Slashdot Mirror

Mainly, I find that during my trip I don't want to spend the time writing in my journal during my trip. But I do it anyway because it makes the post-trip memories much better. Sometimes I'm more into writing than other times, and occasionally I'm really into it and trying to write decent prose. But overall I usually write the gist of my travels, and then edit it after my trip.

I did a 2 month road trip before going to grad school, and had a small leather-bound journal that was given to me as a going away present by a coworker at my job. For me, this journal was PERFECT for what I needed, small enough to not be a nuisance, and I could jot down whatever thoughts came through my mind. When in a hurry I'd quickly scribble a few sentences.

Call me old-fashioned, but there's a certain flavor with a handwritten journal that's lacking with electronic notes. For example I can make little sketches, draw doodles, interspersed in my writing. Or I could collect little 'souveniers'. Like receipts from cool places, a cutout of the paper menu from a cool bar, parts of a flyer for a cool music show, etc. My journal is really colorful and fun to read. Plus, there's alot of stuff in the handwriting, you can get a feel of when I'm hurried, when I'm happily lazy, when I'm annoyed, etc.

Anyway, I would imagine that ultimately any recorded journal would need to be edited after a trip. The traveller might get a hint of saving time by recording directly into a PDA, but I think the overall result would be better w/ a handwritten journal, where a few notes can be jotted down quickly without distracting much at all from the actual trip. But that's just my opinion.

In case anyone's interested, here is a link to my still unfinished travel log. The delay comes from my being in grad school and EVERYTHING ELSE having priority to finishing my journal. Ie, the bottleneck is not because I don't have my travel notes in electronic format.

Would you hire an engineer who recieved his degree from Columbia, Johns Hopkins, or Georgia Tech?

What if I told you taht this individual finished his degree in an off-site evening-degree or distance learning program?

We live in an era of cheap paper. You can buy your diploma from just about anywhere. Sure, some quality is there, but it's not the same. Would you really want to hire somebody who studied rocket science at home? Schools don't make engineers; that's why co-ops, internships, and senior projects are important.

Mirrored on an Internet2 site here: ta-da.

I have made a mirror of the movie so you can spare Tom's the bandwidth.

If you find such a project, please let me know. My organization spends a lot of money on commercial math software.

If you find such a project, please let me know. My organization spends a lot of money on commercial math software.

Here

Man, I misread your shortening of "poontang" to "tang" as the Kool-Aid-like powdered drink mix stuff. "Art made out of Tang(tm)? I gotta check this out!" Then all I found were images like this. Damn. Although this one was pretty funny.

Yep, I'm headed straight for that offtopic mod, too.

Umm... breaking encryption is cool! Especially if you get paid for it!

Man, I misread your shortening of "poontang" to "tang" as the Kool-Aid-like powdered drink mix stuff. "Art made out of Tang(tm)? I gotta check this out!" Then all I found were images like this. Damn. Although this one was pretty funny.

Yep, I'm headed straight for that offtopic mod, too.

Umm... breaking encryption is cool! Especially if you get paid for it!

Re: your sig.

I followed the link, saw the desert eagle wood carving and thought "cool -- nice work!"

Then I went up a few dirctories (you may want to check your .htaccess and.or apache config files) and saw all the effort you put into many "happy birthday ${female_name}" images.

They're very pretty, but I wonder: did you ever get any tang out of those, or was it all done "for art's sake"?

I only ask because I saw a few "Happy Valentine's Day" images as well, but none of those had names.

This is more like the railroads back in the 1800's. Railroads are now required to carry others' cars, in fact they pay rent for the cars on their track.

At one time there were hundreds of small railroads, many of which owned only a single line from PointA to PointB. Their anticompetitive tactics often included predatory pricing, delay of competitors' shipments and outright refusal to carry a competitor's engines or cars. Companies with a monopoly on routes important to others sometimes charged 10X the going rate, "because they could". For another example, railroads commonly charged farmers exorbitant rates to store and carry their grain because they had a local monopoly on storage and transportation.

According to this, this problem was the impetus for the original Interstate Commerce Act of 1887 (ICA), and also the original Sherman Anti-Trust Act of 1890. The latter act is also the basis of the federal actions against Microsoft, BTW. Some other interesting links are available here. This article briefly shows how the development of the railroad system beginning in the early 1800's was a major factor in creating the entire US legal structure we now take for granted, including racial integration, worker protection, eminent domain, elimination of 'blue laws', liability/tort law and public/private partnerships.

