Slashdot Mirror

Funny, I just ordered a few days ago a gel sheet with a known frequency response just to test if I can learn to see more than three primaries if I filter one of my eyes. (I'm not color blind.) Turns out such sheets are manufactured for film shooting purposes, and that they are not expensive - the one I ordered was a sheet of IIRC 21" x 24" and cost $7 + postage.

Here's a plot of the human cone frequency responses both without the filter (colored dashed lines) and with the filter (colored non-dashed lines, scaled to 1.0): http://www.niksula.hut.fi/~sliedes/cones.png. The dashed black line is the frequency response of the filter sheet I ordered.

The sheet is something called "Seven Eighths Digital LED C.T.O." from Lee Filters; they have frequency response curves for all their filters. For example, see this one's curves.

I wrote a python+numpy+matplotlib script to extract the curve parameters from the PNGs they have on their site and to compute and plot the modified frequency responses. I can share the code if someone is interested. It would be interesting to iterate through all the filters they have and figure out which one would be good for changing the frequency response curves a lot without eating too much of all the light.

In my college (Helsinki University of Technology), barely anybody attended lectures. A 500-page book can teach a topic much more clearly than most professors. Seminars you had to attend to but even there it was the reading material that contained all the information.

I was very happy with my school and felt like I was learning very efficiently. The professors had done a great job selecting the course books.

2:3(69)-3(91):FOR: for (int i=1;i<=10;i++)
3:3(78)-3(80):VARIABLE_DECLARATION:i=1
4:3(88)-3(90):ASSIGNMENT:i++
5:3(93)-3(93):BLOCK:{
6:4(103)-4(123):METHOD_INVOCATION:System.out.println(i)
7:5(130)-5(130):BLOCK_END:}
8:5(130)-5(130):FOR_END:}
The comparison is then done on the tokens (the part in all capitals), ignoring the exact names and such used.

Using Plaggie in the same configuration I usually use, and assuming the code surrounding this loop is different enough, the answer is "no". Typical programming plagiarism detection tools compare entire source code files against each other to determine the extent of similarity, not small chunks. Naturally, you don't want students to circumvent the tool simply by reordering methods or such, so you look for chunks of code that are similar, but the verdict is based on how much of the code can be accounted for this way, not just whether a match is found.

Of course, responsible users of tools like these check anything they flag manually.

In addition to Oxyrhynchus, significant finds have been made at Herculaneum and Pompeii.

If those ones take your fancy more than the ones from Oxyrhynchos -- and there are some good reasons why they might -- you might find it useful to have these links at your disposal:

• Oxyrhynchos papyri site (Oxford) -- here's some info on the imaging process, but I think it's rather out-of-date and only covers basic photography in the visible spectrum
• more up-to-date info on more advanced imaging techniques, with regard to papyri from Bubastos
• the Philodemus Project, dedicated to the most important ancient author to be discovered from carbonised books found at Herculaneum

For texts, the Big Two sites are Oxyrhynchos and Herculaneum (though, IIRC, the idea of using multispectral imaging for damaged manuscripts was first got from trying to decipher the Dead Sea scrolls).

What's distinctive about Herculaneum is the finding of the works of the philosopher Philodemos, as noted above. Editions have started to appear in the last two decades; I think there's at least one translation available. Oxyrhynchos is overall much more important, though. Oxyrhynchos doesn't have a Philodemos, but that's more than compensated for by the sheer quantity of papyri -- in the first century of publication only about 1-5% have been edited and published so far, and that isn't because they've been slacking off. No complete literary works have emerged from Oxyrhynchos -- but we do have gajillions of letters to a relative who lives in the next town over, contracts, land deeds, shipping lists, shopping lists; but also a few bits of literary stuff -- tiny bits of lost plays, about a thousand lines of an otherwise lost epic called the Catalogue of Women, heaps of pieces of texts of which we already had complete copies, and other odds and ends. And yes, in response to the sibling post, ancient porn too. (Well, I know of one sex manual by Philainis, at least.)

It shouldn't have just been denied an oral presentation, it should have been caught by the program committee and never reviewed. You can't read 3 sentences of that abstract without knowing that it's garbage.

