M
mmeron
Well, some people make a living this way
Mati Meron | "When you argue with a fool,
(e-mail address removed) | chances are he is doing just the same"
Well, some people make a living this way
Mati Meron | "When you argue with a fool,
(e-mail address removed) | chances are he is doing just the same"
J
Jan C. =?iso-8859-1?Q?Vorbr=FCggen?=
Here often anything greater than a few processor instruction cycle
On a current processor, "a few [...] cycles" is about a nanosecond. Unless
you want to depart at warp speed from a sun going supernova, I see no
application requiring that small a reaction time. The Shuttle avionics runs
on "cycles" of evaluation that are on the order of several milliseconds,
IIRC - and that is in a _very_ dynamic environment.In particular...
....experience shows that's not usually the case. The context-switching
overhead determined by software (OS) architecture, but also by the hardware
(size of processor context, memory hierarchy, etc) seem to have led to
latencies that seem not to have improved much in the past decade, if not
worsened.
....which is very difficult to evaluate on a modern processor in any case.
Jan
On a current processor, "a few [...] cycles" is about a nanosecond. Unless
you want to depart at warp speed from a sun going supernova, I see no
application requiring that small a reaction time. The Shuttle avionics runs
on "cycles" of evaluation that are on the order of several milliseconds,
IIRC - and that is in a _very_ dynamic environment.In particular...
....experience shows that's not usually the case. The context-switching
overhead determined by software (OS) architecture, but also by the hardware
(size of processor context, memory hierarchy, etc) seem to have led to
latencies that seem not to have improved much in the past decade, if not
worsened.
....which is very difficult to evaluate on a modern processor in any case.
Jan
J
Jan C. =?iso-8859-1?Q?Vorbr=FCggen?=
NASA is renowned for its antenna failures - the Hubble space
HST didn't and doesn't have antenna problems.
That one is going 'round the sun, and has had no problems of the sort.
....which isn't having an antenna problem, either.
As far as I can remember, the only probe to have an antenna failure was
Galileo. Feature creep caused by delays and other mods caused that one -
a bit of bad luck was also involved. OTOH, if the delays hadn't been,
another problem would likely have caused it to disintegrate during cruise,
if that hadn't been (painfully) found on another satellite and a work-around
identified.
As usual, you win some and you loose some.
Jan
HST didn't and doesn't have antenna problems.
That one is going 'round the sun, and has had no problems of the sort.
....which isn't having an antenna problem, either.
As far as I can remember, the only probe to have an antenna failure was
Galileo. Feature creep caused by delays and other mods caused that one -
a bit of bad luck was also involved. OTOH, if the delays hadn't been,
another problem would likely have caused it to disintegrate during cruise,
if that hadn't been (painfully) found on another satellite and a work-around
identified.
As usual, you win some and you loose some.
Jan
R
Robert C. Martin
I only know of one antennae failure. Gallileo's antenna failed to
open correctly. As I recall there was speculation that some of the
grease in the struts degraded (or something) during the long years of
storage following the Challenger explosion.
NASA has had it's up's and downs. That's understandable. They are
doing things that nobody has ever done before.
U
Uncle Al
Jan C. Vorbrüggen said:
It sure as Hell did. When they got it up there NASA discovered that
the high gain antenna feed cable intersected space swept out by the
high gain antenna. The technical term for this is "pookie pookie."
The Jupiter orbiter could never deploy its high gain antenna. While
NASA spurted all over the Media for years, data was being received at
a rate recalling Radio Shack computers and their modems. Only a very
small fraction of the collected data was ever relayed. What a
successful mission.
No. It turns out $240 million won't custom-build a NASA FlashRAM card
as good as one can purchase at Radio Shack. That's no surprise. $1
billion could not grind and polish a NASA version of a Keyhole spy
satellite main optic available off the shelf. In pure NASA Korporate
Kulture tradition, the old fart optician who screamed about the
Hubble's mirrror blank being rough ground to the wrong spec was fired
very early on. It cost another $billion to fix it in orbit.
