Slashdot Mirror

On their web site, if they thank L. Ron Hubbard for something, then they are not a charity! (Although they can apparently fix shattered arms in 45 minutes and raise the dead, their help is usually handing out other people's supplies, Scientology leaflets, "pull my finger" healing, and getting in the way.)

EBMT never really worked very well (it needed millions of translations before it'd start to yield anything useful, and even then it needed hand-holding), but perhaps these new researchers have taken it to the next step.

EBMT never really worked very well (it needed millions of translations before it'd start to yield anything useful, and even then it needed hand-holding), but perhaps these new researchers have taken it to the next step.

Pity Wu went and broke it by whacking a minefield of patents around it. Now hardly anyone wants to touch it and those that are interested are scared off by the other sharks who are circling that pond. So another potentially interesting protocol becomes unusable for 20+ years because of simple greed.

That's what people said about coaxial cable, too...

Adding a new set of wires to the poles behind my house is a trivial expense compared to the cost of digging up my street and laying new water pipes. The pole behind my house has two cables for two companies and there are only limited barriers to adding new cables. Water distribution is a classic example of a natural monopoly. The monopoly exists because of the cost structure of providing the good, not because of government regulations. In fact, the opposite is true: government ownership and/or regulation are designed to prevent the natural monopolist from earning more than the normal profit they would otherwise earn if they had put their capital into a competetive industry instead of the monopoly.

but when government regulations restricted the ability to compete directly between cable companies for individual consumers, a few media companies went to the trouble and expense of launching satelites into geo-stationary orbit for a work-around.

Satellites were not a workaround to cable regulations. They are a technology that drastically reduced the cost of shipping television signals to consumers. One satellite can cover an area that would take (thousands?, millions?, a whole lot!) miles of cable to cover.

We have no way of knowing how much better (or worse) our water service would be if it were open to the free market. I'm making an educated guess that it would be quite a bit better.

That's not an educated guess. It only demonstrates that you never took microeconomics 101.

Hey, if they stop serial cables, why not HDTV tuners?

Sorry, I didn't mean my comment to be a correction, but rather to suggest a related idea. Although the HCII doesn't seem to have any current projects directly related to the headmouse situation, I'm sure there are several faculty members who are relatively well versed in alternate ways of interacting with computers restricted by disabilities. The Gestures Project seems most relevant in current research. If the project succeeds in being able to translate gestures into a computer representable form, then it is conceivable that the same technology can be applied to aiding the disabled in using their computers.

I don't believe that Carnegie Mellon has a medical school, however you might want to contact their Human Computer Interaction Institute. They can probably point you in the right direction.

There are lots of ways to distribute software. And sometimes a very good reason to do so.

Carnegie Mellon University has a bio-medical engineering program, with one caveat. A student can only major in it as long as they have an additional major, like mechanical, chemical or electrical engineering. I think that's a good idea, because as some people have been saying, Bio-medical is still a relatively new program and doesn't have much structure yet. http://www.cmu.edu/bme/

This is an interesting question. The databases you ask about don't have MySQL gotchas, that's for sure. Nor do they have PostgreSQL gotchas. They don't have SQLite gotchas either. Or Oracle gotchas, for that matter. But one thing is sure, trust me, they most certainly do have gotchas of their own. Do you know them? Can you work around them? Will they silently corrupt your data? Will it be easy to migrate your data to other RDBMS without changing your applications? How do they scale? Do they fully support SQL92? SQL99? Can you afford them not to? Are their transactions truly atomic? Is your data always guaranteed to be in a consistent state? Are the operations on your data isolated? Are the transactions durable? What is the developers' relation to the decades of scientific research and engineering experience in the field of relational database management systems? Do they fully understand it? Do they know why you need ACID? Or would they rather tell you that you don't? Those are the questions that you have to answer. When it comes to relational databases, it is always a question of which gotchas are you ready to face. And of course, as I have already written, you will be unable to answer that question without at least some basic understanding of relational algebra, set theory and predicate calculus. Those fields are essential to understand what the relational model is all about.

You can find a description of the maintainability index here.