Presumably someone DID review this and deny it an oral, but didn't follow up with the program committee to make sure it was pulled entirely.

I've never been to a conference which pity accepts papers. CVPR, a IEEE conference on computer vision, has a 25% acceptance rate for posters. I think this paper is quite an embarrassment to IEEE.

Meh. As others have noted, it was for a poster session. This conference isn't in my field, but at the conferences I've been to in my field (astronomy and astrophysics), pretty much anything gets accepted for poster sessions. At AAS meetings, I've seen particularly wacky posters in extragalactic astronomy and cosmology all clumped together in a kind of ghetto; and back when I was a grad student, during free time between oral sessions or at the end of the day, someone among my friends and I would say "hey, let's go look at the crazy stuff" and we'd take a look at the posters about space potatoes or the Plutonium Atom Totality or whatever. On occasion, I've even seen oral sessions -- typically one of the last ones on the last day -- devoted to something like "Speculative Ideas in Cosmology" with some of the nuttier talks tucked in there (as well as ones which are almost certainly wrong, but aren't in the same league of crazy as we're talking about, unfortunately).

At my university, the student union would recruit a course reporter from every course. The course reporter gets a token reward for their troubles. Their task apparently used to be to take quality lecture notes, which the student union would sell to students at the cost of copying.

However, in my time, the professors, concerned with the quality of the notes, would typically have prepared the lecture notes themselves. They just handed their notes to the course reporter, who carried them to the student union to be copied and collected their reward (some free lecture notes).

Always when someone uses the "nanotech" buzzword, I'm reminded of a study (from Helsinki Univ of Tech) that nanotech isn't a field of technology. It's just a marketing trick. When you actually dig up the patents, social networks and case studies from corporations, the conclusion emerges that "nanotech" is consists of four different fields of technology that don't "talk to each other". They are measurement instrumentation, materials, pharma/chemicals and semiconductors. For example, a pharmaceutical chemist doesn't talk to the semiconductor physicist. Only instrumentation is actually applied in all fields.

"Advanced materials" doesn't sound as cool as "nanotechnology".

The only one talking bullshit is you. If you had some knowledge about the world, about industries, about business and the overall size of development you wouldn't been so quick to call something bullshit.

To make my case, please in example look to Finnish Association of Graduate Engineers. They have 40000 graduated engineers as their members and this represents only a part of all engineers in Finland. Wikipedia tells that there are 70000 engineers in Finland practising their profession. Every year 7500 student graduate and become engineers. (Referens in second page of the PDF) If you now think "gee, where do they need so much engineers", well the answer is for the service of industries mainly to telecommunication, electronics, software, electricity, machine building etc.. There is just so huge amount industrial production going and the amount of engineers required to work is huge. Multinationals like Nokia, Kone, UPM, Stora-Enso, Wärtsilä, Fortum, ABB, TietoEnator, EADS etc.. and also great number of small and medium sized firms employ lots of engineers.

Going back to Iraq, okay 100000 is a good number for developing/third world country, but not comparable with developed countries. And as there isn't more data about Iraqi collages and universities I won't touch that subject more, only saying that in all Arab countries the university system is very weak / low. As what comes to Iraq being the most industrialized Arab country, that unfortunately doesn't still make much of it as when you compare it to other then developing countries as South-Korea, Taiwan, Thailand, Indonesia and Malaysia it paled in comparison.

And what goes to reason for war, yes the oil was the motivator and I'm not denying it. I only said that first Gulf war happened because of Iraq invaded Kuwait and threatened to invade Saudi-Arabia. The second war, as I said happened as the time was right to try to replace Saddam and hopefully avoid civil war. What comes to Israel, they are not a reason to go war, oil is. Besides Israel has it's own quite large amount of nuclear weapons, so having an Arab state industrializing doesn't make it a threat to Israel. Only if an Arab state would have an nuclear weapon, then that would be threat to Israel. Even so, that wouldn't be a reason to go war. As you said and as I believe too, oil and control of it and control currency on which oil is traded are reasons to go war. Thought not morally right.

cplay is pretty cool for a music player. Browse dirs, create a playlist, and play/pause/stop/ff/rew , prev/next/first/last all your tracks from a curses ui. It's a 1Kloc python script and uses a couple auxiliary programs. Who needs a 10Mb GUI to play a few tunes?