FOR THE EMPIRICALPRICE OF FINALLY DOING HUBBLE CORRECTLY, NASA COULD
HAVE BUILT AND ORBITED NEARLY THREE OF THEM USING KEYHOLE SATELLITE
OFF THE SHELF PLANS AND PARTS.
[snip]
Taking a picture of Martian dirt is not a major triumph. We already
have surface meterological and chemical data. Unless a nice block of
limestone turns up, with included fossils, this has been a huge and
hugely expensive bullshit party. How many Enron ex-executives are
working at NASA?
R
Randy Howard
A very good point, and I am a bit surprised it hasn't been said
previously in this thread. The public seems to have convinced
themselves that space travel is no more difficult than going down
to the corner store for a loaf of bread. It's not even remotely
true, with or without a passenger on board.
NASA shouldn't be getting a "black eye" for such failures, people
should be saying "good try, when are you going to make another
attempt at solving such a difficult problem". Stop and think
for a minute about how effective you would be at debugging your
project if the link between your development machine and the
system under test was on another planet and the delay between
inputs commensurately slow. Have fun. How would you debug
hardware problems remotely if you could not have any physical
contact with the machine? EVER AGAIN?
A
Andrew Thompson
"Randy Howard"...,
| (e-mail address removed) says...
| > NASA has had it's up's and downs. That's understandable. They are
| > doing things that nobody has ever done before.
|
| .... The public seems to have convinced
| themselves that space travel is no more difficult than going down
| to the corner store for a loaf of bread. It's not even remotely
| true, with or without a passenger on board.
Don't be silly, have you ever tried
to purchase bread on Mars?
Even if you can find a baker, it's
stale by the time you're Earthside. [ ;-) ]
| ..How would you debug
| hardware problems remotely if you could not have any physical
| contact with the machine? EVER AGAIN?
Assuming you can get it to obey _any_
commands.
a) Orient front of rover toward big rock.
b) Drive to 3 metres distant from rock, stop.
c) Drive forward 4 metres.
d) Back up 3 metres.
e) Repeat c), d) until problem fixed or rover destroyed.
Easy peasy.
You folks just do not think laterally. ;-)
| (e-mail address removed) says...
| > NASA has had it's up's and downs. That's understandable. They are
| > doing things that nobody has ever done before.
|
| .... The public seems to have convinced
| themselves that space travel is no more difficult than going down
| to the corner store for a loaf of bread. It's not even remotely
| true, with or without a passenger on board.
Don't be silly, have you ever tried
to purchase bread on Mars?
Even if you can find a baker, it's
stale by the time you're Earthside. [ ;-) ]
| ..How would you debug
| hardware problems remotely if you could not have any physical
| contact with the machine? EVER AGAIN?
Assuming you can get it to obey _any_
commands.
a) Orient front of rover toward big rock.
b) Drive to 3 metres distant from rock, stop.
c) Drive forward 4 metres.
d) Back up 3 metres.
e) Repeat c), d) until problem fixed or rover destroyed.
Easy peasy.
You folks just do not think laterally. ;-)
R
Richard Heathfield
Robert said:
Modulo the apostrophes, that's great tee-shirt material.
D
David C DiNucci
Randy said:
While I hesitate to bash NASA (I do think they're doing very important
work, and I worked there myself for seven years), I regrettably prolong
a thread which has been OT from its inception with this excerpt from a
recent news article:
"[Mission manager Jennifer] Trosper said the problem appeared to be that
the rover's flash memory couldn't handle the number of files it was
storing. ... She pointed out that the scientists had thoroughly tested
the rover's systems on Earth, but that the longest trial for the file
system was nine days, half of the 18 days Spirit operated before running
into the problem."
http://www.cnn.com/2004/TECH/space/01/26/mars.rovers/
"Thoroughly tested"? If you're going to send any object, and especially
an object with a computer and software, to a distant planet where it is
supposed to survive for about 90 days, wouldn't it seem prudent to run
at least a 90 day test of the object on earth before liftoff?
