Slashdot Mirror
World's Worst Dog'n'Pony Shows
Thanks to coli for passing along last Thursday's press release from VisuaLABS. This is a company that has been telling investors that they have what they call "GroutFree(tm)" technology, which joins multiple LCD screens invisibly into one, large, flat screen.
On July 3rd, investors were wowed by the demo of the company's "42 inch diagonal flat screen display" prototype. Sheldon Zelitt, VisuaLABS' Chairman and Chief Scientist, said, "It was our great pleasure to share an early look at that technology with our loyal shareholders at the Shareholders' Meeting."
And on July 26th, we got another press release -- this one titled "VisuaLABS Announces That Its Primary Technologies Are Not As Represented And Dismisses Sheldon Zelitt." It turns out that "the large screen GroutFree prototype demonstrated at the Annual Meeting was, in fact, a standard 42 inch plasma television purchased by Sheldon Zelitt ... at a local Calgary consumer electronics retailer ... The Committee believes that no working prototype of a device incorporating the GroutFree technology exists."
While all this was going on, the Pentagon was busy launching two missiles and making them smack into each other. This is the missile defense justification, the one scientists say can't be done, the umbrella that will protect the U.S. and its allies from all those Third World dictators who just have to deliver their nuclear warheads the hard way.
The big test came on July 14, when a target missile (avoiding mishaps) was launched and successfully blown to pieces by its interceptor. Bush was "pleased." CNN showed us the debris radar. And Michael Kelly of the Washington Post stuck it to the "liberal critics," pointing out that "The 'Smart People' Were Wrong." As he wrote:
"In the blink of a video screen going blinding white on July 14, it became impossible to offhandedly disdain a missile defense system as 'weapons that don't work.' It does work."
Yep! So phase one of our missile defense plan is complete. Now we go on to phase two, which is to convince all our enemies to install GPS transmitters in all their missiles.
Oh, you didn't know the test missile had a GPS transmitter on board? Well, you do now.
My favorite part is that the test missile actually launched a Mylar balloon as "chaff" to try to fool the "kill vehicle." Luckily, the balloon didn't have GPS.
So what's your favorite dog'n'pony story? Ever had a demo fail in some especially embarrassing way? Ever cheated? Ever get caught? C'mon, you can tell us...
Update: 08/01 08:00 PM by J : I'm seeing a lot of discussion of the relevance of the GPS. Here's Defense Week which claims the "prototype interceptor was able to find a target warhead partly because the target signaled its location to the interceptor for much of the flight, and the transmissions formed the basis of the targeting orders."
And thanks as always to Slashdot readers for posting more information. monopole points out this link, or take your pick, this one -- they're plans from last year, but still interesting:
SR. DEFENSE OFFICIAL: And we take the GPS data, and we fuzz it up quite honestly, because GPS is a lot more accurate than radars. Okay? [...]
Q: Well, actually, would you then use the degraded GPS, or would you just the regular GPS that you use as a fallback -- (inaudible word)?
SR. DEFENSE OFFICIAL: (Inaudible.)
STAFF: Use the regular GPS.
SR. DEFENSE OFFICIAL: Regular GPS.
Hey, as long as we're on the topic of Microsoft, who remembers the infamous X-Box screenshots brought to you by the good folks at Adobe? :P
I had been working at Apple for about one year and I was on the list of fast trackers (that's an Apple insider term for engineers who get the series A projects) which I would soon regret. I was put in charge of the project to port the two button mouse over to the Mac OS, due to a serious outcry by all the people who were buying Macs and found them to be utterly useless with only one mouse button.
I had my team working day and night to try to get a prototype ready for Comdex but we knew it was not going to be possible. All of our code was in C++ and it was too slow and buggy due to the flaws of the language, and we were looking at doing a total rewrite to try to get to right. That would be the first of many, by the way.
