as a reasearcher in CompSci, I must sheepishly admit to worship the great goddess of X86 (albeit through AMD)... started using Unix (Coherent) in 89 though. And yes, there is something wrong with "lpr foo.ps": it's got to do with dead trees. (I know, you didn't write this, but the AC after you)
That bit that quotes the government statistic has the number L-9-9-8 (as opposed to 1998) in it...
OK, ok, so I haven't participated in the debate on sociological factors. But then, I'm german and I'm still mulling over the meaning of the choice of words in a recent press release of my government: they want people to take more "initiative" where Americans encourage "responsibility"... should I agree?
There are so many messages here who just take it for granted that NT is insecure. NT has a solid security architecture that is more fine-grained than that of Linux. This means it COULD be better. The real problem is that MS Office is designed for a single user and requires you to have the equivalent of ROOT access to run it (OK, I'm exaggerating and I've never had office on my computer so I wouldn't know, but disprove me). You could do exactly the same with Linux (pop up a box in netscape, make the user type their password, mail it home), only that a user has less rights on Linux.
It is my understanding that they were talking about sustained recording. You are talking about a peak rate. I doubt you remember everything if I play *random* notes from your 88 note keyboard to you for an hour (a layman's definition of "bit" in the info-theoretic sense would be the info transmitted over a channel that can transmit symbols from an alphabet of two symbols where either symbol is equally likely, if you build a markov chain and can make predictions about the next note, as you do in music, it's not a bit).
You flippantly mention "remembering the number of cracks in the wall". I learned that each of our eyes is very very inaccurate and that we have a 2 degree section of focus, in which we see with the accuracy that we're accustomed to. I learned that we constantly skip about with our eyes and that our view of the world around us is mostly an internal model, with small parts of it updated frequently with high precision. If what I learned is true, you'd not remember the cracks in the wall., because they wouldn't have your attention.
GNUS + BBDB and some custom macros and you have an information port (reading mail, news, man pages, info docs, and arbitrary other stuff) that is keyboard operable, has grouplens filtering power and mail expiry features that outlook won't have in 10 years' time. Most people just haven't been exposed to what CAN be done. As a matter of fact, I even have toys like X-faces, my own encryption, a uk-to-german filter and address-book synchronisation with my psion siena in it. Thanks to Masanobu UMEDA and Lars Magne Ingebrigtsen for giving me an incredibly powerful tool!
To my mind there were 4 groups of people doing fairly dissimilar things. They all called their research "oodbms" work. 1) relational people wanting to support structured fields (break 1st normal form) 2) persistent programming people wanting transaction / query facilities 3) AI expert systems people wanting a persistent rulebase 4) functional data modelling people experimenting with polymorphism The outcome are two major strands: persistent object stores and object relational systems. In principle they can do the same, but in practice they excel in different areas: if you let the DBMS know about your data types (ADTs - abstract data types) then the query optimiser can make use of that information. That's the great strength of ObRel systems. On the other hand, if you look at how rel systems are used, you find that many introduce lots of surrogate keys (simplistically: serial numbers with no meaning in reality). An OODB would be able to replresent those by object references, which cuts down on the pointer indirections by a factor 2 or three. If all you ever do is chasing pointers that may be a great help. Look at CAD software. There are two papers that I can highly recommend: the oo database manifesto and the (empire strikes back) 3rd generation database manifesto. Look over at dblp.uni-trier.de and type in manifesto.
1) is it possible that he's given up on selling licenses for Office and OSs? MS thinks that in 2003 the infrastructure will be in place to rent out software. First step: lo/cost version for China. 2) is it possible that he's given up on making money off Office and OSs alltogether and moving to do things like billing, databases, erp, eTrade? 3) is it possible that MS will control a different resource than the OS? Say, patents, formats, whatever? But then, he may just be bluffing.
I saw an article a few years (10 or so) ago that claimed that a "proper" application ought to have a robust core, a concise help system (why am I so happy about man pages and so frustrated with Win9 help?) and both beginner and expert modes. Incidentially, I worked for the WordPerfect help desk and I was quite taken by the WP expert mode (switch off the menus and disable the mouse...). When WP went "windows" it lost its command structure and its macros. What a shame.
Hello, I'm just wondering whether the lady who tested Linux may actually not have performed the procedure as described. She says that downloading linux would have taken 5 hours (implying a modem) and later claims that it would have been too difficult to connect to the corporate network. Am I missing something here? I also suggest keeping the tech guy article for future reference on the competence of senior MS technical staff.