-- Dave
D
Dave Seaman
It seems the Spirit is willing, but the flash is weak...
M
Michael N. Christoff
Uncle Al said:
Actually, it apparently was not the flash ram, but the particular software
that accessed the flash ram. Thats why they think they can fix it by
sending a software patch.
l8r, Mike N. Christoff
M
Michael N. Christoff
B
Bernd Paysan
Robert said:
I don't agree. NASA does things that nobody has ever done before, but most
time it fails, it fails because it does things that a lot of people have
done before, just differently. My impression is that NASA has a big NIH
problem. They had it back before Gemini, because they didn't use their
German rocket experts, and their own rockets blew up again and again, until
the Sputnik shock forced them to rethink a bit. On the other hand, Russian
rocket science is dead conservative industrial production. They still use
slightly modified V2 aggregates in their rockets - lots of them, making
failures unlikely, and since they are "mass products" (several thousands a
year), they are cheap.
When NASA says their flash file system is having problems, I'd like to bang
my head against a wall. Why didn't they use a off-the-shelf flash file
system that's known to work? For all NASA engineers: My footer has an
obligation: "If you do it yourself, you better do it right!".
A
Alan Balmer
Don't give up so easily ;-) From the article you reference:
"Trosper said the problem appeared to be that the rover's flash memory
couldn't handle the number of files it was storing."
That sounds like a (possibly) software or (more probably) design and
specification problem, not hardware.
In any case, I'm interested to discover that Radio Shack FlashRAM
cards are far superior. I wonder how the OP determined that?
U
Uncle Al
David said:
HERETIC!
Any child of BillGates' Windoze knows that
1) Memory is unlimited, and
2) You *want* crud files to massively accumulate without any
callable mechanism for erasure. When the partition mysteriously blows
out, the ignorant consumer is screwed into reinstalling a purchased
update. Check out th size of your Windoze *.log files. What do they
do?
Uncle Al partitions a 3 Gbyte C: drive to hold *only* Windoze. It can
be rewritten without sacrificing the rest of the hard drive. It is a
contained garbage dump that can be manually purged. Run ZTREE and get
control of goddamned Windoze file and subdirectry obscenities,
http://www.ztree.com/
http://www.mazepath.com/uncleal/zt.zip
For a nice screen configuration file. You can flip ZTREE underneath
your idiot Desktop by hitting the Windoze key then Esc.
Uncle Al's box came from Gateway (cheap) with Windoze98. Uncle Al
spent months deleting files. The 12,000 original files are now down
to 7200. When things are slow... out go another few dozen. Do you
think this is funny? Look for all *.png files in your Windoze
partition. Does that voluminous pile of crap do anything? No, it
does not. Purge the image files, get back a few MB of hard drive.
(But who cares about a few MB? Memory si unlimited...)
Back to the mission... FlashRAM degrades with use. Each cell is good
for maybe a half-million cycles in commercial chips. Did NASA custom
fabricate its own NASA-crappy FlashRAM instead of visiting Radio
Shack? Did they save 20 grams by putting in half as much memory as
they really needed?
There are no hard drives on Mars! With ambient air pressure being
only 7 torr, there isn't nearly enough air pressure to levitate the
read/write heads when the hard drive spins up. If you seal the hard
drive, it overheats.
A
Alan Balmer
But (some) hard drives are indeed sealed. E.g.
http://www.wmrc.com/businessbriefing/pdf/data03_1/technology/mountop.pdf
J
JatJatIan
As I understand, the Rover uses VxWorks, which is one of the more
popular off-the-shelf RT OSs. VxWorks supports a flash filesystem,
and I'd guess NASA would use it (although they could have written
their own).