Well we told Steve what we needed to do and I suggested he cancel the big two button Comdex demo, and he went friggin nuts. I thought he was going to belt me right in the chopper. No, he insisted that we "just rig the fucking thing up" over and over. So I ended up taking a one button mouse and cutting it in half, right down the center, so it looked like there were two buttons. Of course you would get the same results reagrdless of which button you pressed, but Steve could work through that. Well, as it turns out, the demo got canceled anyways and we all breathed a huge sigh of relief. I think Steve never really liked me from then on though, he's a tough guy to get along with.
All the best,
--Bob
Possiby the target had a GPS receiver in it, either to keep it on track, or to transmit telemetry down to the ground so they could replay the results of the tests?
Now clearly, anything transmitting telemetry is easy to home in on if you seek to the frequency of the telemetry, but that's such a complete cheat I wuld need more evidence before believing it.
When you do a test like this, you don't do it just to get the binary answer of "did it hit?" You want complete telemetry on both missiles to find out exactly what happened, what went right, what went wrong. That the seeker missile would get to take advantage of the telemetry would be ridiculous and a grand scandal.
(NOt that it would make it easy. You couldn't hit something just from its GPS coordinates.)
Has it been over a year since you last donated to the Electronic Frontier Foundation
There's always the anecdote about a company that was giving a demonstration of speech recognition in MS-DOS. (In one version I've heard it was Creative Labs; in another it was Microsoft.)
The marketing flack giving the presentation arrives at the most dramatic and impressive bit, where he demonstrates the system's capability to recognize a set of related commands. He keys the mike, preparing to speak a command, and someone in the front row of the audience shouts "FORMAT C COLON! ENTER!" Someone in the row behind him shouts "YES! ENTER!" and history is made.
In all fairness I have no source for the story, but I once used a DOS-based Creative Labs product that would easily have been capable of such a feat. I believe it was called Voice Commander, and I trained it to recognize the letters of the alphabet, plus some punctuation and DOS command words.
Somewhere along Interstate 90 there is a 300 yard concrete monument to poor preventative maintenance.
I worked for a small company in west Iowa that manufactured concrete paving equipment. Big machines. This particular baby was about 60,000 pounds of steel and hydraulics.
A contractor purchased one of our machines second hand. Since we were a small company, we generally would bend to the wishes of the customer to help with bringing the equipment online. And in this case, this contractor had scheduled a demo of this machine to the Minnesota DOT so we wanted our equipment to look good.
We had spent several days pouring concrete, making road and adjusting the machine and controls. I was involved with the design of the crown adjusting computer. Every time something wasn't quite right, the contractor blamed the crown computer. While fiddling with the controls to satisfy his ego, I noticed one of the hydraulic pressure guages was madly wiggling. I pointed this out to the contractor. I was told it didn't matter and that I didn't know anything about hydraulics. Side note: I am a mechanical engineer ;)
Next day bright and early half of MDOT showed up for the demo. The dump trucks were lining up in front of the machine. Each truck carried approximately 10 yards of concrete. Well there was quite a line of trucks when the machine *sucked* a hydualic pump and destroyed the whole system. It stopped, shook and then just sat down on the gound squeezing concrete out like toothpaste. So much for the show. It lasted maybe 45 minutes.
MDOT left. The trucks dumped the their loads to the side of the road and I went home.
Inhanbla Gmunka
"The probability of a successful demo is inversely proportional to the number of people watching it."
We've got one coming up on Friday.
Shit, why am I replying to SlashDot? I have bugs to fix...
Well, the new facility was all cubicles, with no closets designed in... no place to put junk, so we shoved boxes full of junk under desks and in out of the way corners. We were working on a lab, and the VP of Engineering decided that the test leads on the oscilliscope looked untidy, so those were removed and hidden as well.
When that was done, the VP looked around the lab and said "Now, this looks like a productive lab!" Yeah, a scope with no test leads... that's real productive.
I sent this story to Scott Adams. I don't think he ever used it in Dilbert, but he did send me a reply saying he found the story amusing.
steveha
lf(1): it's like ls(1) but sorts filenames by extension, tersely
Back around 1991-1992 a Company I worked for was making a Warehouse Database Management Control System for a very large cigarette company. They spent something like seven maybe eight months working on it and they had a pretty nice demo for the customer to show them how on-track they were for installation a mere month or three away.
