FreeCalypso still alive in 2022

[Mychaela Falconia]

Hi DS!

> About nanoBTS, be aware that they tend to transmit at fairly high power,

It is my understanding that the maximum power these units are capable
of putting out is 23 dBm.  I live in a two-story building that has
been divided into 4 apartments, specifically in one of the two upstairs
apartments, and I plan on installing the nanoBTS indoors, using an
interior wall as mounting support, without anything protruding outside.
In this arrangement the path of signal from my pirate BTS to the
outside world will have to cross (penetrate with loss) at least one
building wall, or more than one wall (one exterior and some interior)
depending on direction.

I expect that under these conditions, the signal from a BTS that puts
out 23 dBm will probably fade to undetectable (or at least to the
point where a standard phone can't pick it up) within a few meters or
maybe tens of meters outside of my house - but I will be happy to be
proven otherwise.  For the purpose of this test network, I will be
operating in the PCS band, not 850 - otherwise Pirelli phones and
other legacy devices that don't support GSM850 band won't be able to
connect to it.

Now if everyone who got those poison injections end up dying from
blood clots and autoimmune diseases and whatnot over the next few
years, and the remaining people of my kind (purebloods) get to
repopulate the world and rebuild civilization, then I will start
building and deploying a real *production* GSM/2G network (not just
for handset development and testing), and at that point I would
definitely go for GSM850 band, to get maximum geographic range from
each single cell - more bang for the buck - but in that phase we will
no longer be able to use Pirelli handsets, only C139 or our own FC
hardware - or Nokia C3-00, the model which my S.O. likes the best, and
which is thankfully quadband.  But the test network that needs to work
with legacy handsets will need to be a separate, short-range (not
extending much beyond my lab) PCS1900 cell.

> Hence, you may want to use a test PLMN,

By "test PLMN", do you mean setting MCC-MNC to 001-xx or 001-xxx, as
opposed to 310-xxx?  Right now my plan is to use 310-xxx, where xxx is
a code that does not officially belong to anyone: in almost all
official assignments the last digit of MNC is 0, hence most codes with
non-zero last digit don't officially belong to anyone.  But what would
be the benefit of using MCC of 001 instead of 310?  Either way I will
be illegally squatting on an unused sliver of spectrum that is
currently going to waste...  (Yes, I did run my SA across both GSM850
and PCS1900 bands and found such unused slivers, plenty enough for
GSM.)

One complication we have is that our current FCSIM1 cards (GrcardSIM2)
have EF.AD allocated as 3 bytes rather than 4, hence we lack a truly
proper way to tell phones that we use a 3-digit MNC.  Believe it or
not, there is code in TI's firmware (originally there, NOT added by
me) that special-cases MCC 310 (and other USA MCCs up to 316) and
takes the first 6 digits of IMSI (as opposed to the first 5 digits) as
the home PLMN in the absence of 4-byte EF.AD - and I can only hope
that similar special casing for MCC 310 may be present in other
mainstream legacy phone firmwares, such as Nokia C3-00.  But if we
program IMSI to something that begins with 001 and we don't have a
4-byte EF.AD, then there is absolutely nothing on the SIM to tell
phones that we use a 3-digit MNC, and phones would take the first 5
digits of IMSI (not 6 as we want) as the home PLMN.

Why have we got such poor SIM cards that don't have 4-byte EF.AD?
Answer: cost.  The batch of 200 cards we got made for us last year
cost a total of $582.75 after shipping and other overhead, whereas the
MOQ from Sysmocom would cost somewhere around $3000 for unprinted
white cards or $4000 with printing.  But Grcard refuse to customize
their file system layout (or heaven forbid, disclose the knowledge of
how to do it downstream) for orders smaller than 10000 cards, which
would of course be even more cost-prohibitive than Sysmocom's MOQ of
1000 pcs.

M~