In any case VxWorks, is not bug-free. It contains bugs that you'd be
very suprised at. But as it is a niche OS, bugs that might be
considered unacceptable in a general purpose OS are viewed
as funny quirks.
It is very possible that its flash file system sucks up large amounts of
RAM where there are many files, and this might not be considered a
problem as it might be uncommon for RT apps. Perhaps the most common
usage of flash memory is executable images and configuration files.
Thanks for reading,
Jatian
U
Uncle Al
Alan said:
Point taken!
"MOI has successfully deployed these hermetically sealed drives in a
variety of Department of Defense and commercial applications."
Why would NASA tolerate any other department's success? Mars' ambient
temp being 0 C and below, cooling is not much of an issue. But the
power use! Except ultraminiature hard drives are not rare, either.
Some digital cameras have them. But the bouncing! Park the heads.
But there are no reliability studies! Hire some high school summer
interns and do them. But there are UNKNOWN HAZARDS! Ahhhhhh....
Repeat after Uncle Al, "NASA, SNAFU FUBAR." The first Space Scuttle
had woven ferrite core memory. But then, the Space Scuttle was
designed as a low-orbital nuclear bomber and woven core is
EMP-resistant. NASA is the gang who couldn't shoot straight, casting
their own guns out of pot metal and machining their bullets out of
Mil-Spec shitanium.
As I watch this sorry saga unfold I get the overwhelming impression
that NASA is strutting its stuff and demanding vast Media exposure for
what any competent civilian engineering group would consider to be
piffle. Given a $780 million budget, they could have air-dumped a
rover in an Antarctic dry high desert and done a real time real world
reliability test. (Long range P-3 Orion sub-chasers are currently
underemployed. C-5A cargo planes can fly mostly empty. Vandenberg is
pretty good at firing stuff over a pole.) But wait! NASA must file
an Environmental Impact Report! With somebody or other.
NASA, "It will require weeks to determine the true color of Martian
soil." Didn't any of our nation's best and brightest think to paint
on a color calibration chart? I can walk down the street to a camera
shop and buy one, cheap. I'm certain that Pantone could have supplied
something for 1000X civilian price.
Marines don't sit around crying into their inventory sheets. They
hide stuff in their socks, then go out and kill the enemy. A
perfectly budgeted war with digital readouts doesn't mean much if you
lose it to ragheads using wooden clubs. (Except the REMFS will all
get promotions while the body bags are stacked. Which of those two
sets is NASA?)
Last comment: We've got an apparently functional rover (until its
FlashRAM mysteriously craps out in two weeks) within pissing distance
of naked Martian bedrock. Said stone has a very large albedo compared
to regolith, perhaps like limestone. Does anybody here really believe
the rover will make it over to the rock and have a good look? Or will
there be "intervening priorities" that result in the rover's surprise
disablement?
Oh yeah... Is the horrible stupid thing of a rover capable of
translating its position over slabs of smooth dissected rock? You
know where Uncle Al's bet is placed. It wasn't in the specs!
A
Abhi
Michael N. Christoff said:
Just for information through my inbox...
---------------------------------------------
DC Agle (818) 393-9011
Jet Propulsion Laboratory, Pasadena, Calif.
Donald Savage (202) 358-1547
NASA Headquarters, Washington, D.C.
NEWS RELEASE: 2004-040 January 27, 2004
Martian Landmarks Dedicated to Apollo 1 Crew
NASA memorialized the Apollo 1 crew -- Gus Grissom, Ed White and Roger
Chaffee -- by dedicating the hills surrounding the Mars Exploration
Rover Spirit's landing site to the astronauts. The crew of Apollo 1
perished in flash fire during a launch pad test of their Apollo
spacecraft at Kennedy Space Center, Fla., 37 years ago today.