There was only one problem. The design team hadn't actually coded in the database portion of the DBMS. They just used flat files and convinced both their own employer as well the customer that everything was working just fine.
I don't know how anybody found them out but by the time they were done they fired the whole development team and started over again. The second time they actually instituted oversite into the project. What a novel concept.
Beware the wood elf!!!
Agreed. Not just for the bad PR, but because it also hurts the troops when the weapons don't work.
> What really sucks about this case is that they were open and proactive about admitting what parts of the test were not the same as the proposed operational system, and they're STILL getting beaten up over it.
Amen, brother. Chalk up another "victory" for the media using the public's ignorance in order to further its own agenda.
> the successful destruction of that missle (plus the decoy avoidence) is impressive - and legit.
Amen again. The test demonstrated that if you have a functioning tracking system, you can pull off kinetic kill. The next step is to build a tracking system while simultaneously refining your kinetic kill capabilities.
(Yeah, if it were me designing the thing, I'd go for "blows up somewhere near the target with sufficient force that 'close enough' is good enough to destroy the missile", as opposed to a pure kinetic kill approach. But it doesn't take away from the fact that the ability to do KK means you've built some fscking impressive tech.)
Everyone remember when Steve Jobs impressed MacWorld by booting a big wall of macs from a single OS X server? turns out that they were using a forked version of OS X that was abandoned right after the show, so those features aren't present in today's version. But more than that, rumour has it that the OS X server machine was using PC hardware to accomodate ungodly amounts of RAM so it could run a copy of MacOS for each one of the machines.
I always thought that was funny.
chris
New York Times by Tim Weiner, August 27, 1993 "Last year, the GAO audited
seven "Star Wars" tests between 1990 and 1992. The auditors found that
three of the tests were accurately described to Congress. Those three tests
were complete or partial failures. The missile defense program's officials
told Congress the other four tests were successes. That was untrue, the
auditors said.
The inaccurate claims included the success rate of experiments, the
progress of the programs, the sophistication of the tests, the ability of
interceptor missiles to distinguish between a target and a decoy and the
missiles' achievement of accuracy and altitude goals, the GAO reported.
'They have lied about certain functions that their missiles are supposed to
perform,' said a Federal investigator who agreed to speak only if he was
not identified. 'They've used things to enhance the target. The fact is
that you've got something up there solving your guidance problem. And
you've got an incentive to deceive. That's how you keep your program
going.'
A former Reagan administration official, a nuclear physicist who closely
studied the missile defense program in the 1980s, said it was characterized
by 'secrecy, greed, self-deception, deception of Congress and actually even
of the President.' The former official, who remains a Pentagon consultant
and who spoke on condition of anonymity, is not among the accusers in the
debate."
New York Times by Tim Weiner, August 18, 1993 "Officials in the 'Star
Wars' project rigged a crucial 1984 test and faked other data in a program
of deception that misled Congress as well as the intended target, the
Soviet Union, four former Reagan administration officials said."
A guy I worked with once (he was later fired) was supposed to demonstrate the functionality of a database-driven website that he had written (an online community sort of thing) back when this was still a relatively new concept. His resume listed this sort of thing as his specialty. The client arrived, and he started demonstrating by logging in, and voila, the home page customized itself to his account. He went to the preferences page and clicked a few options, and voila, the home page was different now. That's when the client noticed that his name was misspelled. "Oops, must have typed it in to the database wrong, hehe." He started showing the pages again, and each time the client would ask to see a particular link, he'd blow it off and click somewhere else instead. My co-workers and I were passing around looks. Then I noticed that his name was once again spelled correctly at the top of the page.
