I registered to vote in an other county in
NY more than 6 months ago (which is supposed
to delete my NYC registration) but my record still
appears on the site, so it is at least a few
months behind or they are slow deleting
re-registered voters.
I have taught a number of mathematics courses
for prospective and current schoolteachers, and
I must say it has been a very sobering experience.
Given the rewards and requirements of the teaching profession, the kind of
people who are drawn to the career (and those who
stumble towards it) are really
not those who are particularly suitable for it.
For prospective elementary schoolteachers, the
last math course required of them at one university
in California that I taught at is a course titled
"Elementary Problem Solving." The topics of that course were
carefully chosen to have essentially no prerequisite knowledge
of algebra or geometry (mostly
basic divisibilty/primality topics,
counting combinations, and pretty straightforward
topics that a strong 7th grader would get the
hang of in about a month.) The students there
struggled spectacularly with the
topics and were generally unable to manage even
a first level of abstraction. We are not talking
difficult problems- questions like "How many different
ways can you make a sandwich if there are three
different kinds of bread, four different kinds of meat
and a customer can have up to two different kinds of
four varieties of cheese?" More depressingly,
they were in general not at all fazed by
their failures, and spent more energy complaining
about my unreasonable expectations that they
did trying to solve problems. The general litanies I heard were
"I only want to teach 2nd grade- why should I
need to know any of this?" and "I need to pass
this course to become a teacher, and everyone tells
me I'll be a great teacher because I like kids so much!"
I found teaching that course to be not particularly rewarding,
and in fact, the people in my department who most often taught
that course were the ones with the absolute lowest expectations
of their students.
The students tend to think that to teach 3rd grade math, they need only
know the math that a 3rd grader learns.
The idea that a teacher should understand a subject
thoroughly enough to have actual insight is totally alien.
Another comment related the expanded opportunities
available to women as contributing to the problem.
This is very clear. In the bad old days, the acceptable
careers for women were schoolteachers and nurses.
Neither one of these paid well, but since the overall
opportunities for women were limited, there were many
bright, capable women who entered those careers, thus
artificially enriching the level of teachers available
for a fixed salary and prestige.
Now, thankfully, there are many more opportunities
for women so the bright capable ones are no longer
limited to teachers and nurses- they can become
engineers and lawyers and whatever else. Unfortunately, that
means that who is left to go into the field but
less capable people of both genders. (I figure that
both the health care and education crises are
complicated by this effect.) Essentially, the
societal pressures limiting women to "traditionally
nurturing careers" artificially
reduced the cost of getting good teachers. Now, that pressure
has lessened with no increase in salary or respect to
compensate, so there has been an overall decline in
the competence of schoolteachers.
Even with stronger requirements for math and science teachers,
there is little effect. In California and New York, a reasonably
competent school adminstrator can staff all math and science
teachers with uncredentialled teachers-- and many are. In
some parts of California, fewer than 10% of math teachers
are credentialled (with a very weak credential, BTW) and
the remainder have "emergency" credentials that can be
extended indefinitely, with only a slight amount of
administrative imagination. The great need for people
thwarts any effort to raise the credential requirements,
which are pretty much moot anyway.
I don't know what a good solution is but it is clear that
greater resources need to be spent to improve the
situation, if indeed this is something that is important to
people. Everyone seems to be "for education" but given
the costs of the changes that need to be made, support
for significant change vanishes.
I have found sites like
domania.com
very useful for looking up recent sale prices
in an area. This was useful just to get
an idea of prices overall, as well as
specific sale history about the property
I was interested in. Their database is
not complete, and will not include data
from states that don't make it publicly
available (eg Utah) but I found it useful.
It is also interesting to see how much the
house your friend bought was, when it sold,
and things like that.
When I was TAing an undergrad prog course, a prof assigned
a "greatest common substring" assignment.
After people had turned their programs in,
(this was more than 10 years ago when people
actually turned in printouts)
he mentioned that he would run their programs
looking for common substrings in their code and
long substrings would be analyzed for plagarism.
Students were given the option of withdrawing
their programs; many did.
As a programming TA, we routinely ran
a not-very-sophisticated "cheat check" program
which looked for copied code. By comparing
object code (with debug info and variable names
stripped), we would often see code that
differed only by changing variable names.
Sometimes the copying was ridiculously
blatant- at one institution I taught at, every
student was required to include an "honesty
pledge" at the beginning of all programs.
Oftentimes, we would find (and work to expel)
students whose programs differed only in the
honesty pledge. The most memorable was one
student who failed to delete the spaces that
arose when his name was shorter than the much-longer name of the student
he copied the program from.
Sometimes, students are held liable if
their code is copied without their knowledge.
This thwarts the "dumpster-diving" for printouts
that can happen; students are much more
careful with their printouts if they are
responsible for someone else copying their
code, even without their knowledge.
