Mars Rover Spirit Falls Silent ...

[Ned]

The first pictures from opportunity show a really different landscape and features that look like the remnants of a seabed.

[Gatekeeper]

All in all, this turning out to be quite good. Spirit seems to be making a slow, but steady, recovery and now Opportunity is raring to go. In addition to the usual benefits successes like this provide to the scientific community, it's also a big morale booster for the good folks at NASA.

Gatekeeper

[Ned]

Gatekeeper, in a way, I sense this is a morale booster for all of mankind.

[Ned]

Now here is an interesting foto:

http://marsrovers.jpl.nasa.gov/gallery/press/opportunity/20040125b/Pancam_Postcard_WedgeV_bw-B22R3.jpg

Too large to post.

[The Mad Monk]

Okay, over at another site we have a poster who actually works for NASA. This is what he posted about Spirit over the past few days:

01-26:

OK...
So far...

Official Explanation #1: The problem is that the rovers are programmed with Java, and it cannot handle more then 3 tasks at once.

Official Explanation #2 (less favored): It's the Java. Really! Java sucks! We told you not to go with Java! The 16 different flavors of embedded Java are not actually working together very well.

Official Explaination #3 (from engineering, not as favored as #1 and 2): Well, we told everyone that Flash was great. But if you are reading to/from Flash, or writing TO Flash, whenever there is above average solar activitity, then the Flash will corrupt. But flash was just so freaking CHEAP! And, there was a serious solar storm that happened while we were doing science. But it's not OUR fault. You guys said you wouldn't be constantly reading or writing to Flash! Except for those 56 hour straight science things. And the 125,000 hour straight science observations. And running the clock to know when to power back up... and ...

Official Explanation #4: Well, this is just more Columbia management ****. Yuu see, we didn't actually TEST the rover, on orders from management. Because if anything failed, it would then have to be fixed, and that would have made it cost too much. Project cancelled. And even if the fix/correction didn't put us over (we were on the bubble), then we'd have lost our launch from the time it took to work (because we were on the bubble for time as well), and we were told "It's going, regardless. Just send back pretty pictures, so that we don't look like balless space powers, since Japan and ESA (Europe) is going! It has to go! So, make it happen, or be fired".

That's all straight up from the Mars teams and their management.

So... it's either politics as usual and ****ty engineering, or just ****ty engineering. (Java was chosen because the original lead computer scientist wanted to get Java on his resume before he left NASA.)

This, by the way, is suppose to never happen again, but NASA has already bastardized the "checks" put on them from following the CAIB. All hail the future dead astronauts!

01-28:

BTW... latest word I heard on what was Spirit's problem. Right now, the #1 believed cause of their problem is: Out of Disk Space.

Ok, it's actually just out of flash memory space. As Spirit gets to clean up it's memory, it seems to be doing better.

Imagine that. 256 Meg of Flash (which is double their RAM (128 Meg)), just wasn't enough for taking all of those 100,000 jpeg pictures! Well, that an those 12 hour X-Ray science observations, and all the other stuff...

The explanation why/how they filled it all up without noticing? They forgot to count in all the space still taken up by the pictures that they ordered taken and haven't yet downloaded from the rover PRIOR to driving it over Adrionback rock.

What I love is an interview with one of the Rover designers saying "We added 256 Meg of Flash Memory to the Rovers, because on Sojourner, they almost used up all of its 128 Meg Ram, and we wanted to make sure we would have so much storage, we'd never be able to use it all up in a year at constant science observing and taking one picture every second!"

So far... The released list of likely causes/excuses: cheap wiring or cheap switch. Or piss poor coding. Or Java is a piece of **** and cannot even handle bools consistantly.

(Sounds like: Bad design and lack of adequate testing to me, so far.)

NASA...would you PLEASE get your CRAP together?!

[Zylka]

dumb

STUPID

[asleepathewheel]

can you post what website that was off of?

[Dr Strangelove]

I take it that they sent two Rovers at the same time precisely because they were afraid that what might happen actually did happen which in fact actually did happen?

[The Mad Monk]

Counterglow.

[Brundlefly]

Originally posted by The Mad Monk

NASA...would you PLEASE get your CRAP together?!

Kinda puts those quotes from your signature in perspective, don't you think?

[Ned]

I can hardly imagine they programmed in JAVA.

Really?

[Ecthy]

they've mixed up metric and imperial before, after all

[The Mad Monk]

Originally posted by bfg9000


Kinda puts those quotes from your signature in perspective, don't you think?

Not really. The problems spring from the constant underfunding, whch encourages people to cut corners, or else.

[The Mad Monk]

Originally posted by Ned
I can hardly imagine they programmed in JAVA.

Really?

Internet language runs remote-controlled Mars rover

PASADENA, California (Reuters) -- The same piece of software that lets people all around the world play video games on their cell phones is now letting scientists drive the ultimate remote-controlled car across the surface of Mars.

Java, the software developed by Sun Microsystems Inc. in the mid-1990s as a universal platform for Internet applications, gave NASA a low-cost and easy-to-use option for running Spirit, the robotic rover that rolled onto the planet's surface on Thursday in search of signs of water and life.

For the next three months, NASA scientists and engineers at the Jet Propulsion Laboratory in Pasadena will plot Spirit's wanderings with the Java-based Science Activity Planner that operates like a digital Gran Turismo.

"It takes all the raw data in the mission data base and builds a 3-D terrain you can spin around and zoom in," Gene Chalfant, JPL technical staffer, said.

With the same point-and-click skills one would need for, say, online shopping, the NASA team will plan Spirit's daily activities, page through voluminous data and communicate.

"It's a sandbox, in a way, to try different ideas," Chalfant said. "You pick the rock you want to investigate and command the rover to move there and the rover figures out the best way."

The team made virtually no changes to an online version of the program, dubbed Maestro, that lets space nuts page through panoramic color images, check out the rover's wheel-mounted hazard cameras or plan a rover mission just like real scientists.

The site (http:/mars.telascience.org) has been so popular since its January 2 launch that Sun had to provide extra bandwidth to keep NASA's servers up and running, Chalfant said.

The simulated rover drives on a 3-D model of the Martian terrain as precise as the one used by the NASA mission.

"[Scientists] do exactly the same thing you can do," Chalfant said.

Java's journey from mundane to extraterrestrial began nearly a decade ago when JPL scientists began noodling with the programming language to create a command and control system for the 1995 Mars Sojourner, said James Gosling, known as "the father of Java."

The JPL team showed Sun what they had done, and Gosling, a vice president and fellow at the Santa Clara, California-based software and systems developer, was hooked.

"I'm a geek anyway, so it sucks me in," Gosling told Reuters. He spent so much time at the Pasadena space laboratory that he became an advisory board member.

"They are doing things that people think are science fiction," he said. "It's a place to go to have your mind blown. It's hard to find a government agency ... where people are living their dreams."

Although Java's data-handling capabilities initially attracted NASA, the code's ability to transcend the many platforms used by mission scientists and engineers sold the space agency, Gosling said.

"They can have scientists all over the world looking at the data but collaboratively deciding on the way the mission should proceed," Gosling said. "They are all speaking different languages when they talk to the rover but everybody in the control room is using Java."

Separately, Alameda, California-based Wind River Systems Inc., created the embedded software in Spirit and its twin, Opportunity, that manage a wide range of functions, including data collection and communications.

http://www.cnn.com/2004/TECH/space/01/16/space.mars.java.reut/

[Wraith]

--"I can hardly imagine they programmed in JAVA."

Well, it may have been Java, but... see the tagline.

Wraith
"If you try hard enough you can make any language FORTRAN."
-- NASA