"Tarver Engineering" wrote in message
...

"The CO" wrote in message
...

"Tarver Engineering" wrote in message
...

"The CO" wrote in message
...

Actually, it now appears highly likely that it's a problem with
the
FLASH memory management software module.

Perhaps.


snip

Today the claim is a "file management" problem, which seems likely.

Yes, the software that manages it is the suspect, as they stated
previously.

The flash memory controller is being reported as being software

Yes. That's probably not the ideal way to do that, possibly they were
considering
the likelihood of an unrecoverable hardware failure due to rad or
thermal impact
as more likely than using software for the task. It's probably that the
software
can be debugged or a new version uploaded from here, even now, however
changing out a dodgy card would not be possible.

and this
would not be the first time some softhead grabbed up all the memory
themselve, so that their program alone would run better.

I get the impression it's an actual bug that has arisen due to something
that was
not tested for rather than a memory or cycle hog.

It would net be
the first time such a thing has happened and a goo reason why memory
management units are usually hardware.

Well, it's certainly one reason.

Politics, funding and the need to protect certain people, are the
standard
reasons why Governemnt Agencies lie.

No doubt, but I was under the impression you had some specific reason to
be suspicious.
They do seem to be being quite candid at the moment. Were you perhaps
being a little
over critical based on your perception?

Software and hardware both seem to be faulty.

Software seems to be the very high probability candidate. I've heard
nothing more
about the teams investigating a possible hardware anomaly (on the motor
board) and
this seems less and less likely to be the case as time goes on.

It is an age old mistake of
developers to use the interupt that halts the processor and there was
a
scientist on TV making the claim that the processor was being
interupted,
right before the 138 reboots.

Quite so, but the interrupt is being software driven. The type of error
you describe
would be rather unlikely given the amount of development. I suspect
it's rather more
obscure than that, I get the feeling this is more likely to be something
that only occurs
during heavy memory workloads.

snip

Right now JPL is hacking the system and deleting old files and no
longer
necessary software and perhaps the robot will come back.

I think it's a little more involved than that.

snip

All we can really know is that the machine is suffering from memory
starvation right now. Some how there is a leak, much like what
happens to a
PC under Windows.

Doesn't seem to be a simple memory leak. If I may use a *very* crude
analogy,
it looks to bear some resemblance to the issue that arises with DOS when
the
root directory has it's maximum of 255 entries.

I respectfully suggest therefore that your statement is indicative
of
strong personal bias not based on
realistic knowledge of the onboard computer system. How much do you
actually know about the effects of
ionised particles/he protons on computer memory, specifically,
FLASH, or
memory diagnostic routines in VxWorks RTOS environments running on a
20Mhz RAD6000 hardened version of the PowerPC chip?

Quite a bit.

Ok. I'll be interested to see the end result of the rover recovery
efforts.

I was mistaken is claiming the two landers are identical, as the
second
robot has some chips that are a generation newer than the broken bot.

Not aware of that. Thank you. Interesting.

snip

JPL's track record may not be perfect, but I suspect that they are
well
in front of whoever is in second
place. (Russia?)

As was pointed out to the Queen at the completion of the first
America's
cup:

"there is no second place"

LOL. I can agree with that.

The US has to go back to the Moon for strategic reasons,

Strategic? What can you do from the moon that you can't do from
LEO (in the military sense)?

otherwise the Chinese will gain a military advantage within a few
years.

Hmm, seems unlikely that it would be either significant or any real
advantage.
They are very much playing catch-up ball here.

Mars is a dead planet

That is yet to be scientifically demonstrated. Make it 'likely to be a
dead planet' and
I might agree.

and probably is not going to be of much interest.

Sorry, can't agree.

I'd like to see JPL send a robot to one of the Moons of Jupiter

ISTR there is a mission of that kind planned.

instead of sending a man to Mars.

How about 'as well as'.

The CO