The ICA established the Interstate Commerce Commission (ICC), which was responsible to set standard rates for railroad passage, and required every railroad to carry others' cars at a reasonable cost.

At present each railroad pays rent for each car that is recorded as being on its tracks to the car owner, but can use it to earn revenue from a shipper. Thus every railroad is motivated to get the car filled with something and send it off to another place, to either earn money with it or stop the rent accruing. Many cars are now owned by individuals or holding companies, who buy the car and send it out on the track in hopes of receiving rent.

thanks to all those who have offered to mirror/host my site. i'm currently working on a solution so i should be back up again soon. thanks to Asheesh Laroia there is now a temporary mirror here: http://jhunix.hcf.jhu.edu/~alaroia1/dean/ please note: only the html test files work on this mirror. thanks again. dean edwards

I made this mirror based on the Google cache. It has the full source code, as well as the docs he wrote.

This is temporary, of course.

Nope. It's the programmers and info security folks who are the most worried about E-voting, because we know how easy it is to mess with the back end of a computerized system. It's the much less technical politicians and election administrators who wanted Diebold, over our objections.

srytch wrote:
>
> Total lack of ad-hoc polymorphism. This language has such a problem with overloading,
> you need a different operator for adding floats than you do for adding ints.

This is a Good Thing (TM). That means that your language isn't going to do implicit conversions from floats to ints (or vice-versa) behind your back. This was an explicit design decision, due to the recognition that the programmer is smarter than the compiler. You are the programmer so you should decide how types are converted from one type to the other, not the compiler. Of course, if you want, and only if you want, you can easily create a composite type and a library of functions that will do all of the conversions for you. Curiously no one has seen the need to implement such a library.

> Lack of compiler warnings. My all time favorite error in ocaml goes like
> "The value is declared type Foo, but is used here with type Foo".

This is covered as the first question in the OCaml FAQ, and personally I have run in to that situation exactly once in my entire experience programming OCaml. So I consider this particular error message a relatively rare corner case... certainly nothing to condemn a whole language over.

> As a better C, ocaml looks great. But it fails to offer many credible alternatives to what
> even ugly old C++ offers in ad- hoc and parametric polymorphism.

Don't even get me started on C++. Instead, start by reading this. C++ is not even in the same league as your average functional programming language, much less OCaml.

Second, ocaml has parametric polymorphism. Read this. And this... google for more. And here's an example of OCaml's parametric polymorphism:

# let f x = x ;;
val f : 'a -> 'a = <fun>

This creates a parametrically polymorphic function, named "f". And here is how you would use it:

# f 3.14159 ;;
- : float = 3.14159
# f "0wn3d" ;;
- : string = "0wn3d"

The problem is that we're coming from vastly different points of view; so different that it seems we're talking around each other. So I'll start making a bridge. I ought to say I'm sorry for calling you an idiot, but I won't, since it seems you've brought your best thinking to the debate, at least once you got done whining. You think I didn't read your post at all when in fact I spent a long time reading it and working on my reply (notwithstanding the diversion into how many binary numbers are in fact evenly distributed).

You took my argument the wrong way; I have no intention of attempting to map the tortured turnings of your byzantine mind. I was trying to get you to accept that there are random events in the world. You went off and started talking about Laplace's Demon. My tactic was to make you see what a repugnant idea your radical determinism is, but that failed. I'll just put it to you straight. Determinism is dead. Heisenburg's Uncertainly Principle shot it and Godel's Incompleteness Theorem buried it. Einstein might not like it, but there are events in this world that cannot and will not succomb to the mechanistic tyrrany of a Turing Machine. God does play dice.

So skip the coin flip, quantum physics tells us that there are random events. It is not possible to calculate or predict alpha particle emissions.

That's the corner of the debate. You're not an idiot, you're right, it's not very useful in some situations to have a random number generator hooked up as input for a scientific experiment, because you can then never repeat the results. An LCG with a "properly chosen seed" is the correct tool to generate "random" input, when reproducibility of results matters. What you're not seeing is that there are situations, such as cryptography, where repetition is NOT desired, and so truely random numbers are needed and in fact do exist.

Where we actually differ is that you see that "properly chosen seed" as merely the output of a complex system, and I see true unpredictability.

