the phone, start using it fully, stick it in your
>>>> pocket, and never, for the entirety of the life of the hardware,
>>>> update any software on it to a newer version.
>> 
>> I think you can do that it with any phone. You do not need to upgrade it.
>> You only have to accept that communication protocols change and you
>> will not be able to use all services over that long time. Some of them
>> will disappear or start to be broken. You may also not (yet) have the
>> time zone database for the next 10 years etc. It may also happen that
>> the network where your SIM card is for will not be available any more in
>> 10 years...
> 
> The issue is not whether the technology used by the phone remains relevant in the future.  The issue is whether the software distributed on the phone works.

If it is distributed on SD card you can even choose between an old image
and a new image without damaging anything. So IMHO it is not even a
contradiction...

This is why the http://download.goldelico.com/gta04/ has been organized as 
an archive. The plan is to never delete images so that you can always restore
to the one that was delivered.

But if you expect software in a quality that it is bug free and nobody
finds a significant bug within 10 years, it needs a completely different
development process compared to communites, open source and git...

The latter are based on the assumption "if we introduce a regression, you
can fix it in the next release in 4 weeks or fork your own distro".

We would need development processes like in aviation industries. And,
even on deep space missions they can field upgrade the software (e.g.
Voyager communication technology was improved in data rate over
the years by software upgrades).

So I think it is a nice dream, but do we need such a high goal to give a
significant contribution to freedom of mankind? Or do we need to be
better than all the other devices out there? And can we achieve this
at all?

But I admit that we are far away from the best approach we *can* do.

E.g. QtMoko is seen by the owners as the most usable distribution we
have. But as others already said, it is not complete and sometimes still
has rough edges. Replicant is still fighting a little with power management.
Etc.

But we don't even use a bug/issue tracker for these projects. Or I am
not aware of. And we don' point to them.

For the neil-plus-kernel [1] we have it, but it is also not intensively used.
I.e. we lack some more formal organization how multiple developers work
on issues that others have. Currently we are in a "I make it for myself"
mode :)

BR,
Nikolaus

[1]: http://projects.goldelico.com/p/gta04-kernel/issues/




More information about the Community mailing list