If you look at the desription, you'll see that the equation was mainly "calibrated" based on a bunch of projects at HP. But fitting such an equation to a handful of self-selected projects doesn't give you any idea of how statistically valid it is.

Furthermore, the maintainability index contains measures that you would expect to go up as software systems become bigger; therefore, it isn't even a meaningful comparison of software systems of different size (or a single software system over time): maintaining a 1MLOC project is just a lot harder than maintaining a 100kloc project, but you may be doing equally well on both of them.

Particularly amusing is a term of 50 * sin (sqrt(2.4 * perCM)) in the maintainability index, where perCM is the average percentage of comment lines per module. As perCM ranges from 0 to 100, the argument for the sin(.) function will range from 0 to 15 (ponder for yourself what that means about how much you should comment).

Until someone produces sound maintainability data with hundreds of software projects, the use of MI is just bullshit.

FYI, the Software Engineering Institute is Carnegie Mellon University's Department-of-Defense-sponsored research center whose "core purpose is to help others make measured improvements in their software engineering capabilities and to develop the right software, delivered defect free, on time and on cost, every time."

CMM is in fact the Capability Maturity Model, from the SEI.
Here's more info.

There are 5 levels. It's damn near impossible to get a level 5. IBM Federal Systems (later Loral) was certified Level 5. They did shuttle avionics. When I worked for a major defense contractor, it was a huge success when we were certified Level 3.

Two things, though:

1) This is hardly a declaration that "Business Considers..."

2) There is a complete confusion of licensing ("open-source") with development practice ("commercially built").

Two things, though:

1) This is hardly a declaration that "Business Considers..."

2) There is a complete confusion of licensing ("open-source") with development practice ("commercially built").

Some of the work at Project LISTEN may come in handy too. Many of their publications are available for download.

Some of the work at Project LISTEN may come in handy too. Many of their publications are available for download.

More likely it's the other way around. Far more projects fail because of management problems than technical. Take a look at the CMM for Software Engineering sometime. Apart from some procedural stuff that any sensible software engineer would want to do anyway, it's all about management.

Really?

Carnegie Mellon has an Entertainment Technology Center which offers a Masters of Entertainment Technology (MET) degree.

The ETC draws upon Carnegie Mellon's particularly unique combination of strengths and its predisposition towards interdisciplinary related activities. The DaVinci Effect if you will.

Carnegie Mellon has an Entertainment Technology Center which offers a Masters of Entertainment Technology (MET) degree.

The ETC draws upon Carnegie Mellon's particularly unique combination of strengths and its predisposition towards interdisciplinary related activities. The DaVinci Effect if you will.

It sounds good and all, but then you come to this line:

Are you a... master code monkey? ... Then the ETC might be the place for you.

Interesting ...

(to be fair, the ellipses include: "superb 3D modeler? master code monkey? stagemanager extraordinaire? talented screenwriter? mesmerizing mechanical engineer? theatrical lighting expert?" )

This effort joing some other projects targeting cheap PCs at users in developing countries. For example, the PCtvt was recently proposed by Raj Reddy at CMU (an academic rivalry?).

But both efforts are predated by the Simputer, a low cost device that was designed to be shared by Indian villagers. Each user stores their data on a Smart Card, which is plugged into a single Simputer as it is shared by various members of the community.

Last time I check, OOO maintainers are playing the 'we don't need stinking grammar checker' politics on this issue. For the few tools that do (eg: Link Grammar, Queerqueq), they never talk with each other.

Also, something like facial recognition needs large test datasets, and it's never a "solved" problem. There's always a way to do it faster or better or more easily. Other things like Canterbury Corpus or Calgary Corpus are datasets used for comparison between compression algorithms. Meaningful comparisons can be made between different algorithms based on how well they perform on them simply because they've been used enough and are standard enough.

I'm so interested in this that I just registered gpldata.com... finally something useful to do with my free time!!!

All the free tts systems sound the same as they did since the early 80s. Because they all use the same algorithms and data generated by the Navy. The nicer sounding ones that have more complete data sets, improved algorithms and are computationally more intensive are only available through special licensing. (the algorithms have multiple patents, the data has copyrights, etc).