If you'd like a random number of your very own, try HotBits which uses radioactive decay as its source of entropy, or if you prefer something cooler, try Random.org who can serve up random bits in whatever package you like, even via CORBA! Now, if you'll excuse me, it's time for me to take my completely predictable dog out for a walk.

jaz

if you're going to copy and paste, the least you could do is credit the source: http://www.ugrad.cs.jhu.edu/~russell/classes/enigm a/how.html

I thought this was kinda cool, so I looked around for a java applet and found one: Its pretty cool.

Here's a mirror, folks.

People, when you mirror things for Slashdot, your home cable modem probably won't work very well....

I have made a mirror of the page, as it is becoming exceedingly slow.

At 8:00 a.m. EST, my Mom IM'd me at school to tell me that the Internet as slow at home. At 9:00, my friend who uses my wiki told me that he had been letting it load for about an hour and it wasn't loading. The Xbox-Linux project's wiki also is hosted on my server, and it was inaccessible.

All these web sites are hosted on my little Linux box in my basement, on my parents' cable-modem with 40 KB/s up to the Internet. What happened to me was that the Google logo, linking to the image search for "julia fractals", had my friend Jonathan's site as the top hit. (The exact hit was this page.)

The page was all-but inaccessible, as was my server. I eventually SSH'd in, copied the files to my JHU web hosting space, and set up an Apache redirect to serve the files from there. JHU (my university) has a pretty big pipe, I've learned over the years I've been here. :-).

I mentioned this in a blog entry I made on the topic. It seems that now the search finds some other first hit, the .edu.au site mentioned in the story. Perhaps that's because my server was "Cthuugled" (eaten alive by Google, that is), and no one could reach the first link for so many hours.

At 8:00 a.m. EST, my Mom IM'd me at school to tell me that the Internet as slow at home. At 9:00, my friend who uses my wiki told me that he had been letting it load for about an hour and it wasn't loading. The Xbox-Linux project's wiki also is hosted on my server, and it was inaccessible.

All these web sites are hosted on my little Linux box in my basement, on my parents' cable-modem with 40 KB/s up to the Internet. What happened to me was that the Google logo, linking to the image search for "julia fractals", had my friend Jonathan's site as the top hit. (The exact hit was this page.)

The page was all-but inaccessible, as was my server. I eventually SSH'd in, copied the files to my JHU web hosting space, and set up an Apache redirect to serve the files from there. JHU (my university) has a pretty big pipe, I've learned over the years I've been here. :-).

I mentioned this in a blog entry I made on the topic. It seems that now the search finds some other first hit, the .edu.au site mentioned in the story. Perhaps that's because my server was "Cthuugled" (eaten alive by Google, that is), and no one could reach the first link for so many hours.

Like, for example, this idiot. If he'd just been born with an overly common name like mine he would have learned to get over it ages ago.

But some people just take this bizarre pride in knowing that their name is theirs alone.

The original is already slashdotted, so I have created a mirror:

http://myweb.jhu.edu/bananas/haiku.html

SUV's have an "interesting" dynamic, which is that they're not safer for you to drive (except when you hit another SUV), but they are more dangerous to the rest of the people on the road. So while I can imagine some situations where an SUV would be useful (imagine having a bunch of kids with hockey gear, or having a career that involves moving a lot of equipment), but for the most part, I think that SUV drivers are somehat anti-social -- they drive vehicles that are huge, expensive, consume absurd resources, and are great at killing other drivers in accidents. Specifically, in accidents, SUV drivers die a little more often than non-SUV drivers (because they don't wear seatbelts more often for some reason), and kill non-SUV's that they collide with. So while people are "sold" SUV's as being safe, the statistics don't support that. I'm not too sure how this applies to MP3 players, though.

It tells you that Microsoft spent millions of dollars producing documentation that shows that Windows 2000 meets an inadequate set of requirements, and that you can have reasonably strong confidence that this is the case.

Intersting Document on EL

Obviously this is a developing field. The best models seem to use phrases from the original text, anyway the Mac OSX example above shows that it is useful to users willing to take it with a massive grain of salt, even if we are not into full computational sentience yet.

When it works even a little better it will replace all those awful grade school teachers who assign paraphrasing as a homework assignment. The reporters who might have been replaced by it will have already lost their jobs, except for the ones in AhaIndia of course who will paraphrase for the rest of us, usually at a marginally better level than the machine.