"Through recorded history explorers have had both the honor and
responsibility of naming significant landmarks," said NASA
administrator Sean O'Keefe. "Gus, Ed and Roger's contributions, as
much as their sacrifice, helped make our giant leap for mankind
possible. Today, as America strides towards our next giant leap, NASA
and the Mars Exploration Rover team created a fitting tribute to these
brave explorers and their legacy."
Newly christened "Grissom Hill" is located 7.5 kilometers (4.7 miles)
to the southwest of Spirit's position. "White Hill" is 11.2 kilometers
(7 miles) northwest of its position and "Chaffee Hill" is 14.3
kilometers (8.9 miles) south-southwest of rover's position.
Lt. Colonel Virgil I. "Gus" Grissom was a U.S. Air Force test pilot
when he was selected in 1959 as one of NASA's Original Seven Mercury
Astronauts. On July 21, 1961, Grissom became the second American and
third human in space when he piloted Liberty Bell 7 on a 15 minute
sub-orbital flight. On March 23, 1965 he became the first human to
make the voyage to space twice when he commanded the first manned
flight of the Gemini space program, Gemini 3. Selected as commander of
the first manned Apollo mission, Grissom perished along with White and
Chaffee in the Apollo 1 fire. He is buried at Arlington National
Cemetery, Va.
Captain Edward White was a US Air Force test pilot when selected in
1962 as a member of the "Next Nine," NASA's second astronaut
selection. On June 3, 1965, White became the first American to walk in
space during the flight of Gemini 4. Selected as senior pilot for the
first manned Apollo mission, White perished along with Grissom and
Chaffee in the Apollo 1 fire. He is buried at his alma mater, the
United States Military Academy, West Point, N.Y.
Selected in 1963 as a member of NASA's third astronaut class, U.S.
Navy Lieutenant Commander Roger Chaffee worked as a Gemini capsule
communicator. He also researched flight control communications
systems, instrumentation systems, and attitude and translation control
systems for the Apollo Branch of the Astronaut office. On March 21,
1966, he was selected as pilot for the first 3-man Apollo flight. He
is buried at Arlington National Cemetery, Va.
Images of the Grissom, White and Chaffee Hills can be found at:
http://www.jpl.nasa.gov/mer2004/rover-images/jan-27-2004/captions/image-1.html
The Jet Propulsion Laboratory, Pasadena, Calif., manages the Mars
Exploration Rover project for NASA's Office of Space Science,
Washington, D.C. JPL is a division of the California Institute of
Technology, also in Pasadena. Additional information about the project
is available from JPL at http://marsrovers.jpl.nasa.gov and from
Cornell University, Ithaca, N.Y., at http://athena.cornell.edu.
E
Edward Green
David C DiNucci said:
Oh my #$*&%'ing G*d.
Ok ... obligatory disclaimer about difficulty of ...
Screw it! This is a a languid courtesan, reclining seductively, and
saying "bash me, big boy"!
I am almost flabbergasted into textlessness. The fact that a system
.... any system, not just a computer ... may work correctly in some one
or two delta range yet fail in some 10 or 20 delta range ... some
newbie tyro university graduate wet behind the ears neophyte kid might
make this mistake in a small project, and the old seasoned pro salt
seen it all manager would take this as a teaching opportunity. But in
an entire organization, a huge project putting a robot on a distant
planet, and not once did this occur to anybody!? Nobody with
sufficient experience or a long enough memory reviewed the test plans?
Anybody who has operated MicroSoft windows has had the same
experience: crap accumulates with operation.
This is the type of thing which used to get me flamed when I chatted
in the all-computer-pro group of a local ISP: since I can no longer
write even a Basic program, how can I dare comment ... oish. Wet
nurse tyro-ish newbie error. Inbred self-congratulating community.
Industrial experience of a polywog. Etc.
Help me out here, Uncle Al: I'm running out of germane insults. ;-)
Size of institutional memory buffer needs to be increased. Political
ossification makes effective project review impossible.