Michaela, Thank you for detailing this. I think we need to bookmark this and refer everybody who needs to understand what it is all about.
On Sep 10, 2017 1:33 PM, "Mychaela Falconia" <mychaela.falco...@gmail.com> wrote: > One other problem I have with what this "viva astoria" person is saying > is the following: > > > Then I have the following impression and question: it seems to me that > the > > newly developed firmware is neither [one quality] nor [another quality] > > than the original TI firmware? > > The fundamental fallacy in this person's choice of wording is that > they seem to think that one could somehow choose between running > either "the original TI firmware" or FreeCalypso on the same hardware, > and questioning how the latter is better than the former. But the > fundamental flaw in this person's understanding of the situation is > that there is NO "original TI fw" which one could actually run on any > actually available hardware as an alternative to our FreeCalypso > firmwares. > > We never got any firmware (source or otherwise) directly from TI, and > we don't have any fw (again, source or otherwise) that directly > corresponds to TI's internal mainline. Instead what we have is a > Prometheus-liberated copy of a firmware which TI delivered to a > *specific downstream customer* (call it Company 1), and we did not > even get what TI gave to Company 1 in a pristine form, instead what we > got was that plus Company 1's own local modifications. Those > Company 1 modifications were either adaptations to C1's specific > hardware quirks which are not applicable to our different FreeCalypso > hardware, or just plain bugs introduced by Company 1's sw engineers > who had no clue what they were doing. We had to find and remove the > latter deleterious Company 1 modifications, identify the bits specific > to their hardware and conditionalize them so they get included only > when we build our fw to run on the scarce historical hardware from > that Company 1, and introduce some new (appropriately conditionalized) > hardware-specific quirks of our own for targeting our own FC hw. > > If you wish to run "the original TI firmware" instead of FreeCalypso, > the only way you might be able to do so would be if you manage to find > an original TI-made D-Sample or Leonardo board (not easy: in my many > years of searching I have scored one D-Sample board, but never found a > Leonardo) and run whatever fw that board already comes flashed with. > If you change TI to Company 1 (anyone with half a clue can figure out > what that company is), you might be able to find some of their hw (no > longer made for many years and not available NOS any more either), and > you could run their original fw instead of FreeCalypso if you like - > but then you would simply be running an older fw version that still > has some bugs that were fixed in our newer FreeCalypso fw versions. > > But if you are not a lucky owner of an original TI-made D-Sample or > Leonardo board and you are not interested in heavily used (not NOS) > hardware from Company 1 in beat-up condition, then the only newly made > hardware you can get with the Calypso chipset is our FCDEV3B - see my > post from less than 48 hours ago announcing the availability of these > boards. And when you do use the hardware made by us rather than some > other company, then the only TI-based firmware available for our hw is > our own FreeCalypso fw, of which we have 3 different versions: > Magnetite l1reconst, Magnetite hybrid and Citrine. > > Of course if you do buy one of our FCDEV3B boards, no one forces you > to use our FreeCalypso fw: you could simply ignore the FreeCalypso fw > image in the flash and run OsmocomBB software instead (which only runs > a teensy-tiny piece of code on the Calypso itself and feeds it to the > Calypso boot ROM serially, hence there is nothing to flash), but then > you won't be able to take advantage of the per-unit RF calibration > performed on our factory production line. We calibrate the Tx power > levels on each board with an external measuring instrument (R&S CMU200) > and save these per-unit calibration values in the flash file system, > so if you use our official FreeCalypso fw (any of our 3 versions), the > Tx power levels put out by our modem meet the official specs. But > OsmocomBB does not know how to read these RF calibration tables in > TI's format and uses a hard-coded table instead - according to > OsmocomBB head honcho Harald Welte, the latter table is only good for > some particular Mot C1xx phone, and thus stands no chance of being > correct for our FCDEV3B with different RF hardware, thus if you use > OsmocomBB, your Tx power levels will most certainly be out of spec. > > M~ > _______________________________________________ > Community mailing list > Community@freecalypso.org > https://www.freecalypso.org/mailman/listinfo/community > _______________________________________________ Community mailing list Community@freecalypso.org https://www.freecalypso.org/mailman/listinfo/community