boss, head of IT, wanted us to come up with a contingency plan for ethernet. At first we looked at each other trying to figure out what he meant. Evidently, he wanted an alternative to ethernet that still provided networking just in case ethernet failed
I sympathize with your situation because your
boss clearly doesn't have enough of a clue to
have any idea how to properly allocate
resources. But...
token ring
fddi
cddi
firewire
fibre channel
I'm pretty sure you can run TCP/IP over all
these things (and many more). So, you can
give your boss the answer, and if necessary
give him enough pricing information to
get him to shut up. (Start with fibre channel!
I want to see how much it costs to replace
the entire ethernet network, down to the
PCs on the secretaries' desks, with fibre
channel!)
Do you agree that there might be a possibility that your beliefs are not rational (again, without going into whether or not they are so)?
I think he sort of answers that question in the
essay you linked to. He says that "it is true
that Christianity is ultimately based on faith".
There are many philosophical viewpoints on what
are valid ways of obtaining knowledge. Some
people think the
only valid source of knowledge
is empirical observation and rational thought.
Others think that,
if there is such a thing as a supernatural
being, that being could impart knowledge to
people through some sort of mystical revelation,
i.e. God controls the universe, so if he wants,
he can make you know things. (It's a pretty
reasonable conclusion if you
first assume God does
exist and does control the universe.)
Some people think empirical observation and
spiritual revelation are both valid, but if
the two are in conflict, revelation should
take precedence.
These different viewpoints are differences
in philosophy. I learned in computer science
class (when I learned about diagonalization
and the incompleteness theorem) that logical
thought cannot give you all the answers.
There are true statements that logic
can't lead you to and can't support.
Many people during The Enlightenment believed
that the forward march of reason was
inexorable, and reason could, given enough
time, solve any problem. They were wrong,
although they weren't even proven wrong
until less than 100 years ago.
My point is, the question of what avenues
for obtaining knowledge are valid is an
open philosophical question. (One might
even say a timeless question.) There have
been relatively recent developments that
have changed our views of this question.
In light of that, is it wrong thinking
to believe in faith over reason? Maybe
it is, or maybe it isn't. But many
Christians have a pretty simple philosophy
on it: they believe in faith and revelation
over reason, but they do investigate
Christianity in an intellectual sense
enough to be sure that it's a defensible,
basically consistent point of view.
(It doesn't have to be perfectly consistent
and complete, because none of the other
views of the world are either.)
This creates a lot more room at the UGL/FAC for study areas, which were pretty lacking.
Agreed. This move is really a good idea. It
relocates the books to one or more of the
(literally) 17 other libraries on campus,
and the FAC gets more study and gathering area.
Which makes a lot of sense when you realize
that the FAC is where all the freshmen (who
often haven't learned about the PCL) go to
"study" anyway. And by "study", I mean
"bring textbooks so they can get together
and chat and talk on their cell phones
right next to the signs that say 'quiet
please'".
One day, the ENTIRE network goes down. Nothing will bring it back up, until someone happens to yank the power strip connected to the new machine with a 3Com network card in it. The network IMMEDIATELY comes back up. I don't know why a 3Com network card would bring down an entire network, but it DID.
It can happen easily. Ethernet works because
all the cards
figure out if there's a collision and back
off and stop transmitting if someone else
is trying to use the bus. If a card wants,
it can hog the bus.
I've seen it happen twice -- once, we had a
defective AUI to 10BaseT tranceiver (back
when Unix machines came with Ethernet but
didn't have 10BaseT plugs on the back).
The second time, it was because of a
cheap ethernet hub that went nuts. (This
was all back in the days before switches
were cheap enough for us to afford.)
That was when I was working
at a NASA facility, and the robotics group
downstairs had purchased the hub to hook
up more computers than there were ports for
in someone's office. Once the problem was
found, I, being in the systems group (a/k/a
system administration), was concerned that
if they tossed it in the trash, someone would
fish it out and plug it in again, so I asked
them to destroy it. Luckily, they had a
small workshop and were able to use tools
such as a drill press and a few blunt objects
to ensure nobody mistook it for a working
piece of electronic equipment.