The research is interesting - and I'd like to understand Barzilay's notation is that APL or calculus of statement? - in the paper (pdf) I found on google. Also see the papers on her site.

Of course structured text is easier, and news stories are known to have most of the meat in the beginning, but this is great stuff.

One interesting older system is ThoughtTreasure which was built to understand a story and answer questions about it. The author also did work on news analysis ("NewsForms") too. There are tools out there, I've been making a survey myself too. If anyone has information about practical NLP tools for real world tasks please post.

On your specific points:

• Agreed that NT has access controls on every object. However they are not visible and not used very much by end users and administrators. The UNIX ones are simple and very easy to understand. Here you have the choice between complicated (you do know the difference between discretionary and inherited rights filters?) and pervasive (every object) versus simple and pretty much only on files (which almost every OS object is anyway).

  Many (if not most) Windows programs get it wrong. Heck even Microsoft has been released games that can only be played if logged in as administrator.

  Linux does let you do delegation, but that is mostly left as a user space implementation issue. That is the purpose of setuid/setgid, group memberships, sudo etc.

• The Windows acceditation is a crock. It is in a non-networked environment with no floppy disk or CD drive. Show me anyone who deploys that way. Here are some relevant articles: Win2K evaluation IBM/Suse evaluation. I have one specific question: if the Windows architecture is so fantastic, why did the NSA choose Linux to acheive their goals? Why did Microsoft claim that fundamental design flaws in Windows were the reason they couldn't release the Windows code? (And we won't even go into the ability of any process in a desktop session being able to send messages to any other process which is probably the flaw Microsoft alludes to).
• And you deploy Microsoft patches immediately without worrying that they will break the other products you run and use? You can get Linux advisories from whatever distro you use. There are also services like CVE. At least with Linux you can choose to fix things yourself. With Microsoft, you are stuck with whatever amount of time and problem severity they determine. If they don't want to fix something for 6 months, there is nothing you can do about it.
• SCE is nice, but is only needed because the whole OS has so many places where ACLs are applied. And it doesn't do things like registry access control (you have to use regedit) or the filesystem. So you do have to use a number of tools, and understand everything. In Linux you have to understand chmod. In either case, a clueless admin will do way more harm than the OS you picked to run.

Hi, I did a two month cross-country trip in the summer of 2000, and you can see part of my unfinished journal . I guarantee you that you'll pick up tons local flavor just about anywhere you go, not just through radio but through cuisine, live music, scenery, and just talking to locals. Here are some tips, in case you or anybody else is interested.

Firstly, to have the best experience, avoid interstates as much as possible. Obviously if you need to get somewhere fast, the interstate is essential. But travelling on the US or state highways will bring you into contact with literally orders-of-magnitude more local culture.

I drove to the West coast (from Jersey) and about 95% of the drive westward was off the interstate. It was totally awesome. The scenery of smaller roads is usually way better. And you pass lots of farms and produce stands, find little country kitchens to get good homecooked food, small bars and honky-tonks to hang out in, etc. And in the smaller towns it's much easier to meet people and have a good chat in the small bars too. Before my trip I lived in NJ, Boston, and Philly, and found it far easier to talk to folks in bars outside the cities. I then drove back to the East Coast in a few days on the interstate, but that was because I had to be back in Baltimore for my graduate school starting date.

I'd also suggest trying to keep your trip as flexible as possible. Ie, make it up as you go. That way you can stay places you like, leave those you don't, etc. The more deadlines you place on yourself, the harder it is.

And finally, here's a tip I didn't find out until about halfway through my trip. You can sleep in national forests for free. I forget the details, but if you're something like 200 feet off the road, you can stay there for 2 weeks. Much much easier on the wallet that way than paying $5 to $15 for small campgrounds. Although it is kind of freaky being in the middle of the woods in the middle of nowhere at night by yourself. But also really cool too. And you can shower at truckstops, though these are usually found along the larger roads, for 2 or 3 bucks.

Let's see, to post something roughly on topic, there were parts of the trip where no radio is to be picked up for miles. Ie, in the deserts of the southwest and some parts of the deep South, etc. Satellite radio would be great for truckers and other folks driving through these ares, though I didn't have it. If you're driving alot, sometimes it's nice to hear a real human talking live, it gives some form of interaction, even if it's one way. I did pick up alot of tapes on my way from various truck stops.