Before everyone had their own computers and
had to actually turn up to a lab to use VT100s, we
would often look at the 'last' logs and see if
suspiciously similar code was written by students who
were sitting next to each other in the lab. Or
sometimes if we weren't sure if something was
on the up-and-up, we could see how much time the
student spent on the program, when and from where.
More than once we had a student whose friends
were telnetting in to get the assignments from
our assignment distribution program.
In the same vein, we could look around their
directories to see if they had written test data,
tried some other strategies, etc. to see if they
were genuine.
These days, I make sure my exams test
the knowledge that would be obvious if you
actually wrote the code. If you didn't write your own code, you'll get hammered on the exams
and fail, even if I don't figure out where your
code came from.
Programming is vulnerable to copying. Oftentimes,
for short assignments, there aren't that many
possible good solutions to a problem, so repetition
is not nescessarily the sign of forbidden collaboration. In general, the longer the
assignment, the easier it is to spot. But over
the years, there have been some amazingly boneheaded cheaters... Even if they do manage
to sneak in someone else's program, almost always
they get nailed on the exam. And catching a
student cheating is depressing; I feel an
obligation to the students who do the work
to make sure that the student is expelled
or disciplined as sternly as possible.
Oftentimes, that is ugly and sad, but needed.
Here's some advice: if you are enough
of a moron to cheat and get caught, 'fess up
and throw yourself on the mercy of whomever
decides. It is so sad to watch a student
deny the obvious and people are so much
sterner when the student thinks the story
is fooling someone...
My understading of the approach the project
takes is that it misrepresents the
perspective that almost all working
mathematicians have.
Working mathematicians rarely think about
their proof in terms of the reduction to
the primitives of set theory.
Constructing proofs takes a great deal
of mathematical preparation, creative energy and fluency in current results.
Figuring out what level of detail is
appropriate for a proof (and the audience
of a proof) requires understanding what
is "straightforward" and what is not.
If a typical modern proof were reduced to
the axioms of set theory, it would be perhaps
hundreds of
thousands of pages long and the key ideas
that are of importance to the proof would
probably be.0001% of that amount, and be made
totally opaque to any reader since the ideas
would be scattered unintelligbly throughout
the bulk which would be describing
"straightforward" mathematics in an indigestible
manner.
Caveat: there are mathematicians in
a very speciallized subsegment of the highly
technical field of set theory who do like to
think in terms of reduction to primitive axioms.
However, this is a tiny fraction of the
mathematical community and arguably the wrong
part of mathematics for a casual observer to
try to digest. Certainly, the attention that
ZFC set theory gets by informal approaches is
grossly disproporionate to its use within the
overall fields of research mathematics.
An analgous overly-reductionist treatment of biology would be that by understanding
Schrodinger's equation, all of chemistry and
thus biology is
understood. It may be true that chemical understanding
can be broken down into solutions of the fundamental
equation of quantum mechanics,
but the fact of the matter is that for the
vast majority of chemical phenomena, that
is not a useful approach to take. That approach
would be to unwieldy to yield good understanding
and new results. And for
biological phenomena, which are in principle
based entirely on chemical phenomena, reduction
to quantum mechanics is practically
never the useful approach to take.
Modern mathematics is difficult to approach
since mathematics is such an old field. Almost
all of the mathematics learned by a typical
undergraduate is based on research that was
done in the 1800s. Mathematics has flourished
and continued its progress
since the 1800s, of course, but for many
fields, learning the prequistes (working through
the 1900s, in effect) takes a great deal of
energy and can take several years. In some
fields it is possible to understand and
appreciate more modern results, but in most
fields, years of study beyond the undergraduate
level are needed to get up to speed for
research mathematics. Fields in science
(physics, chemistry, biology, etc.) tend to
be based on research that was done much
more recently than the 1800s and it is much
easier to learn the prerequiste material
than in mathematics. There are no "easy
shortcuts" to the frontier of mathematics
(in general) precisely because it is an old field
that has been richly developed by generations
of work.
Rob Kirby, a prominent topologist at UC Berkeley, has been active in trying to improve the
journal situation for mathematicians.
The idea is to boycott the high-priced journals
by not submitting to them, and instead
submit to journals, especially electronic ones,
which are free or reasonably priced. Here is his
orignal letter and here is
an updated price list.
A number of research mathematicians take these
considerations into effect when deciding where
to submit, so perhaps things will improve.
The most preposterous thing about high-priced
journals is that the "value-added" part of
a journal is the peer review, which is done
almost always for free. When an article is
submitted it is sent out for review to someone
whose research is close enough to understand the
work. Getting an article to review is a chore;
it can take many months to thoroughly review
an article, many are poorly written and have
annoying minor mistakes, and there is no
recognition or pay associated to it. When it
turns out that the journals are priced outrageously, that is the final straw for many.
In general, reviewing articles is considered
a nescessary public service, and since the
editors of the highest-priced journals tend to
be the super-big shots, it is not easy to
refuse to review something. Hopefully,
things will improve! The xxx archive is
great for preprints but the reviewing
process is an important part of disseminating
research so it will take more than that for
things to get much better.