The clients eventually left, only mildly impressed since he follow any of their suggestions on using the site while in the meeting, and I got suspicous. I tapped in to his computer (honestly, if you're wearing a Flyers tie, don't make your password "Flyers") and grabbed the files. I realized there was no database connectivity in here at all! It was all static pages that could only be browsed in one path, and the changes you made just actually linked you to a page statically programmed to show the changes. That's why his name had changed spelling!
Ok, not too impressive, but true =)
Slay a dragon... over lunch!
(Note: This have little in common with what the article talks about, as we didn't cheat for the demo)
f 1.gif
f l2.gif
The best demo I recall was back in time when I worked in porting games, mostly from Atari/Amiga to Macintosh/Apple IIGs
We were 4 developers, living in a big flat, with shitload of hardware and documentation around us, working days and night with weird deadlines.
One of my friend was porting Shufflepuck Cafe from the Mac to the IIGs. This guy was a very organised man, not a hack/compile/run/crash/hack kind of guy (and back in those times cross-compilations were pretty costly, in term of time). I always was amazed that he could code for 2 or 3 days, compile, get 3 or 4 compilation errors, fix them and add some other features, compile again, etc, etc. When the app ran and crashed, he fixes the code, and add some other features (at least, that's how I remember him). So, he did the things well, and had almost finished the port (it was about 70% done), when Broderbund software asked to be shown the avancement of the port.
The only thing they did not realize was that almost everything was made, except graphic output routines.
For the one that don't know about it (ie: most people), Shufflepuck cafe was supposed to look like:
(Remove the slashspaces after 'shuff', and yes that's 'shuff1' and 'shuffl2')
* http://macjeux.free.fr/TheSite/lesinfossuite/shuf
* http://macjeux.free.fr/TheSite/lesinfossuite/shuf
(yes, this was more or less cutting edge at the time) and you may understand that, without graphics, there much less to show.
We smacked a Couple of RectFrame to represent players paddles, and showed the game that way (no ball, only the paddles). The app was mostly complete, with sound, etc. Just no output. We demoed an app, with a nice black opening screen, choosed an opponenent by clicking in the black, which gave us two rectangles moving on the screen, and even managed to mark a few points against one of the opponent...
Cheers,
--fred
Jamie:
Jesus Christ, this isn't the place for bitching about the Bush administration. I mean, I know Slashdot is as much of a political forum as it is a technical forum, but the idea is to incite original and thoughtful commentary. You're just parrotting propaganda (don't bullshit yourself-- it's still propaganda, even if you agree with it).
(For the record: I don't agree with the Bush administration's position on missile defense. I didn't vote Republican. I don't even like Bush's policies. I just get sick of Jamie's unoriginal re-packaging of pure rhetoric.)
That said, here's my favorite story:
When I was in 9th grade, I was assigned the task of developing a new programming language, then implementing a compiler for it in C. I had two months to do it.
It wasn't much, but I actually did get the job done. The language was very FORTH-like-- stack based, a small set of basic commands, a few primitives for defining subroutines. The compiler generated assembly code (nasm was used to assemble it), and I had implemented a rudimentary peep-hole optimizer program that could be run as a second "pass".
My computer teacher was rather impressed with it-- he had never programmed in C, so it was a new experience for him to audit my code. After I put on a miniature dog'n'pony show for him, he decided to show it off to the principal and school district superintendent (I lived in rural PA-- the superintendent looked over 3 schools, so it wasn't as big a deal as it seemed).
The day before the principal and superintendent came to see the compiler, my workstation at school crashed, and all my data was lost. The backups we had were more than a month old-- and definitely not functional.
Rather than cancel the presentation, my teacher wrote me a note, excusing me from classes. I was to work on getting the program re-written-- or, failing that, to get some suitable phony in place.
The next day, the superintendent showed up. We ran him through the process of developing a program, then running it through the compiler and using the executable. We showed him a few demo programs, then showed him the output assembly files. He was very impressed that a 9th grade student could do so much. He left happy, and I actually got a neat write-up in the school paper.
To this day, I thank heaven that he never hit Ctrl-Break and stopped that QBASIC mock-up from executing.