The other thing is that you'll also find alot of Christian radio stations the further you are from big cities. If you're not religious, they can actually kind of interesting/amusing for short times, depending what they're doing. And always look for radio stations in the lower part of the spectrum, ie around 88, 89, and 90. This is where most of the college stations lie, and they definitely play the best sort of stuff of all genres.

Anyway, you'll have a great time. The most important thing to pick up the local culture try to avoid the interstate as much as possible. You'll find out that American culture (ie, Americana) truly does exist, and you can feel different flavors of it as you distinguish between New England, southern Appalachia (Dixie), Cajun Bayou country, the Great Plains and "Wild" West, the Southwest and Mojave, the West Coast, the Rockies, etc.

RHEL is getting certified at EAL2, which is really weak.

Even the Windows 2000 EAL4 certification only protects against "inadvertent or casual attempts to breach the system security." No real security here. For more info, read Jonathan Shapiro's article.

> $37,600 in US vs $26,600 in Germany does mean something.

No, it doesn't. Once you remove the few percent of top earners in the US from the calculations, the figures look very different. The US has much more peaks at the top end than Germany, which inflates simple per-capita income. Face it, Bill Gates' or Warren Buffett's income doesn't benefit you personally one bit--you neither travel more nor have more leisure time because of it. There are figures available that normalize the data in various ways.

There are also other things that skew the figures. An interesting article on the topic appeared in The Economist ealier this year. But since it's premium content now, I've googled for it and found a copy elsewhere.

The whole system seems a little more responsive, although with everything sitting on a Mach kernel I don't think MacOS X will ever achieve the low latency that Linux pulls off. Mach's cool but you pay a price.

The unresponsiveness was not due to the fact that they use a kernel based on Mach, but simply due to the fact that the GUI wasn't optimised very well. In Panther, they added tons of new special-purpose functions which are much faster than the general-purpose routines. You just have to take care the conditions for calling them are fulfilled.

Even now, there's still a lot more GUI processing going on in the Mac OS X window manager than in most (all?) XFree Window managers. I think your remark would be more appropriate if it said "The whole system seems a little more responsive, although with the whole GUI being based on pdf and vector graphics I don't think Mac OS X will ever be as responsive as bitmapped systems such as Mac OS 9 and current XFree and Windows versions".

And even that may prove to be false in the future, as until now the GUI has become more responsive with each version and Apple keeps telling its developers that performance is one of their primary goals. Also, giving the front-most application precedence for screen updates in the window manager/server has little to do with the kernel or pre-emption, but is more of a design choice.

They are also doing this thing called "prebinding" which I assume is equivalent to "prelinking" in the Linux world --

performing dynamic linking a single time and saving the intermediate results so that applications can launch faster. If you look through the installation logs for Panther you see that it includes a new dynamic linker and there are many log messages of the ilk: "Prebinding xxx application."

If you look at the process list in top or with ps you see that there are FAR fewer system processes than before. I'm not sure whether this is because they really aren't running, or if the OS is somehow hiding them (which would be very un-UNIX-like).

Dude, I know you're joking, but one of the professors here in the physics department at Johns Hopkins has actually published a paper with Isaac Newton as the co-author! The paper is in the American Journal of Physics, which is a journal published by the American Association of Physics Teachers. This isn't much of a research journal, but they still allowed the joint-authorship with Isaac Newton.

It's utterly ridiculous. I showed this to some faculty members here, and they were convinced that the "I. Newton" co-author (no institution was listed with him, it just said 'deceased') had to be somebody else. Further detective work in some of the cited articles proved this "I. newton" was THE Isaac Newton of 300+ years ago.

It's ridiculous that someone could put his name as a co-author, and perhaps more ridiculous that the co-authorship was accepted by the journal. Students here joke around that we're all now going to co-auther papers with Einstein, Schrodinger, Heisenberg, you name it.

Anyway, the point is that being dead might not necessarily stop you from being included in other publications, even if you have never even met the person.

Dude, I know you're joking, but one of the professors here in the physics department at Johns Hopkins has actually published a paper with Isaac Newton as the co-author! The paper is in the American Journal of Physics, which is a journal published by the American Association of Physics Teachers. This isn't much of a research journal, but they still allowed the joint-authorship with Isaac Newton.

It's utterly ridiculous. I showed this to some faculty members here, and they were convinced that the "I. Newton" co-author (no institution was listed with him, it just said 'deceased') had to be somebody else. Further detective work in some of the cited articles proved this "I. newton" was THE Isaac Newton of 300+ years ago.