The high school student Josh has a remarkable
new proof and it is inspring to see original
work done in such an estabilished area
by someone so young. Normally, it takes
years of study past the
undergraduate level to even get to the point
where you can understand the statements of
new results in mathematics.
A few clarifications though:
The proof is a nicer, shorter proof of a
known theorem. It is important to understand
the difference between a proof of a new result
(where the truth or falsehood was not known
beforehand) and a proof of an existing result (where it the result was known, but perhaps
using a complicated method or advanced results
from some other work on other questions.)
New proofs of existing results are important if
they are improvements of earlier proofs, as
in this case, but there is a different flavor
to them. Clever approaches that were not seen
on the first proof can sometimes surface later;
there is a remarkable example of a result
(not that far removed from the field of this work)
that when first proven in the 1800s, took several
hundred pages- modern proofs can prove the
same result in three lines (the impossibility
of a finite projective plane embedded in the
xy-plane authentically, if you are curious.)
Thus it is worth noting that the article mentioned
appears in the American Math Monthly, which does
not tend to publish hard-core new research results, but
instead elegant, elementary proofs of known
results. Just meant to clarify; this is a remarkable result (and let me repeat the
earlier plug for the Geometer's Sketchpad- quite
a remarkable piece of software...)
If this gets going in the US, does that mean
that we can apply for federal funding for
lightsabers as part of a "faith-based"
social improvement program?
From this release,
it appears the policy is less severe than originally described:
Forwarding a personal e-mail is unlikely to breach copyright
laws. A court would need to find that the contents of the e-mail
were an "original literary work". For example, if the e-mail was
simply a joke that everyone had been re-hashing for years, it is
doubtful it would have the necessary originality to be protected
by copyright. Similarly, a casual exchange of personal
information or office gossip would probably not be original
enough to have copyright in it.
Still it seems remarkable to have criminal penalties
associated with unauthorized forwarding. Canada,
which has a policy that correspondence, written
and otherwise, is the property of both parties
(complicating lots of "Collected Letters of blah"
books) at least restricts the remedy to civil
cases, not criminal ones.
Printing support in the beta was great for
Postscript and networked printers, but for
"commodity" inkjets it was very weak, and
since Apple was putting the burden on
Epson, etc. to write their own drivers, it
didn't happen and doesn't seem real likely
to happen quickly, particularly for older
inkjets. What is the incentive on the
printer manufacturer's part? They'd rather
sell a new printer instead of write a
driver for a printer they sold 3 years ago
for $100.
The original Wolfenstein was from Muse software in
1981 for the Apple II, and was one of the first
with (almost recognizable) speech- "Achtung!"
Check out this page
for a screenshot from the Apple II if you want
a little memory trip. The page also has info
about running it under emulation on a PC.
Apparently, the father of the little girl is
Dave Thielen, known for stirring up some trouble
himself. For example, here is
a copy of the page he used to pretend to sell
absentee ballots on ebay last November, and
it seems that perhaps the apple has not fallen far
from the tree here...
Having worked in five different universities
where a large fraction of the day-to-day
sysadmin and networking tasks were done by
(paid) students, there is no question that in general
this works well for both sides.
Universities need a great deal of diverse
IT support and cannot afford to pay
salaries appropriate for trained grown-ups. Competent, curious students with interest
can learn a great deal of valuable, marketable things in this setting very quickly, without
the "experience" needed for comparable positions
in industry. Academic tech support is generally
more forgiving for gaps in expertise than
the "real world" and is a good place to learn
the basic necessities.
The disadvantage, from the university side, is
the regular turnover as students learn enough
to get excellent jobs. The disadvantage from
the student side is seeing how preposterously
screwed up things can be and having to put
up with various politics, awful documentation and limited resources. Oh, wait- since things are
often like that outside the university, maybe
that is an advantage...
The danger of an entirely-student run
network comes from the regular turnover, so
perhaps that would be a problem. The better-
run academic IT departments that I know have a number of permanent
staff who were former students, and who decided
that there were some nice things about the
academic computing environment (or who
just wanted massive bandwith to play with)
and stayed around.
The bureau for decades has engaged in a
little-known technique called "data swapping," in
which a few key pieces of information about one
person are switched with those of another person
with a similar background living nearby. For
example, to mask a data file containing the ages
and incomes of six people, researchers would
randomly rearrange the income levels so that
within one census block, a 21-year-old originally
listed as making $20,000 is now listed as making
$15,000, while a 50-year old making $15,000 is
now listed as making $20,000. The process
allows researchers to continue to draw valid
observations from the file, since the swapping
doesn't change the totals for each data column
within a census block.
Is anyone else worried about this in conjunction
with reverse-engineering the published data? What if my identity gets
swapped with some lowlife in the census data
for "anonymity" purposes and then the credit
company manages to hook up his profile with mine
after "figuring out" the census data?
