Complexity of emergency-only baseband firmware development

[Mychaela Falconia]

Hi Denver,

> It would be great if I could find some existing hardware instead of building
> my own.  The Neo FreeRunner seems like a good candidate given what you said
> later (more on that below).
> [...]
> I realize that building a phone is a substantial undertaking so I'd like to
> avoid that if possible.

Yes, building your own phone hw of any kind is very very hard, not
only in the actual manufacturing phase, but also in creating the
design to be manufactured.  The process can be somewhat simplified by
skipping the design and making a 100% verbatim clone of a hw product
that was made by someone else in the past - see below for my thoughts
on doing just this with Openmoko's Neo.

> I do have a Neo FreeRunner so perhaps that should be my first goal: to run
> FC Magnetite and see how well it works for me.

Our community contributor David Matthews can probably help you with
Freerunner issues, including firmware flashing, but I can absolutely
assure you that you WILL NOT be able to see any observable differences
between Openmoko's official firmware and FreeCalypso Magnetite: if
someone were to give you a used Freerunner without telling you what
firmware version it runs, you won't be able to determine it
behaviorally, only by looking at the ID strings returned by the modem.

> > Our Magnetite firmware has not been fully deblobbed yet, but:
> >
> > * We have a clear roadmap toward a fully deblobbed version;
>
> Does it seem like a substantial amount of work?
> Or is there not much left now?

The latter - not much work left, but I need newly made FreeCalypso
hardware with an active manufacturing and marketing company behind it
in order to make the work meaningful.

> Would it not still offer wifi support?  If so, then it seems to be a good
> candidate for my use case.

If you understand WiFi and know how to use it, then yes, for your use
case the Freerunner does offer a practical advantage over a dumbphone.
My analysis of the FR having no advantage over a dumbphone is from the
perspective of someone who has no use for WiFi on a phone.

> And I would be unlikely to find a dumbphone that would do wifi anyway.

Actually the Pirelli DP-L10 (which I currently use as my everyday
personal phone with the WiFi disabled) does just this, but its WiFi
VoIP hardware is completely undocumented, hence it's a no-go for FOSS.

> Of course, as you said, the community is mostly gone at this point.  Bringing
> that back would be a challenge, but if I was able to work on the FreeRunner
> software full-time for a while, perhaps I could make that happen.

Yes, the effectively-dissolved status of the community is one of the
two killer problems with the Freerunner.  The other killer problem is
the no-longer-made status of the hardware, reduced to an extremely
scarce set of second-hand units.  The two problems feed each other - I
don't see how the community can be rebuilt without a new company that
actively manufactures *and markets* new Freerunner units that are
verbatim-identical to Openmoko-made ones, but it is a two-way street:
anyone who steps up into the role of the new Freerunner manufacturer
will need the help of an active software developer and user community
in order to succeed.

Now here is something I need to emphasize: restarting new production
of GTA02 (Neo Freerunner) units verbatim-identical to Openmoko-made
ones *is possible*, and if someone were to fund it, *I can do it*.
The complete design files for the GTA02 are published, and all of the
essential components that would be needed for a Freerunner clone
including the original LCD+touchscreen module can still be obtained in
volume.  However, such a venture would involve a quite substantial
cost:

* If I were to do it, I would need about 10 kUSD upfront to start
  phase 1: to start acquiring all of the components needed for making
  a new GTA02 motherboard, and once all of these components have been
  acquired, produce the first experimental batch of these new GTA02
  motherboards in exactly the same manner as what we are currently
  doing for the FCDEV3B.

* The WiFi add-on, the Bluetooth add-on and the plastic case will be
  separate steps with separate unknown costs after we have recreated
  the base motherboard.

* My guesstimate for the total cost when it's all said and done is
  somewhere between 50 and 100 kUSD.

Now if anyone would like to actually pursue what I just outlined,
let's wait another couple of weeks first to see how our FCDEV3B pans
out.  Everything we are doing on the FCDEV3B will also be required for
a Freerunner clone or for a SIM900-like packaged modem module or for
any other kind of end product we can come up with, and since the
FCDEV3B is now only 2-3 weeks away from the moment of truth, I believe
that we should prove it working first before we start thinking about
what we should do next.

To the funding people who are reading this list: once we get our
current FCDEV3B built and working, we shall come to a crossroads, and
we'll have a choice as to what kind of potential end product we should
pursue next.  I have at least 3 possible choices in mind already:

Option 1: build a new GTA02 motherboard as outlined above.

Option 2: build a packaged modem module that would compete with SIM900
etc, but with full recompilable source for the official shipping
firmware, no restrictions against users loading their own fw, and
published schematics for the guts of the module for transparency.
This packaged modem module can then be used by other people seeking to
build FOSS smartphones or any other applications.

Option 3: build a Handset Motherboard Prototype (HSMBP), a board just
like the motherboard of a dumbphone handset, but in a prototype form
factor that allows everything to be fully exercised on a bare board
without plastics.  This HSMBP will primarily allow us to play with and
polish TI's dumbphone UI for the 176x220 pixel color LCD configuration,
and may later be turned into a complete FreeCalypso libre dumbphone
product.

Which of these choices will we end up pursuing after the FCDEV3B?
Answer: it won't be up to me, but rather up to the funding people to
decide.  So if you are one of the latter, you may want to start
thinking about which project direction you would like to fund going
forward.

Sincerely,
Mychaela Falconia,
The Mother of FreeCalypso