Actually, I looked into the issue a while ago of connecting phone to the PC's audio interface and found that there's a class of devices used in broadcasting called telephone couplers (or hybrid couplers) that are intended for addressing some of the problems of attaching telephone lines to other audio circuts (namely the line voltage matching, echo cancellation/duplex issues, and also take the 8KHz upper frequency limit of conventional POTS specs out of the equation)

I was looking for this last aspect, defeating the 8KHz filter, essentially, to record from phone lines so I could do remote live musical performances. the thing is, these things can be pricey... I found some that were as little as $100-200 and ones as much as $500-1000.... and you generally need at least two.... one on each end. I gave up based on price.

however here are some links I found during my research, including some DIY info.

http://www.sagebrush.com/phontech.htm
http://www.bradleybroadcast.com/2001/telephone.htm
http://www.audiotheater.com/phone/phone.html
http://www.taiaudio.com/right/sales/salescatalog/t elephoneinterface/teleinter.html
http://www.omnicronelectronics.com/PC/Computer_Acc y.htm
http://www.hut.fi/Misc/Electronics/circuits/telein terface.html
http://www.dplay.com/tutorial/Mac2tel.html

hope it helps

I agree that it's hilarious, every time.

However, it is possible to "sharpen" blurred images, if you have more than one (differently) blurred image of the same object. Then each blurred image contains different information, and with suitable assumptions about the blurring mechanism you can reconstruct a sharper image.

I had to do this at school a couple of years ago, and it seems they still give the same assignment: http://www.cis.hut.fi/Opinnot/T-61.5040/Htyo2006/o mm06_project.pdf . The idea is to construct a Bayesian model, and feed several blurred images to it. The model assumes that blurring works so that each blurred pixel is a weighted average of the high-resolution pixels surrounding the center of the blurred pixel, and some white noise of course. Obviously the exercise is quite elementary, but the idea is solid.

If we were really, really optimistic about CSI, maybe the surveillance cameras have a low resolution, but a high framerate, and the criminals/cars/whatever typically stand still for long periods of time. Then it would actually be possible to do what they do! :)

I had heard something like this before... just found it. Here's a group that has a virtual air guitar that works by wearing gloves. They limit you to 4 chords so you can't play anything "wrong." There is also a separate solo mode.

Nowadays the students have to take the course twice, once in a developer/tester role and again as a project manager, architect or in another specialist role.

The course is actually one of the best in school and greatly benefits everyone: the students get real-world experience in long term project and the clients get a low-cost, good quality (usually) software solution. The course overview

Well, actually Helsinki University of Technology did that as well:
http://www.acoustics.hut.fi/publications/files/the ses/lpeltola_mst/

Try also Chronically single. It is still relatively unknown and updates too rarely, but most slashdotters should recognize themselves in it.

Well, that was a mistake. You need to learn a little about Self-Organizing Maps and Learning Vector Quantization.

If you do not believe me, go to the source:

Self-Organizing Map of Symbol Strings with Smooth Symbol Averaging
http://www.google.com/url?sa=U&start=3&q=http%3A// www.cis.hut.fi/panus/papers/wsom03ssom.pdf&ei

Online Algorithm for the Self-Organizing Map of Symbol Strings, Neural Networks, 17, 2004, pp. 1231-1239.
http://www.cis.hut.fi/panus/papers/online_ssom.pdf

And yes, this is like a bad Star Trek episode - the one where Jordi is caputured and told just to make things work...

Listen my hostile little friend, read some science instead of ranting. Here's some things that will help you smooth your jagged idea of mind:

This is most important [but I suspect you cannot handle the ideas naked, so I will paste some easy-reading help below]:
Online Algorithm for the Self-Organizing Map of Symbol Strings [PDF]

Self-Organizing Maps
Dynamic-time warping

On the other hand if you want to secure name-to-address bindings, put a digital signature in a DNS TXT record...