Sounds like the potential for a nightmare
scenario...
Scientific journals already have begun to
become increasingly paperless. I refuse
to submit any of my work to journals that
do not at least have an electronic version,
and there are plenty of researchers who
submit only to electronic-only journals.
Many of the issues are obvious, like
No extra cost for beautiful color
charts and images
Easier to search from one's desk, instead
of tromping around from library to library
or ordering obscure journals.
Some electronic-only journals are
free or much less expensive than print ones.
Rob Kirby, a prominent mathematician,
has an excellent summary
of the ridiculousness journal pricing (profit margins on the order of 40%) and
it is great to see experts working to try
and straighten things out.
Gotta love Florida- they scan everyone's
image who attends the Superbowl, but can they
afford machines that actually count votes?
OS X doesn't have to run on every x86 box
on
OS X on x86?
·
· Score: 1
As has been pointed out, Apple makes money (when they do make money)
on hardware. Releasing OS X to run on cheap
hardware sold by other people would kill off
their biggest revenue stream quickly, and would
seem like a suicidal move from the stockholder
perspective (but would be great for users, so it
is nice to wish for.)
But OS X on x86 doesn't have to
be like a Windows release that is expected to
run on every x86 box. It seems like if Apple
is going to port OS X to x86, they would have
their own boxes, their own hardware, and
could specify a limited set of premium hardware
(and charge more for it)
to keep themselves from sinking their revenue
stream. In other words, they could decide that
they wanted to move processors without moving
to the the entire x86 platform. These days,
Athlons are such a good price/perf they might
think about it more than usual, but it still
seems unlikely.
If they were to move to the Athlon,
I don't seem them supporting all motherboards,
(maybe only their own?)
all legacy devices, and so on. What would be
the incentive? Why would they do extra work so that customers
don't need to buy something new, preferably
from Apple? Also remember that the idea of
a reasonably high "minimum specification" would
sit well with a company that has been putting
an emphasis on quality components all the
way through. One reason that it is hard to
compare prices Mac vs. PC is that you can't
get a Mac without Ethernet, you can't get
a general-purpose desktop Mac without the best, easiest-to-open
case there is, you can't get a Mac with
crappy video (remember those pages about
how even the very first iMacs could be dismantled
and 21" displays hung off of them?) , and so on.
Those things add up.
By the way, there was an little
blurb on sharkyextreme
(which usually ignores Mac stuff)
the other day about how
"Shocking as it may sound, many things we take for granted in today's PC,... either started at or was proven successful at Apple Computer." Well, it doesn't sound shocking
to all of us...
One great feature of Yahoo is that it is
sensibly simple in that it uses text when
it is appropriate, and not some whizbang
graphics that essentially communicate the
same info as text would. Thus, Yahoo is
useful and visitable by anyone from
Lynx users to cell phone users to the
latest greatest browsers. Google also has
a nice "low graphics" appearance but most of the
"walled garden" type portals tend to be
bloated with superfluous graphics and it
would be sad to see those dominate more
than they already do. I don't understand
why "big-time" web sites feel the need to
make over their graphics every couple
of months and needlessly complicate the
presentation of information, sigh.
This large gap in emulation performance
(667 MHz Merced about the same as 150 MHz Pentium)
is remarkable. When the first 60 MHz 601
PPC processors came out, they were slower
than the 40 MHz 68040s on emulated code but
IIRC would benchmark at least as fast as
30 MHz 68030, about a factor of two clock-
speedwise and
one architecture back (whatever that means...)
If that were the case here,
the Merced chip should rate at least a
300 Mhz P-II and it was nowhere close to that.
In Apple's case, the software 68000 emulators were
improved over time so that things got
better for a fixed processor, so perhaps there is work to be done
on the software side. I am not familiar enough
with the chip family to know how much an
improvement is feasible.
Significant processor changes always are
going to have setbacks performance-wise,
and clearly the x86 architecture is
longer overdue for an upgrade, but it
seems like emulation will give pretty
unsatisfactory results if this test was
a good indication.
There are some nice G3 and G4 clusters out there,
they are just not very cost-effective. Here's
a howto on
building a G4 cluster from a national lab, and
there are some prebuilt systems like those
running Black Lab Linux they
were showing at MacWorld New York last summer.
Evidently, there was already a comparable
tax on blank media of about 50 cents/disk in Denmark, though I
don't know if the same tax applies to
hard drives as in the French case.
Apparently the tax is currently on haitus
to resume again this spring. I believe the
amount is done in
terms of minutes of recording time.
The objections there about the presumption
of guilt sound the same as those in the
French case. The tax has spawned
widespread objection and significant
protest, which led to the current postponement.
I also remember plans for a significant
tax on blank CDR media in Canada, of
about $2.50 per CD but that has been
whittled down and has been around for a while at
about $.14/disk, not too noticable I
would think.