It's ridiculous that someone could put his name as a co-author, and perhaps more ridiculous that the co-authorship was accepted by the journal. Students here joke around that we're all now going to co-auther papers with Einstein, Schrodinger, Heisenberg, you name it.

Anyway, the point is that being dead might not necessarily stop you from being included in other publications, even if you have never even met the person.

The article points out one fundamental observation about the spam problem - people keep buying stuff from spam emails. I would like to suggest a somewhat controversial solution that nonetheless seems like it would be effective. I for one think that it is extremely difficult if not impossible to find a solution that will keep spam out. But we can stop spam forever if we destroy its market. ISP's have the ability (though maybe not the right) to discourage their users from responding to spam messages. If the return rate of users to spam drops by a few orders of magnitude, even the extremely marginal cost of sending email can destroy spammers' profit margins. I have a brief writeup proposing a policy and enforcement mechanism that is not without its concerns, if anyone is interested. spam.pdf

Quantitative Physiology is taught by Denny Freeman who took over the course from his advisor Tom Weiss who wrote an excellent two volume textbook on 'Cellular Biophysics'. The first volume is on 'Transport' and the second is on 'Electrical' properties. The text takes an approach which is based on the physical interpretation of the time evolution of the solutions to the governing differential equations (good stuff for EECS types who are interested in Signals & Systems).

The course stems from a nearly 50 year research program in 'Auditory Physiology' at the Research Laboratory of Electronics. The 'engineers' who were trained at RLE learned to apply the quanitative methods of engineering and the physical sciences to biological problems. Inside of RLE, both Tom and Denny where trained in the inter-institutional Eaton-Peabody Laboratory, which served as the training ground of for dozens of well known researchers (including the founder of the Biomedical Engineering Department at Johns Hopkins). The modern version of this training philosophy is characterized by the HST Speech and Hearing Bioscience and Technology Program.

Quantitative Physiology is taught by Denny Freeman who took over the course from his advisor Tom Weiss who wrote an excellent two volume textbook on 'Cellular Biophysics'. The first volume is on 'Transport' and the second is on 'Electrical' properties. The text takes an approach which is based on the physical interpretation of the time evolution of the solutions to the governing differential equations (good stuff for EECS types who are interested in Signals & Systems).

The course stems from a nearly 50 year research program in 'Auditory Physiology' at the Research Laboratory of Electronics. The 'engineers' who were trained at RLE learned to apply the quanitative methods of engineering and the physical sciences to biological problems. Inside of RLE, both Tom and Denny where trained in the inter-institutional Eaton-Peabody Laboratory, which served as the training ground of for dozens of well known researchers (including the founder of the Biomedical Engineering Department at Johns Hopkins). The modern version of this training philosophy is characterized by the HST Speech and Hearing Bioscience and Technology Program.

Quantitative Physiology is taught by Denny Freeman who took over the course from his advisor Tom Weiss who wrote an excellent two volume textbook on 'Cellular Biophysics'. The first volume is on 'Transport' and the second is on 'Electrical' properties. The text takes an approach which is based on the physical interpretation of the time evolution of the solutions to the governing differential equations (good stuff for EECS types who are interested in Signals & Systems).

The course stems from a nearly 50 year research program in 'Auditory Physiology' at the Research Laboratory of Electronics. The 'engineers' who were trained at RLE learned to apply the quanitative methods of engineering and the physical sciences to biological problems. Inside of RLE, both Tom and Denny where trained in the inter-institutional Eaton-Peabody Laboratory, which served as the training ground of for dozens of well known researchers (including the founder of the Biomedical Engineering Department at Johns Hopkins). The modern version of this training philosophy is characterized by the HST Speech and Hearing Bioscience and Technology Program.

Funny that you mention Python twice but the OP never mentioned it once.

FYI, python "handles failure" quite nicely, and it does it in the manner that it should. Namely, it refuses to guess when faced with ambiguity, and it propigates exceptions nicely (and more easily at the code level than does Java, btw).

FYI, it is not "impossible to do shared memory". Fact is, python does shared memory aplenty.

You sound like the programmers I've encountered that know one or two languages, and subsequently feel threatened by others. My suggestion to you is to look past the lowly forrest in which you find yourself.