Compare a public domain TTS like rsynth to a free, but commercial quality TTS like festival or Bell Lab's. It's funny how rsynth sounds a lot like the mac (although rsynth doesn't have a bunch of predefined settings to do different voices, you have to set all the parameters yourself to make it sound exactly like Bruce).

TTS technology doesn't move terribly fast. the TTS that was in the Mac 21 years ago is basically the same technology 30 years ago. But that's no excuse for Apple not to have moved on to using diphonemes or triphonemes like other systems. Apple is behind, but in the TTS world, 20 years behind is not all that far behind. (unlike say the harddrive world, where 20 years behind is the difference between 100s of gigabytes to 10s of megabytes. ouch)

If you really want to build your own, why not start from a team with open-source software or published robot designs?

RoboRoos - Currently the software release seems to be down, but a friendly email to them would probably fix it.
CMDragons - My team (whee shameless plug). Our complete 2002 software is available (runs on Debian). RedZone robotics is currently selling a robot based on our design.
RoboRoos - These guys have done very well the last couple of years, and have a fair amount of information online as well.

Remember, only the *best* team has to beat humans by 2050. It's a lofty goal, but with something like 150+ teams pre-registered for 2005 (total for 4 leagues), I think we might be able to reach it. I've been doing RoboCup since 1999, and you would not believe the progress just in the last 6 years. I do think its self-centered BS for any particular team to claim *they* will reach the 2050 goal, but I don't think its so bad to say *some* team will.

Humanoids need to advance in hardware, but look at the other robot and simulation leagues and you will see that strategy is coming along quite well already. When I started working on small size, humans with joysticks could beat a RoboCup team. Now humans get demolished. Speed is exactly the computer's advantage -- they are dumb in high-level strategy, but they outdo you in driving and kicking accuracy. Plus they never blink and have a reaction time of 100ms for almost any sort of change. Humans have *at best* 100ms reaction times, and much slower if they get into an unexpected situation. Check out our videos

That's why my team uses Debian :P

Personally we've never really found a use for Realtime extensions; Robots are a pretty soft-realtime application becuase the complex algorithms need to be interruptable for debugging. It's way too easy for vision or path planning to take more than the alloted time.

Plain Linux does pretty well at handling the event loop; Most teams are simply event driven off of vision frames, with the motion control loop handled by custom embedded hardware on the robot.

There are also wackos who request all sorts of information (over and over again) from every FBI office and then sues them because it doesn't match her delusional reality.

One of the more famous examples is Dr. David S. Touretzky's "Gallery of CSS Descramblers", which contains more than 20 different examples of code that is (assumed to be) illegal under the DMCA.

The page also prominently displays Dr. Touretzky's name, email address and a photograph of him. It was explicitly created to draw attention to the absurdity of the DMCA law, through civil disobedience:

If code that can be directly compiled and executed may be suppressed under the DMCA, as Judge Kaplan asserts in his preliminary ruling, but a textual description of the same algorithm may not be suppressed, then where exactly should the line be drawn? This web site was created to explore this issue, and point out the absurdity of Judge Kaplan's position that source code can be legally differentiated from other forms of written expression.

Buncha snot-nosed little whippersnappers, that's what. When I wrote my first image compression program, I had to push the chads out of my paper tape. I had to open a window to let the heat from the processor out. I had to... Look, without Lenna, we can't be sure this is an image algorithm at all. Probably just some genetic code gotten out of hand. Where's my raid, anyway? Oh. It's stuffed full of JPEGs. Including a beautiful hires one of lenna, considerably more than her head. See above link. :)

Have you seen this?

That's hosted in the US BTW. The last I heard the authors of that page were not stuck in a dungeon somewhere awaiting fingernail extraction.

Yes that is a danger, but even there many people post without fear of anything beyond a lawsuit, not physical harm by government goons. At least here it is somewhat more feasible to commit civil disobedience.