If you actually look at the code to express
something simple such as (a+b)^2 in MathML
it looks quite unwieldy compared to how
simple it would look in TeX. At first, this
seems like a huge turn-off for those of us who
are used to typing in TeX or HTML by hand. But
we need to remember what MathML is
trying to be: the low-level format to exchange
mathematical ideas.
The standard proposes to do lots, including:
Facilitate conversion to and from other mathematical formats, both presentational and semantic. Output
formats should include
graphical displays
speech synthesizers
input for computer algebra systems
other mathematics typesetting languages, such as TEX
plain text displays, e.g. VT100 emulators
print media, including braille
Anything which will allow input and output into Mathematica
and TeX both (let alone the others) is going to not be something that
you can not type directly by hand, so for this standard
it would be unfair to expect that. Instead,
it is important to make sure that the standard
includes the important mathematical notions that
will port from
TeX and computer algebra systems.
(to me, that means all of TeX and LaTeX except the
page-layout specific features, and most of Mathematica,
Maple, and Matlab...)
It may be that the standard is trying to do
too much or that it would only be useful to
the mathematical elite, but given the ambitious
role it is clear that the standard will need
to be complicated and presumably not suitable for unaided
digestion or production.
See the standards page here, for the 12 line code for the expression
for (a+b)^2.
The article seems to think that it is an
all-or-nothing choice for operating systems.
With the
size of modern drives, there is not a forced
choice of just one OS since it is easy to set
things up as multiple boot.
I currently use OS9, LinuxPPC and the OS X
Public Beta on a G4 and my Powerbook.
I switch regularly back and forth for different
needs and expect that many LinuxPPC users
work in the same mode. I am a big fan of
the Mac on Linux project which is great for doing
a few things under MacOS without restarting,
in a manner similar to launching the Classic
Environment under OS X.
OS X and LinuxPPC have a great
deal of functionality in common but there I don't
think that the kind of people who use LinuxPPC
are likely to abandon LinuxPPC entirely
for OS X. One issue is hardware support;
I expect that to be better for consumer-type hardware under OS X but that remains to
be seen. Printer support, for example, is
currently very weak under the Public Beta but
hopefully that will change soon. Some
of the big scientific programs that I have
compiled took a while to configure under LinuxPPC; there
doesn't seem to be any real point to
going through the configuration issues again
just to get them working under OS X instead
of LinuxPPC. Instead, I see myself continuing
to switch back and forth for various tasks.
I expect there will be more effort to port
scientific computing projects to OS X, which
will be great, but again, I don't see it as
a question of total immediate replacement.
I am dating myself, of course, but for me
the most addictive games were a number
of the Apple II classics:
Wizardry, an amazing step forward in computer RPGs
Snake Byte, too simple and clever for my good
Lode Runner, excellent gameplay
In each case, the gameplay was extremely well
done. Wizardry, in particular, with its own
spell names which are still stuck in my head only
about 15 years later (Mahalito, anyone?)
Slashdot Mirror
I registered to vote in an other county in NY more than 6 months ago (which is supposed to delete my NYC registration) but my record still appears on the site, so it is at least a few months behind or they are slow deleting re-registered voters.
For prospective elementary schoolteachers, the last math course required of them at one university in California that I taught at is a course titled "Elementary Problem Solving." The topics of that course were carefully chosen to have essentially no prerequisite knowledge of algebra or geometry (mostly basic divisibilty/primality topics, counting combinations, and pretty straightforward topics that a strong 7th grader would get the hang of in about a month.) The students there struggled spectacularly with the topics and were generally unable to manage even a first level of abstraction. We are not talking difficult problems- questions like "How many different ways can you make a sandwich if there are three different kinds of bread, four different kinds of meat and a customer can have up to two different kinds of four varieties of cheese?" More depressingly, they were in general not at all fazed by their failures, and spent more energy complaining about my unreasonable expectations that they did trying to solve problems. The general litanies I heard were "I only want to teach 2nd grade- why should I need to know any of this?" and "I need to pass this course to become a teacher, and everyone tells me I'll be a great teacher because I like kids so much!" I found teaching that course to be not particularly rewarding, and in fact, the people in my department who most often taught that course were the ones with the absolute lowest expectations of their students. The students tend to think that to teach 3rd grade math, they need only know the math that a 3rd grader learns. The idea that a teacher should understand a subject thoroughly enough to have actual insight is totally alien.
Another comment related the expanded opportunities available to women as contributing to the problem. This is very clear. In the bad old days, the acceptable careers for women were schoolteachers and nurses. Neither one of these paid well, but since the overall opportunities for women were limited, there were many bright, capable women who entered those careers, thus artificially enriching the level of teachers available for a fixed salary and prestige. Now, thankfully, there are many more opportunities for women so the bright capable ones are no longer limited to teachers and nurses- they can become engineers and lawyers and whatever else. Unfortunately, that means that who is left to go into the field but less capable people of both genders. (I figure that both the health care and education crises are complicated by this effect.) Essentially, the societal pressures limiting women to "traditionally nurturing careers" artificially reduced the cost of getting good teachers. Now, that pressure has lessened with no increase in salary or respect to compensate, so there has been an overall decline in the competence of schoolteachers.