Okay, on this map I am certain its now Krieger (#27), which is the home of the Math dept, since the Physics dept moved to Bloomberg (h00j lab space in the basement!). And this was also the home of Homewood Academic Computing lab (though the basement of Garland seems to be the center of some computer operations... but jhunix was housed in Whitehead hall.

I remember during rainy days, the trick was to get from one side of campus to the other without getting wet- you could use the tunnel from DUnning to Remsen, then move to Mergenthaler, Gilman, Ames, Krieger, to Maryland. No quick path to Barton (without using the steam tunnels).

They even have the same primary author, James Clark, producer of the ubiquitous expat XML parser.

DSSSL is Scheme-based, and Scheme is a rarefied dialect of LISP.

Thus turns the circle.

Security experts have been saying for years that the security of the Windows family of products is hopelessly inadequate. Now there is a rigorous government certification confirming this.

(Originally taken from rec.humor.funny).

I share your desire for a "UNIX Philosopy meets GUI" future. However, I lack your enthusiasm for these particular projects.

There are already some small and versitile commands for X11. For example, I use:
XLoadtime
XLassie
dclock

All that you really need to integrate these small tools into your desktop is a panel widget that supports swallowing other X11 apps. Sadly, support for that has been dropped from GNOME and KDE long ago in favor of their own proprietary "Applet" extensions.

If Win2k gets a higher rating than Linux, then why do we have stuff like this happening?

Isn't it odd that a "comprehensive security rating" can overlook something as serious as a complete remote compromise?

EAL4 - methodically designed, tested and reviewed

EAL4 permits a developer to maximize assurance gained from positive security engineering based on good commercial development practices. Although rigorous, these practices do not require substantial specialist knowledge, skills, and other resources. EAL4 is the highest level at which it is likely to be economically feasible to retrofit to an existing product line. It is applicable in those circumstances where developers or users require a moderate to high level of independently assured security in conventional commodity TOEs, and are prepared to incur additional security-specific engineering costs.

An EAL4 evaluation provides an analysis supported by the low-level design of the modules of the TOE, and a subset of the implementation. Testing is supported by an independent search for vulnerabilities. Development controls are supported by a life-cycle model, identification of tools, and automated configuration management.

Jonathan Shapiro wrote a great article analyzing the Windows Common Criteria certification; much of it applies to Linux as well. Among other things, it explains why Windows can get certified even with its remote root exploits: "An EAL4 rating means that you did a lot of paperwork related to the software process, but says absolutely nothing about the quality of the software itself. There are no quantifiable measurements made of the software, and essentially none of the code is inspected."

While this has implications for government acceptance of Linux, which is good, it turns out that it wasn't all that significant when W2K achieved it, and means even less that a system running linux got EAL2. It's probably most interesting that it was an IBM system running SUSE system, not RedHat.

if you're curious about some of the history of microsoft and the certication of windows for government work, click here, and look elsewhere for the story of ed curry. its been linked to here on slashdot before.

if you want to know more about what the eal4 certification that windows 2000 sp3 currently has, click here.

I can imagine some distributions of this translation system that take this code - with improvements - and precook large corpuses to create translators. Anyone want to write the Mozilla and OpenOffice plug-ins for the new menu item "Edit/Translate Language"?

From the article: The original work along these lines dates back to the late 1980s and early 1990s and was done by Peter F. Brown and his colleagues at IBM's Watson Research Center.

IBM's pioneering work was written up in a student-friendly workbook available online. Feel free to try coding it and see how well you do. Do remember though, the state of the art has progressed a lot since IBM's work. This workbook only covers the basics.

You will find that debugging statistical translation system is really hard. You can write test cases, but they take one hour to run each time. You can look at the result of your test cases, but since you cannot work the answer out by hand, you can never by sure if the numbers you are computing are correct. As an example of how tricky it can get, in Brown university's cs241 last fall, amongts the four teams, only two teams managed to correctly implement Model-3, and the workbook goes up to Model-5.

There are two reason why a three way translation is a bad idea. First, it is already difficult to find large amounts of text translated two-way and available in digital format. Restricting your approach to three-way translated text would reduce the amount of text you could train on so much, it would offset the advantage you would get from the three-way text.

Second, training for statistical translation is really expensive. If running one single test case can take an hour, running a full training can take a whole week. Under these conditions, you are always very careful how you spend your cpu cycles. Until better cpus come along, training three-way and cross referencing each language with the other could well take a month of processing (or two).