Man I *knew* that term sounded familiar. For all youngsters out there you should get familiar with the Hacker's Dictionary. God forbid you start "inventing" slang that's been around since before you were born. http://www.tf.hut.fi/cgi-bin/jargon?search=automag ically

In my book, the "Ultimate Prank" took place in Stockholm, and it was carried out by Finnish technology-students. the prank itself is not widely known around the world, but it is IMO the best prank ever made. The story:

There is a well-known (to Finns) statue of Paavo Nurmi in Helsinki. Apart from that big statue, there are lots of smaller replicas of that statue. What does this have to do with pranks? Read on.

HMS Wasa is a famous Swedish warship that sank on her maiden-voyage. Year was 1628. the ship was re-discovered in 1961. The ship had sunked to depth of 62 meters. As the ship was re-discovered, it made big headlines in Sweden.

As it happens, there were some Finnish students from Helsinki University of Technology were visiting Stockholm when Wasa was re-discovered. And they decided to play the ultimate prank. They bought a miniature Paavo Nurmi-statue and went to the site where Wasa was discovered. The area was closely guarded, but somehow the students managed to dive in to the ship, made their way to the captains quarters, and placed the statue there. No-one knew that they had been down there, and they haven't told how they managed to do it.

As the ship was re-srufaced, experts started to study it. Only to find a relatively modern statue in the captains quarters. I can only imagine the collective "Huh?" they had when they found it :). the students had engraved their contact-information to the statue, and the students in question held a big news-conference about the incident, further annoying the Swedes.

further info (in Finnish though)

I'm guessing the story goes something like this

What about an semantic interface based on Websom http://websom.hut.fi/websom/milliondemo/html/root. html interface? Just an idea

What i like to see is an interface to semantic web like this : http://websom.hut.fi/websom/milliondemo/html/root. html

It's not your house - it's all the stuff IN your house. Worst offenders are generic PCs; specifically their case design (RFI/EMI-wise) is absolutely clueless (e.g., see http://www.ac6v.com/comprfi.htm/ for theory and fixes). Second place is firmly held by very, very crappy power supplies that let all the noise OUT of the PC on the power buss (ie. into your wiring). Sam's very useful Notes on the Troubleshooting and Repair of Small Switchmode Power Supplies) will get you started with theory and what [often] goes wrong (disclosure: I'm his sidekick). But:

Pet Peeve: as soon as Name-Your-$14-PC-PS-Manufacturer gets their UL sticker (meaning they can start selling in the US!) the ENTIRE L-C filter from the input of the power supply PCB is shorted out with a series of jumpers. Right, the 120VAC wires go through the save-thy-ass fuse right into the rectifier! No caps, no chokes, nuthin'! (Ok, so what do you expect from a $15 460W PS?) This allows all the noise on the power lines to enter the PC (and fry it - use a surge protector!) *AND* it allows all the noise IN the PC to escape back out and corrupt others (ie. your receiver, TV, etc.) (See: http://cms-emc.web.cern.ch/cms-emc/pdffiles/PhDfil es/PS&filters.pdf section 3.2 Switching mode power supplies for a nice overview). Oh, yeah, and I'm *SURE* all of you have your grounded cable actually grounded, right?

I got a 250W ATX knock-off case for $29 that came with a PS included. Turned the PC on, *ALL* AM stations vanished! Right... I opened it up and shure enough, a jumper from fuse to rectifier. All caps to ground were missing as well (from various points in the circuit). A few moments with a soldering iron (jelly-bean components, salvaged from dead *quality* PSes) and you can't tell the PC is on by listening to AM dial. Day and night difference!

Don't feel bad if you never though of it, this guy obviously never did either... and he should have. http://techreport.com/reviews/2004q4/psus/index.x? pg=1

But, how do you later chop up the files? I'm glad you asked: I use a hacked version of text-only (yeah!) soundgrab. You can get my latest version from http://repairfaq.ece.drexel.edu/filipg/soundgrab/

My signal comes from a battery-powered (ie. avoids issues with 'corrupted power' ;) digital sony walkman, via a 20-odd foot coax cable to my PC. The further away your receiver is from the source of noice, the better off you are![*] Linux records it from a SoundBlaster Live! with rawrec to a wav file. I have a series of templates (.sg files) for different shows and just fudge them a bit then export to MP3. Piece of cake! I've done a bunch of Dave's shows that way (~700MB worth) for inclusion in a weakness of mine (don't worry, they get some equally-illegit music ;).