Troy, I think there's one more difference: Unix processes have separate address spaces. Each address space starts at 0 (2gig, whatever), hence they overlap. For a thread change you keep the virtual-to-physical memory mapping, for a process change you need to change it (which means flushing the TLB - translation lookaside buffer, lots of traffic between MMU, which is part of the CPU-"chip" and the main memory). Also, I think the main Symmetrical MP problem is memory bandwidth.
not "transport". And it's IP addresses, not TCP/IP addresses. And are they really bound to domains? I'm not an expert on this, so I better shut up now...
The ideas are really applied steganography. That is, one tries to perform undetected communications. Only, if you were to do this without telling people, they'd denounce you as sneaky. So how is it done? Firstly, we want it to be transformation-proof. I'd say turning it into analog and back should preserve the watermark, though it might "smudge" it. So what do we do? We create some piece of data that is seemingly random and forward-error-correct (introduce redundancy so that it can be perfectly reconstructed in the presence of errors... preferrably with the correct error model). Then we encode it into the "analog" data. How do you encode info? You modulate it. Frequency modulation and amplitude modulation are familiar from the radio, phase modulation is in every modem. Pick and choose, I'm not an expert, so I don't know what would work best with music. Now add this to your music and encode the music as an mp3. This could pose a problem, since mp3 uses compression, knowing what the ear (and psyche) pick up on and leave out the bits we don't notice so much. So some of the info might get lost, but if you try to encode 128 bits, which FEC-ed turn into maybe 2k, into something that is uncompressed maybe 20mb, that should be quite well spread, un-noticable, unless you know what you're looking for and fairly indestructible (depending on your FEC you might be able to chop parts off or out...). Now how would you go about extracting this? You turn the mp3 back into 20mb of raw data and perfrom some signal processing. U-Boats are found by doing a convolution of the data against itself, GPS uses similar techniques to extract the spread-spectrum very-low-bitrate data from the noisy background. Once you have it, extract the original data and hey presto. Now, I'm into databases, not security. I'm just trying to apply *from memory* what I *should have learned* half a decade ago. Please excuse any inaccuracies and those cryptographers, please feel free to correct me.
I now hope people see how this could be used, un-noticeably, to identify some signature. My scheme depends on a secret, making it impossible for users to detect forgeries. I don't know whether that would be useful, but it should be enough to get a litigation going.
I seem to remember reading an article from some Swiss professor. I believe he used some words that you use at the end of your post. I vaguely remember the topic of the article: WW III will be fought between those who wish and don't wish to allow computers to attain consciousness. I found it through a link from CNN about a year ago.
I've been changing the font on a TI99-4A at 11. I had soldered a parallel cable onto the 6501 of my C64 floppy drive within the first week of getting it. I've done a 250.000 dollar military project at 17, and I didn't have a girlfriend until I was 21. Does that count as geek?
However, I've started a BA in CS in 1992, graduated in '95 and am close to completing my PhD in distributed OO databases. I believe I have both perspectives.
Uni taught me things that I wouldn't have considered before:
- skyscrapers (built by engineers) typically don't fall down. Can we say that about software? - what is the difference between patent/copyright? - would I be able to sleep after building a medical system and not doing proper testing? (Ethics) - why are Mac-Fans so fanatical over their toys? - can you write programs without using variables (functional programming) - what does a secretary think how the windows desktop works? - can you name 10 different ways to sort data and what the tradeoffs are? - programming using randomness - why are neural networks NOT like the brain? - what are the similarities and differences of the phone and an ethernet? (packet/vc, in-band/ out-of-band signalling, flow control, statistical multiplexing, etc.) - what is the fundamental difference between the mode of communication in phone and fax? - do you understand the difference between identity and equality? - how could one parse natural language? - how does human vision work and what can we learn from it? - programming by specifying the problem (prolog)
Now besides those clear questions, I learned a number of invaluable skills:
- traversing a hierarchy of levels of abstraction - understanding tradeoffs - understanding what my goal is or should be - planning time - why memorisation is as important as understanding - talking to non-technical people - that technology and politics are intermingled
I hope that at least people out there agree with me.
Slashdot Mirror
as a reasearcher in CompSci, I must sheepishly ... started using Unix (Coherent) in
admit to worship the great goddess of X86 (albeit
through AMD)
89 though. And yes, there is something wrong
with "lpr foo.ps": it's got to do with dead trees.