Even with stronger requirements for math and science teachers, there is little effect. In California and New York, a reasonably competent school adminstrator can staff all math and science teachers with uncredentialled teachers-- and many are. In some parts of California, fewer than 10% of math teachers are credentialled (with a very weak credential, BTW) and the remainder have "emergency" credentials that can be extended indefinitely, with only a slight amount of administrative imagination. The great need for people thwarts any effort to raise the credential requirements, which are pretty much moot anyway.
I don't know what a good solution is but it is clear that greater resources need to be spent to improve the situation, if indeed this is something that is important to people. Everyone seems to be "for education" but given the costs of the changes that need to be made, support for significant change vanishes.
It is also interesting to see how much the house your friend bought was, when it sold, and things like that.
- When I was TAing an undergrad prog course, a prof assigned
a "greatest common substring" assignment.
After people had turned their programs in,
(this was more than 10 years ago when people
actually turned in printouts)
he mentioned that he would run their programs
looking for common substrings in their code and
long substrings would be analyzed for plagarism.
Students were given the option of withdrawing
their programs; many did.
- As a programming TA, we routinely ran
a not-very-sophisticated "cheat check" program
which looked for copied code. By comparing
object code (with debug info and variable names
stripped), we would often see code that
differed only by changing variable names.
- Sometimes the copying was ridiculously
blatant- at one institution I taught at, every
student was required to include an "honesty
pledge" at the beginning of all programs.
Oftentimes, we would find (and work to expel)
students whose programs differed only in the
honesty pledge. The most memorable was one
student who failed to delete the spaces that
arose when his name was shorter than the much-longer name of the student
he copied the program from.
- Sometimes, students are held liable if
their code is copied without their knowledge.
This thwarts the "dumpster-diving" for printouts
that can happen; students are much more
careful with their printouts if they are
responsible for someone else copying their
code, even without their knowledge.
- Before everyone had their own computers and
had to actually turn up to a lab to use VT100s, we
would often look at the 'last' logs and see if
suspiciously similar code was written by students who
were sitting next to each other in the lab. Or
sometimes if we weren't sure if something was
on the up-and-up, we could see how much time the
student spent on the program, when and from where.
More than once we had a student whose friends
were telnetting in to get the assignments from
our assignment distribution program.
- In the same vein, we could look around their
directories to see if they had written test data,
tried some other strategies, etc. to see if they
were genuine.
- These days, I make sure my exams test
the knowledge that would be obvious if you
actually wrote the code. If you didn't write your own code, you'll get hammered on the exams
and fail, even if I don't figure out where your
code came from.
Programming is vulnerable to copying. Oftentimes, for short assignments, there aren't that many possible good solutions to a problem, so repetition is not nescessarily the sign of forbidden collaboration. In general, the longer the assignment, the easier it is to spot. But over the years, there have been some amazingly boneheaded cheaters... Even if they do manage to sneak in someone else's program, almost always they get nailed on the exam. And catching a student cheating is depressing; I feel an obligation to the students who do the work to make sure that the student is expelled or disciplined as sternly as possible. Oftentimes, that is ugly and sad, but needed. Here's some advice: if you are enough of a moron to cheat and get caught, 'fess up and throw yourself on the mercy of whomever decides. It is so sad to watch a student deny the obvious and people are so much sterner when the student thinks the story is fooling someone...Working mathematicians rarely think about their proof in terms of the reduction to the primitives of set theory. Constructing proofs takes a great deal of mathematical preparation, creative energy and fluency in current results. Figuring out what level of detail is appropriate for a proof (and the audience of a proof) requires understanding what is "straightforward" and what is not. If a typical modern proof were reduced to the axioms of set theory, it would be perhaps hundreds of thousands of pages long and the key ideas that are of importance to the proof would probably be .0001% of that amount, and be made
totally opaque to any reader since the ideas
would be scattered unintelligbly throughout
the bulk which would be describing
"straightforward" mathematics in an indigestible
manner.
Caveat: there are mathematicians in a very speciallized subsegment of the highly technical field of set theory who do like to think in terms of reduction to primitive axioms. However, this is a tiny fraction of the mathematical community and arguably the wrong part of mathematics for a casual observer to try to digest. Certainly, the attention that ZFC set theory gets by informal approaches is grossly disproporionate to its use within the overall fields of research mathematics.