Cheers

[*] "The solution to pollution is dilution" - Evil chemistry maxim applied to the wonderul and friendly world of RFI ;-)

That is interesting. I was very tempted towards the AFL, but ultimately steered towards the more-familiar LGPL. I forgot to post one more link that I also found useful:

License Comparison

This one gives a bit more analysis per license (and groups them nicely), but doesn't have such a nice, concise grid as the other link. -- Paul

From: Telephone line audio interface circuits

Safety issues of telephones

The telephones should be designed so that they do not cause danger to the user. The 48V DC voltage in telephone lines does not cause immediate danger to the user, but the AC ring signal (70-120V AC) can give a nasty shock. Telephone wires are also exposed to any different environmental effects (nearby lightning, ground potential differences in buildings, interference from power lines) which can cause that there are sometimes high voltage spikes on the telephone wires.

Fucking gay students from HUT. Stop this trolling shit.

$ cmp file1.dat file2.dat
file1.dat file2.dat differ: char 20, line 1

$ md5sum file1.dat file2.dat
a4c0d35c95a63a805915367dcfe6b751 file1.dat
a4c0d35c95a63a805915367dcfe6b751 file2.dat

$ xxd -s 1 -ps file1.dat | xxd -r -p | md5sum
84e6e0a21e2c4c9ef53f3762fdc90bc8 -
$ xxd -s 1 -ps file2.dat | xxd -r -p | md5sum
a63151008e5f8fc116ba947fd8af8c5a -

file1.dat:

00000000 d1 31 dd 02 c5 e6 ee c4 69 3d 9a 06 98 af f9 5c
00000010 2f ca b5 87 12 46 7e ab 40 04 58 3e b8 fb 7f 89
00000020 55 ad 34 06 09 f4 b3 02 83 e4 88 83 25 71 41 5a
00000030 08 51 25 e8 f7 cd c9 9f d9 1d bd f2 80 37 3c 5b
00000040 96 0b 1d d1 dc 41 7b 9c e4 d8 97 f4 5a 65 55 d5
00000050 35 73 9a c7 f0 eb fd 0c 30 29 f1 66 d1 09 b1 8f
00000060 75 27 7f 79 30 d5 5c eb 22 e8 ad ba 79 cc 15 5c
00000070 ed 74 cb dd 5f c5 d3 6d b1 9b 0a d8 35 cc a7 e3

MD5(file1.dat) = a4c0d35c95a63a805915367dcfe6b751

file2.dat:

00000000 d1 31 dd 02 c5 e6 ee c4 69 3d 9a 06 98 af f9 5c
00000010 2f ca b5 07 12 46 7e ab 40 04 58 3e b8 fb 7f 89
00000020 55 ad 34 06 09 f4 b3 02 83 e4 88 83 25 f1 41 5a
00000030 08 51 25 e8 f7 cd c9 9f d9 1d bd 72 80 37 3c 5b
00000040 96 0b 1d d1 dc 41 7b 9c e4 d8 97 f4 5a 65 55 d5
00000050 35 73 9a 47 f0 eb fd 0c 30 29 f1 66 d1 09 b1 8f
00000060 75 27 7f 79 30 d5 5c eb 22 e8 ad ba 79 4c 15 5c
00000070 ed 74 cb dd 5f c5 d3 6d b1 9b 0a 58 35 cc a7 e3

MD5(file2.dat) = a4c0d35c95a63a805915367dcfe6b751

See you in the next life.

The coldest spot on earth is in a laboratory in Finland:

http://boojum.hut.fi/Low-Temp-Record.html

Dome A is the coldest naturally occuring spot.

then you should loak at this and this and this and ...