I have, in my basement, a Sun 4/670MP.
[... ]
Date on the manufacturing plate: 1983
That's really impressive considering that
any Sun 4 system would be a SPARC system,
and SPARC systems weren't introduced until,
what, 1989 or 1990? And I'm fairly sure
the big multiprocessor SPARC systems like
anything in the 4/6x0 series werent
introduced until something like 1991 or
even 1992.
Is it possible the system is an upgrade from
some Sun 3 system that used a similar chassis
and the same power supply or something?
Thus, the OS will force you to upgrade your monitor to an HDCP compliant one if you want to watch HD.
When it comes to viewing on computer, this is an
exquisitely stupid idea. Why? Because right
about now (2005) is when lots of people are
becoming convinced to ditch their CRT and get
an LCD instead. Lots of businesses have already
converted, and home users are starting to do so
in numbers.
So when they try to force this on people, a
large number of people are going to react
thusly: "Buy a new monitor to view this
content?! BUT I JUST BOUGHT A NEW
MONITOR!!"
On the other hand, people seem to be
replacing their television sets more
slowly, so maybe this technology can
gain a foothold there.
The recent CAFTA trade treaty forces the banana republics of Central America (no offense intended) to adopt virtually every Copyright, Patent and Trademark law verbatim as dictated by the USA
. ..
Every country will eventually be coerced into doing the same, either with trade/financial incentives and punitive sanctions for the unwilling, or worse.
Hmm, somehow I think China may remain an
exception to this rule. They don't like
to kowtow[1] to the US anyway, and oh by the
way, they have, compared to the US,
around 4 times as many
young men who could be drafted into the
military if necessary.
Oh yes, and they also make cheap electronics
which they'd be happy to export for cash,
and I don't think doing it through the black
market or grey market would upset them very
much.
[1] Hmm, I just learned that the English
word "kowtow" comes from Chinese. How
appropriate.
I work in the building industry and the amount of stupidity i see in architects designs are beyond comprehension. But because its been designed by them, we have to find a way to build around it. the problem is architects don't work with engineers, however they call on them to "make things work".
Any time two groups of people interact, people
from one group and going to experience
stupid people from the other group, and vice
versa. My dad is an architect, and he has a
great story about how they designed a church
with a gymnasium directly above the sanctuary
because of space constraints. They worked
things out very carefully with some engineers
and chose a specific product to isolate the
gym floor so that noise from basketballs, etc.
wouldn't make its way into the sanctuary
while people were praying. (They designed it
so that both rooms could be used at once.)
They made it very clear to the builders that
no similar products should be substituted
for the specific one they'd chosen. But the
builder realized
they could pocket some money if they
specifically ignored those instructions and
put a much cheaper product in instead, so they
did.
A few months later, the church's gym floor
started "sinking". It was a couple of inches
below where it was supposed to be, and if I
recall correctly, it was uneven as well.
The church got very angry and complained to
the local TV news, which aired a story about
it. There was talk of a lawsuit.
Eventually, it was found that the builder
had caused the problem by substituting a
cheaper product, and the builder got to
replace the entire gym floor.
So, just like builders have stupid architect
stores, architects have stupid builder stories.
Because, in reality, both do stupid things
sometimes.
Who was the DUMBASS from this school's administration that decided to sell 1000 laptops for less than 1/15th of what they could have fetched on eBay?
That's a very good question. I think the CNN
article was a bit misleading in that it didn't
describe the true market value of the actual
laptops. Someone from the area has stated
that they were 500MHz G3 laptops. A little
research on eBay of completed items that
are comparable indicates they would sell
for something like $300 to $325 on average,
depending on configuration.
So, that's still $250 per laptop down the
drain. Given 1000 laptops, that's $250,000
taxpayer dollars wasted if they could've
gotten the real value for each one.