An analgous overly-reductionist treatment of biology would be that by understanding Schrodinger's equation, all of chemistry and thus biology is understood. It may be true that chemical understanding can be broken down into solutions of the fundamental equation of quantum mechanics, but the fact of the matter is that for the vast majority of chemical phenomena, that is not a useful approach to take. That approach would be to unwieldy to yield good understanding and new results. And for biological phenomena, which are in principle based entirely on chemical phenomena, reduction to quantum mechanics is practically never the useful approach to take.
Modern mathematics is difficult to approach since mathematics is such an old field. Almost all of the mathematics learned by a typical undergraduate is based on research that was done in the 1800s. Mathematics has flourished and continued its progress since the 1800s, of course, but for many fields, learning the prequistes (working through the 1900s, in effect) takes a great deal of energy and can take several years. In some fields it is possible to understand and appreciate more modern results, but in most fields, years of study beyond the undergraduate level are needed to get up to speed for research mathematics. Fields in science (physics, chemistry, biology, etc.) tend to be based on research that was done much more recently than the 1800s and it is much easier to learn the prerequiste material than in mathematics. There are no "easy shortcuts" to the frontier of mathematics (in general) precisely because it is an old field that has been richly developed by generations of work.
The most preposterous thing about high-priced journals is that the "value-added" part of a journal is the peer review, which is done almost always for free. When an article is submitted it is sent out for review to someone whose research is close enough to understand the work. Getting an article to review is a chore; it can take many months to thoroughly review an article, many are poorly written and have annoying minor mistakes, and there is no recognition or pay associated to it. When it turns out that the journals are priced outrageously, that is the final straw for many. In general, reviewing articles is considered a nescessary public service, and since the editors of the highest-priced journals tend to be the super-big shots, it is not easy to refuse to review something. Hopefully, things will improve! The xxx archive is great for preprints but the reviewing process is an important part of disseminating research so it will take more than that for things to get much better.
A few clarifications though: The proof is a nicer, shorter proof of a known theorem. It is important to understand the difference between a proof of a new result (where the truth or falsehood was not known beforehand) and a proof of an existing result (where it the result was known, but perhaps using a complicated method or advanced results from some other work on other questions.) New proofs of existing results are important if they are improvements of earlier proofs, as in this case, but there is a different flavor to them. Clever approaches that were not seen on the first proof can sometimes surface later; there is a remarkable example of a result (not that far removed from the field of this work) that when first proven in the 1800s, took several hundred pages- modern proofs can prove the same result in three lines (the impossibility of a finite projective plane embedded in the xy-plane authentically, if you are curious.)
Thus it is worth noting that the article mentioned appears in the American Math Monthly, which does not tend to publish hard-core new research results, but instead elegant, elementary proofs of known results. Just meant to clarify; this is a remarkable result (and let me repeat the earlier plug for the Geometer's Sketchpad- quite a remarkable piece of software...)
Printing support in the beta was great for Postscript and networked printers, but for "commodity" inkjets it was very weak, and since Apple was putting the burden on Epson, etc. to write their own drivers, it didn't happen and doesn't seem real likely to happen quickly, particularly for older inkjets. What is the incentive on the printer manufacturer's part? They'd rather sell a new printer instead of write a driver for a printer they sold 3 years ago for $100.
Check out this page for a screenshot from the Apple II if you want a little memory trip. The page also has info about running it under emulation on a PC.
Apparently, the father of the little girl is Dave Thielen, known for stirring up some trouble himself. For example, here is a copy of the page he used to pretend to sell absentee ballots on ebay last November, and it seems that perhaps the apple has not fallen far from the tree here...
Universities need a great deal of diverse IT support and cannot afford to pay salaries appropriate for trained grown-ups. Competent, curious students with interest can learn a great deal of valuable, marketable things in this setting very quickly, without the "experience" needed for comparable positions in industry. Academic tech support is generally more forgiving for gaps in expertise than the "real world" and is a good place to learn the basic necessities.
The disadvantage, from the university side, is the regular turnover as students learn enough to get excellent jobs. The disadvantage from the student side is seeing how preposterously screwed up things can be and having to put up with various politics, awful documentation and limited resources. Oh, wait- since things are often like that outside the university, maybe that is an advantage...
The danger of an entirely-student run network comes from the regular turnover, so perhaps that would be a problem. The better- run academic IT departments that I know have a number of permanent staff who were former students, and who decided that there were some nice things about the academic computing environment (or who just wanted massive bandwith to play with) and stayed around.
Is anyone else worried about this in conjunction with reverse-engineering the published data? What if my identity gets swapped with some lowlife in the census data for "anonymity" purposes and then the credit company manages to hook up his profile with mine after "figuring out" the census data? Sounds like the potential for a nightmare scenario...
2001-03-19 23:12:44 Lameness filter is kicking in.
Many of the issues are obvious, like
No extra cost for beautiful color charts and images
Quicker distribution, particularly internationally.
Generally wider distrubution
Easier to search from one's desk, instead of tromping around from library to library or ordering obscure journals.
Some electronic-only journals are free or much less expensive than print ones.