This year FIRST has really placed the emphasis on software. They've given us an easy-to-build chassis and gearbox, a game that requires at most one manipulator (for moving tetras), and a boatload of awesome tools. We get a CMUcam2, which lets our robot track things using a camera. (This also offers some interesting possibilities for funny things, like making a cart that follows a student around as they scout teams or something... we're planning on building both the FIRST drivetrain/chassis and our own, and using the FIRST one as a testbed... I want to convert it into a human-following robot cart once we're done using it ;-) They've also apparently written lots of software modules to make it easier to use gyros, position encoders, and the like, and combined them to make a plain-text-scriptable autonomous mode thing, that allows you to write the robot instructions for what to do. (This gives teams with "intelligent" robots an advantage, as more people will take the dead-reckoning route if it's easy and reliable, so smarter bots will face less competition.) Personally I'm a programmer, and our job was usually neglected in previous years, so I look forward a great deal to a season where programming becomes a major portion of our robot, and not some little detail to be filled in at the end.

For those who aren't familiar with the US FIRST Robotics Competiton, here's a quick summary.

Dean Kamen started an organization called For Inspiration and Recognition of Science and Technology (FIRST) because he felt that students were not being inspired to pursue science and engineering. His usual analogy is that while we have immense respect for athletes, celebrities, and entertainers, we don't recognize engineers and scientists in the same way, and he wants to change that.

The practical implementation of this is the FIRST Robotics Competition. Each January, the kickoff from Manchester, NH is broadcast to teams across the country (and world) on NASA TV, and they find out about a new game. They also receive a kit of parts, and they then have six to seven weeks to design, build, program, practice driving, and ship a robot to play that game.

This year's game, as many are, is just complex enough that I will not try to explain it fully. Essentially, you earn points by stacking small tetrahedrons ("tetras") on the large tetra-shaped goals. There are 9 of these in a grid. You get 3 points for each tetra of your color stacked (upright) on top of a goal, and 1 point for each that is inside the goal but not stacked. Then you get 10 points for each row of 3 goals where your color is on top, and you get 10 points at the end if all three robots in your alliance (there are two alliances, red and blue, with three teams each) are in your end zone. You also receive bonus tetras (placed directly on top of the goals on your end of the field) for certain actions during autonomous mode: placing vision tetras (these have a green stripe for the camera to track) on the goals in the middle (1 bonus tetra for putting it on the side goals, 2 for the middle) and knocking down the tetras magnetically hung from the goals on your side (1 bonus tetra, and the knocked-down one stays in play; it otherwise would be removed).

The structure of the match is 15 seconds of autonomous mode, where the robots can't (electronically) receive communications, and must navigate on their own. This is made much more interesting this year by them throwing a CMUcam2 (a small serially-controllable robot vision system--quite cool!) into our bag of sensors. Then the remaining 1:45 of the match is human-controlled. Scoring is probably another "coopertition"-style deal where the winner gets 2x the loser's score or something similar to keep good teams from kicking bad teams' asses completely.

Ever heard of Sherman Austin? Maybe he should have taken more corporate advertising on his site.

It's always fun to see your work linked in a slashdot discussion. Check out LITA for more information. We're still far from a finished product to send to Mars. But, within the next few years, it would be realistic to expect something that can reliably, autonomously scout out unknown areas and return enough information for scientists to come up with detailed plans and give intelligent instructions on how to proceed. Anyone who wants to explore a volcano would love that, and as for driving across the desert, well, that's exactly what we're doing.

Of course, you'll still need a bunch of experts back at mission control, but even if you sent a human to Mars, they'd need the constant advice of a bunch of experts to do much more than the basics. It won't be all that long before we can send a robot with AI that performs close to as well as a small child.

I'd say that the current state of the art could dramatically decrease the number of people needed to control a robot (from several dozen to a handful), with a slight decrease in reliability. This decrease in reliability can be compensated for by have an increased number of relatively inexpensive rovers. Current research could the efficiency of robots even more.

Also, although the environment is certainly a little different from that of Mars, we did have a robot which was able to explore for hundreds of kilometers (mostly autonomously) in the antarctic and Atacama Desert. This was back in the late 90s.

We have not had much success with exploring volcanoes or driving across the desert with robots how can you claim we can make this huge leap.

This is actually an active area of research. Check out the Carnegie Mellon Robotics Institute's Mars Autonomy Project.