The only Moria I've run across was a 2-D non-graphical overhead view. That Moria was created in the early 1980s, loosly based on Rogue. Is that at all similar to the one you're describing?

The only Moria I've run across was a 2-D non-graphical overhead view. That Moria was created in the early 1980s, loosly based on Rogue. Is that at all similar to the one you're describing?

The disk drives (1541 and 1581) were notorious for getting out of alignment. There was actually a small BASIC program written to help "knock" it back into alignment. You can see it on this page, near the bottom.

Unless there is an official hand recount, most ballots are seen only once with no checks. It wouldn't be hard for someone to count a few votes incorrectly.

Yes, it would. Vote-counting happens with individuals supervising from both major parties. Contrary to what you imply, votes aren't counted by a single individual each with no oversight.

Further, many jurisdictions require recounts whenever a vote is closer than a given percentage -- so in cases where it really matters, votes will be seen more than once.

It wouldn't be that hard for a broken (or tampered with) computer to count a large number wrong, or one person systematicly count them incorrectly.

The former is a very real problem, but one that more traditional counting mechanisms don't have (or make easily traceable).

All your complaints about possible exploits only reenforce my belief that you have to be able to check your vote after it is counted.

If you continue to think that the system as it stands are inadequate, there's plenty of research that indicates how it ought to be done (while avoiding the pitfalls that concern me). See here.

I replied to (presumably) your anonymous post as well, but just so you can learn something *specifically* about MTBF ratings in the drive industry, please read:

ASSUMPTIONS FOR RELIABILITY MODELS, part of Latent Sector Faults and Reliability of Disk Arrays, a dissertation by HANNU H. KARI of the Helsinki University of Technology.

Note for everyone though: "the actual reliability may vary dramatically and, for some units, MTBF can be only a fraction of the average MTBF."

I replied to (presumably) your anonymous post as well, but just so you can learn something *specifically* about MTBF ratings in the drive industry, please read:

ASSUMPTIONS FOR RELIABILITY MODELS, part of Latent Sector Faults and Reliability of Disk Arrays, a dissertation by HANNU H. KARI of the Helsinki University of Technology.

Note for everyone though: "the actual reliability may vary dramatically and, for some units, MTBF can be only a fraction of the average MTBF."

OK, burn coal or gasoline and the nitrogen stays in the air and doesn't chemically combine with anything. If you don't burn it in air....

But if the air gets hot (like in any thermal engine), some nitrogen reacts with some oxygen and you get some NOx. And hydrogen is pretty much the same or somewhat worse than gasoline with respect to this property depending on how you look at it.

If you look at the autoignition temperatures (the lowest temperature a fuel will burn) hydrogen's is 530 C. And the autoignition temperature of gasoline is about 260 C. So you can make gasoline produce much less NOx than hydrogen with some effort.

Hydrogen's flame temperature is 2045 C in air. Gasoline's is 2197 C, almost the same. This is the worst case temperature for the two. Sloppy engines will probably put out about the same NOx.

(Info from http://www.fuelcellstore.com/information/hydrogen_ safety.html,
http://www.hut.fi/Units/AES/projects/renew/fuelcel l/posters/hydrogen.html and http://hypertextbook.com/facts/2003/ShaniChristoph er.shtml.)

Put an RSA-8192 signature using the 288 bit hash "SHA1|MD5" and PKCS-2 padding on the magnetic stripe. That would both contain the information on the front as well as sign the data as legitimate.

Only a secure black box (a la nCipher) at the state DMV would contain the private key. The public key would be given out publicly, especially card scanner manufacturers. The FBI or NSA could have a root certificate, and could countersign the public keys of each of the 50 states and other agencies issuing drivers' licenses.

We really need the federal government demanding a particular standard for the mag stripe system. Let the states design the fronts as they please, but at least make them all electronically readable. Don't build a central federal database with the information, as the public haaaaates that. Keep data storage the same, just make it possible to authenticate the data on the cards.

How much data can be stored on a mag stripe? Maybe you could fit a 10k image of the person on the mag stripe for a card reader to display. Like the other data, this would be hashed, identifying the photograph as authentic.