Rob Kirby, a prominent mathematician, has an excellent summary of the ridiculousness journal pricing (profit margins on the order of 40%) and it is great to see experts working to try and straighten things out.
Gotta love Florida- they scan everyone's image who attends the Superbowl, but can they afford machines that actually count votes?
But OS X on x86 doesn't have to be like a Windows release that is expected to run on every x86 box. It seems like if Apple is going to port OS X to x86, they would have their own boxes, their own hardware, and could specify a limited set of premium hardware (and charge more for it) to keep themselves from sinking their revenue stream. In other words, they could decide that they wanted to move processors without moving to the the entire x86 platform. These days, Athlons are such a good price/perf they might think about it more than usual, but it still seems unlikely.
If they were to move to the Athlon, I don't seem them supporting all motherboards, (maybe only their own?) all legacy devices, and so on. What would be the incentive? Why would they do extra work so that customers don't need to buy something new, preferably from Apple? Also remember that the idea of a reasonably high "minimum specification" would sit well with a company that has been putting an emphasis on quality components all the way through. One reason that it is hard to compare prices Mac vs. PC is that you can't get a Mac without Ethernet, you can't get a general-purpose desktop Mac without the best, easiest-to-open case there is, you can't get a Mac with crappy video (remember those pages about how even the very first iMacs could be dismantled and 21" displays hung off of them?) , and so on. Those things add up.
By the way, there was an little blurb on sharkyextreme (which usually ignores Mac stuff) the other day about how "Shocking as it may sound, many things we take for granted in today's PC, ... either started at or was proven successful at Apple Computer." Well, it doesn't sound shocking
to all of us...
One great feature of Yahoo is that it is sensibly simple in that it uses text when it is appropriate, and not some whizbang graphics that essentially communicate the same info as text would. Thus, Yahoo is useful and visitable by anyone from Lynx users to cell phone users to the latest greatest browsers. Google also has a nice "low graphics" appearance but most of the "walled garden" type portals tend to be bloated with superfluous graphics and it would be sad to see those dominate more than they already do. I don't understand why "big-time" web sites feel the need to make over their graphics every couple of months and needlessly complicate the presentation of information, sigh.
If that were the case here, the Merced chip should rate at least a 300 Mhz P-II and it was nowhere close to that. In Apple's case, the software 68000 emulators were improved over time so that things got better for a fixed processor, so perhaps there is work to be done on the software side. I am not familiar enough with the chip family to know how much an improvement is feasible.
Significant processor changes always are going to have setbacks performance-wise, and clearly the x86 architecture is longer overdue for an upgrade, but it seems like emulation will give pretty unsatisfactory results if this test was a good indication.
There are some nice G3 and G4 clusters out there, they are just not very cost-effective. Here's a howto on building a G4 cluster from a national lab, and there are some prebuilt systems like those running Black Lab Linux they were showing at MacWorld New York last summer.
The objections there about the presumption of guilt sound the same as those in the French case. The tax has spawned widespread objection and significant protest, which led to the current postponement.
I also remember plans for a significant tax on blank CDR media in Canada, of about $2.50 per CD but that has been whittled down and has been around for a while at about $.14/disk, not too noticable I would think.
The standard proposes to do lots, including:
Facilitate conversion to and from other mathematical formats, both presentational and semantic. Output formats should include
Anything which will allow input and output into Mathematica and TeX both (let alone the others) is going to not be something that you can not type directly by hand, so for this standard it would be unfair to expect that. Instead, it is important to make sure that the standard includes the important mathematical notions that will port from TeX and computer algebra systems. (to me, that means all of TeX and LaTeX except the page-layout specific features, and most of Mathematica, Maple, and Matlab...)
It may be that the standard is trying to do too much or that it would only be useful to the mathematical elite, but given the ambitious role it is clear that the standard will need to be complicated and presumably not suitable for unaided digestion or production.
See the standards page here, for the 12 line code for the expression for (a+b)^2.
OS X and LinuxPPC have a great deal of functionality in common but there I don't think that the kind of people who use LinuxPPC are likely to abandon LinuxPPC entirely for OS X. One issue is hardware support; I expect that to be better for consumer-type hardware under OS X but that remains to be seen. Printer support, for example, is currently very weak under the Public Beta but hopefully that will change soon. Some of the big scientific programs that I have compiled took a while to configure under LinuxPPC; there doesn't seem to be any real point to going through the configuration issues again just to get them working under OS X instead of LinuxPPC. Instead, I see myself continuing to switch back and forth for various tasks. I expect there will be more effort to port scientific computing projects to OS X, which will be great, but again, I don't see it as a question of total immediate replacement.
- Wizardry, an amazing step forward in computer RPGs
- Snake Byte, too simple and clever for my good
- Lode Runner, excellent gameplay
In each case, the gameplay was extremely well done. Wizardry, in particular, with its own spell names which are still stuck in my head only about 15 years later (Mahalito, anyone?)