FreeCalypso website update

Mychaela Falconia falcon at ivan.Harhan.ORG
Mon Nov 23 08:01:32 CET 2015


Hi DS,

> Although I don't want to make a promise that I might not keep,
> I'll try to look into the L1 issue. It's obvisously the major
> obstacle to a fully working gsm-fw. As you said L1 is a very
> complex piece of work, if we could obtain the source that would
> be a major advancement!

Here are the two places where I feel we should look:

1. The Chinese world, i.e., the world of phone manufacturers.  We
   would need someone who is fluent in the Chinese cellphone manuf
   culture and who would be willing to do the grunt work of trawling
   through forum sites, posting solicitations requests on the right
   sites in Chinese, talking to people in the industry etc.

2. Former TI employees, particularly in France.  If my understanding
   is correct, all those people who worked on GSM chipset stuff at TI
   France got laid off and were out of a job when that business unit
   closed.  I've heard that unemployment in France is pretty bad, thus
   some of those people may be seriously pissed off - possibly pissed
   off enough to be willing to share some sources they may have kept.

One gotcha that would need to be kept in mind with route 2 (former TI
employees) is the general industry principle that the stuff engineers
work on at any given moment is always one or two generations ahead of
what is out in the field.  Anyone who worked at TI on their wireless
chipsets at the time of their closing undoubtedly must have viewed the
Calypso as very old news, and all their work would have been on some
newer chipset.  An L1 source for some chipset newer than Calypso would
not do us much help: we already have L1 for LoCosto in full source
form, but it only pays lip service to backward compatibility with the
older Calypso, and as we learned the hard way, this lip service level
of backward compat isn't good enough for usability.

In other words, what we are looking for is not just any random TI L1
source, but specifically one targeting the Calypso and not some other
chipset.  (As an aside, an L1 source for Calypso+ would be nice, but
only if it's accompanied by the rest of the fw suite for Calypso+ in
full source form as well.  If we got such an article, we could build
an E-Sample board from the PADS file we have for it and use it as a
basis for building our own Calypso+ phone.  In contrast, for the
regular Calypso a good L1 source is all we are missing; we already
have the rest.)

The Chinese phone manuf route may be more promising in that those
folks weren't given some vaporware R&D stuff, instead they were given
turnkey-working firmware deliverables for the chips actually shipped
in the field to make cellphones out of, and for a long time the "corn
staple" chipset was the very same Calypso/Iota/Rita we are working
with.  It is certain that there were a ton of Chinese phone manufs who
had Calypso firmware at one point, and it also seems likely that many
(if not most) of them had it in mostly source form - the particular
level of distrust which TI had for that one specific GSM/GPRS modem
manuf from whom we got our copy of TCS211 seems to have been more the
exception than the norm.  As a case in point, it appears that both
Compal (Mot C1xx) and Foxconn (Pirelli DP-L10) had their fw deliverables
from TI in full or almost-full source form.  Now if we could only find
those Chinese folks who probably still have a copy buried somewhere...

> Maybe TI would be willing to provide this
> source, considering they have exited the business?

Approaching TI is something I've been meaning to do for a while now,
actually.  I am definitely not holding my breath for any help from
them, but it seems like contacting them in good faith ought to be an
act of due diligence.  If nothing else, a lot of smart alecs keep
telling us "why don't you ask TI", and it would be nice to be able to
truthfully retort with "we have tried".

The biggest difficulty with asking TI will be figuring out *whom* to
ask, whom to approach.  It appears that the office most relevant to
this matter (the one in France) has been closed.  BTW, the link in my
previous post was broken; here is the (hopefully) working one:

http://www.riviera-buzz.com/features/news/item/94-texas-instruments-in-villeneuve-loubet-to-close-with-major-job-losses.html

I doubt that approaching TI's headquarters in Dallas, Texas will do us
any good: to the best of my understanding, they never had any
involvement with the GSM chipset business, and they'll just roundfile
our inquiries without even trying to understand what we are asking
about.  In order to have a fleeting chance of success with approaching
TI, we need to approach someone who remembers that they were once in
the GSM baseband chipset business and who would know at least at a
general level what the heck we are talking about.

> I don't know
> anyone at TI myself, but this is an avenue to pursue.

One of my coworkers at my current day job used to work for TI at their
San Diego office; she worked on 3G stuff when they closed the shop.  I
have not yet figured out whether TI's San Diego office was closed
completely or if some small remnant of it (working on something else
entirely unrelated to baseband chipsets) is still around; I am about
to investigate in the next few days, as I have their address and it's
in my local area.

I will also ask my coworker if she remembers the name of the manager
she worked for back at TI; if that person is still with TI and is
contactable, s/he may be a good place to start, as that group was
building a 2G/3G workable system by combining LoCosto for GSM with
some other company's chip for 3G.  (The two protocol stacks ran on two
separate processors, and handover from 3G to 2G during a call involved
transferring the call state between processors - oh fun!)

M~


More information about the Community mailing list