(I know, you didn't write this, but the AC after you)
That bit that quotes the government statistic has the number L-9-9-8 (as opposed to 1998) in it ...
OK, ok, so I haven't participated in the debate on sociological factors. But then, I'm german and I'm still mulling over the meaning of the choice of words in a recent press release of my government: they want people to take more "initiative" where Americans encourage "responsibility" ... should I agree?
... and I've not touched the mouse of my computer
in at least 3 days (while writing on a diss).
---> different people, different criteria.
Hello, the homepage of Dr Daugman, the guy who has invented some of the technology can be found here. Look for "iris recognition".
There are so many messages here who just take it
for granted that NT is insecure. NT has a solid
security architecture that is more fine-grained
than that of Linux. This means it COULD be better.
The real problem is that MS Office is designed for
a single user and requires you to have the equivalent
of ROOT access to run it (OK, I'm exaggerating and
I've never had office on my computer so I wouldn't
know, but disprove me). You could do exactly the
same with Linux (pop up a box in netscape, make the user type their password, mail it home), only that a user has less rights on Linux.
It is my understanding that they were talking
about sustained recording. You are talking about
a peak rate. I doubt you remember everything if I play *random* notes from your
88 note keyboard to you for an hour (a layman's
definition of "bit" in the info-theoretic sense
would be the info transmitted over a channel that
can transmit symbols from an alphabet of two
symbols where either symbol is equally likely, if
you build a markov chain and can make predictions
about the next note, as you do in music, it's not a bit).
You flippantly mention "remembering the number of
cracks in the wall". I learned that each of our
eyes is very very inaccurate and that we have a
2 degree section of focus, in which we see with
the accuracy that we're accustomed to. I learned
that we constantly skip about with our eyes and
that our view of the world around us is mostly an
internal model, with small parts of it updated
frequently with high precision. If what I learned
is true, you'd not remember the cracks in the wall., because they wouldn't have your attention.
GNUS + BBDB and some custom macros and you
have an information port (reading mail, news,
man pages, info docs, and arbitrary other
stuff) that is keyboard operable, has grouplens
filtering power and mail expiry features that
outlook won't have in 10 years' time. Most
people just haven't been exposed to what CAN
be done. As a matter of fact, I even have toys
like X-faces, my own encryption, a uk-to-german
filter and address-book synchronisation with
my psion siena in it. Thanks to Masanobu UMEDA
and Lars Magne Ingebrigtsen for giving me an
incredibly powerful tool!
To my mind there were 4 groups of people doing fairly dissimilar things. They all called their research "oodbms" work.
1) relational people wanting to support structured fields (break 1st normal form)
2) persistent programming people wanting transaction / query facilities
3) AI expert systems people wanting a persistent rulebase
4) functional data modelling people experimenting with polymorphism
The outcome are two major strands: persistent object stores and object relational systems. In principle they can do the same, but in practice they excel in different areas: if you let the DBMS know about your data types (ADTs - abstract data types) then the query optimiser can make use of that information. That's the great strength of ObRel systems. On the other hand, if you look at how rel systems are used, you find that many introduce lots of surrogate keys (simplistically: serial numbers with no meaning in reality). An OODB would be able to replresent those by object references, which cuts down on the pointer indirections by a factor 2 or three. If all you ever do is chasing pointers that may be a great help. Look at CAD software.
There are two papers that I can highly recommend: the oo database manifesto and the (empire strikes back) 3rd generation database manifesto. Look over at dblp.uni-trier.de and type in manifesto.
1) is it possible that he's given up on selling
licenses for Office and OSs? MS thinks that in
2003 the infrastructure will be in place to rent
out software. First step: lo/cost version for China.
2) is it possible that he's given up on making
money off Office and OSs alltogether and moving to
do things like billing, databases, erp, eTrade?
3) is it possible that MS will control a different
resource than the OS? Say, patents, formats,
whatever? But then, he may just be bluffing.
I saw an article a few years (10 or so) ago that ...).
claimed that a "proper" application ought to have
a robust core, a concise help system (why am I so
happy about man pages and so frustrated with Win9
help?) and both beginner and expert modes.
Incidentially, I worked for the WordPerfect help
desk and I was quite taken by the WP expert mode
(switch off the menus and disable the mouse
When WP went "windows" it lost its command structure
and its macros. What a shame.
Hello, I'm just wondering whether the lady who
tested Linux may actually not have performed the
procedure as described. She says that downloading
linux would have taken 5 hours (implying a modem)
and later claims that it would have been too
difficult to connect to the corporate network. Am
I missing something here?