As a side note, one of its founders is Kevin Lenzo, of YAPC and Perl Foundation fame.

- Barrie

"2600.com had a similar case where they were ordered not to link to any site that had decss. Never mind the fact that google and pletny of other sites did and still do."

Personally, I believe they lost this case due to attacks on their character, rather than actual laws. 2600 was portrayed as group of crazed anarchists who like to flaunt the system. As a result, the judge had little sympathy for them.

IMO, this is why 2600 is forbidden from even linking to DeCSS, but Felten's Gallery of CSS Descramblers is still up. The MPAA would have a much tougher case fighting a CS professor at a famous university, with the money to win.

If anyone considered the 2600, ruling to be precedent, it seems like they would try to use it to get the gallery shut down. IMO (IANAL), that the gallery itself hasn't been shut down speaks to the stupidity of that ruling.

I think the 2600 case shows how character assasination can win you a case by playing to the judge's prejudices.

The MPAA did not have a good faith of any kind not subjective not objective they did not do their job right they did not even read the C&D the Ranger spit out at them they just sign off on it.

As I've said, the 9th Circuit Court of Appeals disagreed.

Yes they did violate my First Amendment rights it is called abridging the freedom of speech and by making me find a new ISP is abridging my freedom of speech.

Do you care to point to some case law which backs that up?

Marc Brandon from FOX sends an email to the MPAA "If you have not already initiated an Investigation, please review www.internetmovies.com."

Amazing. So you admit that the investigation began by a personal investigation, and yet you still accuse the MPAA of "not even read[ing] the C&D the Ranger [an automated tool to find copyright infringements] spit out at them".

I will let you all be the judge on this matter.

Based on the statement by Debra Shapiro, I think the MPAA acted in good faith. Shapiro might not have been correct when she said "internetmovies.com is offering downloads of trailers and features of member company product [....] They are not licensed by us.", but that alone should have been enough for a cease and desist.

If you had only sent in a proper counter-notification this all could have been avoided. Next time fill out something like this.

I hope in the end the courts change the DMCA where it reads fairly.

It'd be nice if they raised the standard from "good faith" to "reasonable investigation", but that's the job of the lesislators, not the courts.

By the way, here's a Do-It-Yourself Counter Notification Letter. Be sure to use it to mitigate your damages if the **AA ever comes after you for something you didn't do.

FWIW, one of the primary reasons that neither OOo, NeoOffice/J, nor AbiWord have grammar checkers is that no good open source grammar checker tools exist yet. Just about any grammar checker worth its salt is completely closed source. There are some options available, though...

The Link Grammar Parser is one that I've actually been keen on integrating with NeoOffice/J for quite some time. I have ideas on how to do so but have not yet had time to devote to it. I had been waiting for an OSS license for a couple of years for it, but unfortunately I think their new license was incompatible with GPL and can't be used in OOo or NeoOffice/J. Aside from that parser, I don't really know of any other good OSS grammar projects for English, much less all the foreign languages that need to be supported too!

If you have knowledge of any other OSS grammar engines (or contacts for acquiring and freeing source code for a defunct grammar checker like Correct Grammar please contact me!

ed

It is time for software developers to donate potential patents to the FSF. If copyright is no longer a defense against one's code being ripped off by commersial competitors. Microsoft must be secretly wishing that this guy wins his case.

Perhaps, also the Gallery of CSS Descramblers could come is useful. Greg Aharonian's filing is taramount to saying that code is not speech, after all.

If you google the words "universal library" you'll find this link http://www.ul.cs.cmu.edu/html/ at Carnegie Mellon. Why is Google doing something different?

PSP == Personal Software Process

http://www.sei.cmu.edu/tsp/psp.html

or the Personal Software Process ;)

At Carnegie Mellon, considered to be one of the best tech schools in the world, all students (even CS and ECE majors) are required to take Computing Skills Workshop. The very first lesson in the course is on security and passwords. It also covers UNIX commands, file management, and access rights.

After a few weeks, most people realize that they can skip the classes and only show up for exams, so it's not really a waste of time for those who do not need it. However, for those who do, it ensures that they have a baseline level of computing knowledge, which helps keep our network safer.