BG> I doubt that USR were equally as amused though.

db> Their problem .. if nothing else, it just serves to prove
db> that USR are as fallable as the other modem manufacturers.  :-)

BG> I actually feel a bit sorry for them, as it wasn't really their fault.

Corse it was, the fuckwits should have been checking that the
flashrom could be erased in its entirety in the final QC test.

Pathetically shithouse quality control.

BG> Especially as it looks very much like they'll need to go back to USR for
BG> repair.  I bet they're sorry they didn't socket the flash-ROM chip now...

Well, that one is more forgivable given the number of
programming cycles that its supposed to be able to do.

Corse they have some absolutely massive fucking brain farts on that
stuff, like the Sportsters without even a fucking eprom socket. The
dork responsible for that should be taken out the back and fucking shot.

db> At what cost to users (considering USR didn't force
db> the users to flash the new firmware into their modems)?

BG> It's not an SDL problem though.  Turns out that USR use both Intel and
BG> AMD flash-ROM chips, and SOME of a recent batch of AMD devices had been
BG> internally flagged as "read-only" in their higher
segments, a problem
BG> which was only discovered when the latest SDL exceeded a certain size,
BG> thereby writing to the chip in areas hitherto unused. It's not the sort
BG> of problem which would show up in normal testing with normal SDLs either.

Pigs arse its not, the final QC check should have checked that you could
actually use the entire flash rom, not just what the current SDL uses.
@EOT:

---