I also suggest keeping the tech guy article for
future reference on the competence of senior MS
technical staff.
Troy, I think there's one more difference:
Unix processes have separate address spaces.
Each address space starts at 0 (2gig, whatever),
hence they overlap. For a thread change you
keep the virtual-to-physical memory mapping,
for a process change you need to change it (which
means flushing the TLB - translation lookaside
buffer, lots of traffic between MMU, which is
part of the CPU-"chip" and the main memory).
Also, I think the main Symmetrical MP problem is memory bandwidth.
not "transport". And it's IP addresses, not ...
TCP/IP addresses. And are they really bound
to domains? I'm not an expert on this, so I
better shut up now
The ideas are really applied steganography. That is, one tries to perform undetected communications. Only, if you were to do this without telling people, they'd denounce you as sneaky. So how is it done? ... preferrably with the correct error model). ...).
Firstly, we want it to be transformation-proof. I'd say turning it into analog and back should preserve the watermark, though it might "smudge" it. So what do we do? We create some piece of data that is seemingly random and forward-error-correct (introduce redundancy so that it can be perfectly reconstructed in the presence of errors
Then we encode it into the "analog" data. How do you encode info? You modulate it. Frequency modulation and amplitude modulation are familiar from the radio, phase modulation is in every modem. Pick and choose, I'm not an expert, so I don't know what would work best with music.
Now add this to your music and encode the music as an mp3. This could pose a problem, since mp3 uses compression, knowing what the ear (and psyche) pick up on and leave out the bits we don't notice so much. So some of the info might get lost, but if you try to encode 128 bits, which FEC-ed turn into maybe 2k, into something that is uncompressed maybe 20mb, that should be quite well spread, un-noticable, unless you know what you're looking for and fairly indestructible (depending on your FEC you might be able to chop parts off or out
Now how would you go about extracting this? You turn the mp3 back into 20mb of raw data and perfrom some signal processing. U-Boats are found by doing a convolution of the data against itself, GPS uses similar techniques to extract the spread-spectrum very-low-bitrate data from the noisy background.
Once you have it, extract the original data and hey presto.
Now, I'm into databases, not security. I'm just trying to apply *from memory* what I *should have learned* half a decade ago.
Please excuse any inaccuracies and those cryptographers, please feel free to correct me.
I now hope people see how this could be used, un-noticeably, to identify some signature. My scheme depends on a secret, making it impossible for users to detect forgeries. I don't know whether that would be useful, but it should be enough to get a litigation going.
I seem to remember reading an article from
some Swiss professor. I believe he used some
words that you use at the end of your post.
I vaguely remember the topic of the article:
WW III will be fought between those who wish
and don't wish to allow computers to attain
consciousness. I found it through a link
from CNN about a year ago.
I've been changing the font on a TI99-4A at 11.
I had soldered a parallel cable onto the 6501 of
my C64 floppy drive within the first week of
getting it. I've done a 250.000 dollar military
project at 17, and I didn't have a girlfriend
until I was 21. Does that count as geek?
However, I've started a BA in CS in 1992,
graduated in '95 and am close to completing
my PhD in distributed OO databases. I believe
I have both perspectives.
Uni taught me things that I wouldn't have
considered before:
- skyscrapers (built by engineers) typically don't
fall down. Can we say that about software?
- what is the difference between patent/copyright?
- would I be able to sleep after building a
medical system and not doing proper testing?
(Ethics)
- why are Mac-Fans so fanatical over their toys?
- can you write programs without using variables
(functional programming)
- what does a secretary think how the windows
desktop works?
- can you name 10 different ways to sort data and
what the tradeoffs are?
- programming using randomness
- why are neural networks NOT like the brain?
- what are the similarities and differences of the
phone and an ethernet? (packet/vc, in-band/
out-of-band signalling, flow control, statistical
multiplexing, etc.)
- what is the fundamental difference between the
mode of communication in phone and fax?
- do you understand the difference between
identity and equality?
- how could one parse natural language?
- how does human vision work and what can we learn
from it?
- programming by specifying the problem (prolog)
Now besides those clear questions, I learned a number of invaluable skills:
- traversing a hierarchy of levels of abstraction
- understanding tradeoffs
- understanding what my goal is or should be
- planning time
- why memorisation is as important as
understanding
- talking to non-technical people
- that technology and politics are intermingled
I hope that at least people out there agree with me.
so long,
os10000