(Operator | in traditional cryptographic notation means "concatenate"; IE, attach the 128 bit MD5 hash to one of sides of the 160 bit SHA-1 hash to make a 288 bit hash function. SSH uses this. The MD5 algorithm has been shown insecure, as this output shows:

$ md5sum file1.dat file2.dat;sha1sum file1.dat file2.dat
a4c0d35c95a63a805915367dcfe6b751 *file1.dat
a4c0d35c95a63a805915367dcfe6b751 *file2.dat
2783c4ff4a3f20d25f2598a8b052b890c37dca c4 *file1.dat
3c35410823ef00b12d020981c1cf8564c0f89b cc *file2.dat
Click me for a site talking about this MD5 collision

SHA1 is also suspect. So combine the two for security. The fact that SHA1 and MD5 use opposite byte orders makes things even more secure.)

Does the new format handle Gamma correction when viewing over the web? PNG was supposed to do this, but does not do a great job.

Liquid helium-3 (3He) displays similar properties. There is a minimum in its Pressure vs Temperature "melting" curve at 0.32 Kelvin and 29 atmospheres. If one cools a sample of the liquid to, say, 0.20 Kelvin and apply a pressure of 30 atmospheres, the sample would remain liquid. Heat it isobarically to 0.3 kelvin and the liquid will begin to solidify. This new substance is the SECOND liquid to display this property. You can look at the melting curve here:
http://boojum.hut.fi/research/theory/helium.html

• brickOS,
• LegOS,
• and Java-based one.

> ftp 216.52.171.81
Connected to 216.52.171.81.
220 ProFTPD 1.2.10rc3 Server (Superconnect FTP) [bubbles.hou.corp.sc.com]
Name (216.52.171.81:): anonymous
331 Anonymous login ok, send your complete email address as your password.
Password:
230 Anonymous access granted, restrictions apply.
ftp> cd /pub/tcl/sorted/file/X-Files1.21b
250 CWD command successful
ftp> ls
ftp> ls -l
200 PORT command successful
150 Opening ASCII mode data connection for file list
index.html
CHANGES.html
X-Files1.21b.tar.g z
226 Transfer complete.
remote: -l
47 bytes received in 0.0049 seconds (9.35 Kbytes/s)
ftp> get X-Files1.21b.tar.gz
200 PORT command successful
150 Opening ASCII mode data connection for X-Files1.21b.tar.gz (115245 bytes)
226 Transfer complete.
local: X-Files1.21b.tar.gz remote: X-Files1.21b.tar.gz
115723 bytes received in 0.54 seconds (210.45 Kbytes/s)
ftp> 221 Goodbye.
> tar -tzf X-Files1.21b.tar.gz
CHANGES
Readme.install
Regi stration.form
X-Files.1x
X-Files.tcl
tar: Skipping to next file header

gzip: stdin: invalid compressed data--crc error

gzip: stdin: invalid compressed data--length error
tar: Child returned status 1
tar: Error exit delayed from previous errors

Here's an actual example of two different binary files having the same md5sum.

As for simply removing the gAMA chunk from PNG files (as you describe) will alleviate the situation for most browsers ... but then there's Safari, which applies gamma correction to PNG images lacking a gAMA chunk. Pretty much kills color matching should a web author care about Safari.

Recommended reading: The Sad Story of PNG Gamma 'Correction'.

For a systematic view of the problem of PNG support as it relates to gamma correction, you can just skip to the pretty chart.

The solution you offer -- removing the gAMA chunk from all PNGs -- is suboptimal for real-world cases. Sorry.

As for simply removing the gAMA chunk from PNG files (as you describe) will alleviate the situation for most browsers ... but then there's Safari, which applies gamma correction to PNG images lacking a gAMA chunk. Pretty much kills color matching should a web author care about Safari.

Recommended reading: The Sad Story of PNG Gamma 'Correction'.

For a systematic view of the problem of PNG support as it relates to gamma correction, you can just skip to the pretty chart.

The solution you offer -- removing the gAMA chunk from all PNGs -- is suboptimal for real-world cases. Sorry.

"Please post citations (links) to such proofs."

Of course