Now, having said that, it takes time and
effort to list things on eBay, and flooding
eBay with 1000 similar laptops is likely to
drive prices down. But still, there are
companies out there who buy used computers
in bulk and resell them. They probably
could have gotten maybe $200 per
computer from such a company, with no
effort at all. So they are still wasting
$150,000 even if they had gone the easy
route.
The worst part of all this is that $150,000
could pay, depending on salaries and the
cost of benefits, for a teacher for 2 or
maybe 3 years. Letting citizens get a
nice break on a laptop is neat perk, but
is it really worth taking $150,000 out of
your school district's budget? And even
if it didn't come out of the school's budget,
is it really fair to the taxpayers who had
to pay the money in but really don't need
a laptop (or already have one, etc.)? It
amounts to redistributing wealth, but in a
totally arbitrary manner. If the school
district is already so well-funded they
really don't need the money, then what they
should've done is sold the computers for a
fair value and then sent rebate checks out
to the taxpayers. Or, put it in a rainy
day fund at the very least. Or establish
a foundation, or a small scholarship.
Next up: Bullshit powered battery. John Dvorak would probably be able to fuel his entire home from the stuff he spews!
Being a geek, I would like to point out that,
while your joke is funny, there is a technical
point that needs to be made. And that is this:
actually, you can generate elecricity
from bullshit. Here's another article about it too.
(Please note that
I realize both of my sources are web sites that are mostly
full of fringe nutcases.
But, since they're
different types of fringe nutcases,
that filters out their bias and makes what
they have to say perfectly valid. (Right?))
LPs
This will continue to be a niche format. CDs provide the same quality sound playback for the human-audible range of sound.
Actually, they provide better quality playback
for the human-audible range, because they
have much lower noise.
I imagine that it might be useful if you were a dog and had to listen to ultrasonic music, otherwise... not useful.
While it's true that CDs cut off sharply
above 20 kHz and thus can't produce
ultrasonics at all, it's a misconception
that LPs don't also have high frequency
limitations. It's tempting to believe
that, because they're analog, they are
producing the sound with infinite detail,
but it's just not true: the higher frequency
sounds require smaller features in the groove,
and those small pieces are easy to wear off.
After a few playings, ultrasonics, if they
ever were present on the LP in the first place,
are gone because the ridges that correspond to
the ultrasonic frequencies
are just too tiny (and therefore
thin and weak) to stand up to the stress of
colliding with the needle. It's much the
same as the concept of keeping a really
fine edge on a knife -- the finer it is, the
sharper it is, but as you get finer and finer,
the faster you lose the fineness of the edge
you've put on there.
The bottom line is that CDs have LPs beat in
the area of
signal-to-noise ratio and they also have them
beat in frequency response. While it is
possible to hear the difference between LPs
and CDs (because they each introduce their
own kinds of distortions), it is tough to
make an argument that LPs are superior
unless it's based on a personal
preferences for the distortions you can
hear.
One of the differences between gaming and music is that music is passive while gaming actually requires your participation.
Music isn't passive if you actually listen
to it. At least, good music isn't. Often,
there are all kinds of interesting things going
on, and in some cases it either requires natural
talent or (as in my case personally, since
I have very little natural talent) years
of training
just to be able to appreciate certain things
the composer throws in.
Correct me if I'm wrong, but :
Traffic shaping only affects UPSTREAM data.
Well, not necessarily. The crux of the
idea is this: TCP adjusts the rate it
sends based on whether (and when) it
receives ACKs back for things it sent.
And, two things can affect whether it
receives ACKs: packet loss in the
direction it's sending, or packet loss
in the direction the ACKs are going.
To make this more concrete, let's imagine
a scenario. You are at home, you have
ADSL, and you're downloading an ISO of
your favorite Linux distribution from a
web server. The web server is using a
TCP connection, and the TCP connection
needs to reduce its rate if it's sending
too many pieces of that Linux ISO too
quickly. But, it doesn't know what packets
you got -- it only knows what ACKs you
sent back. So, if it is missing some
ACKs, it assumes that this
is because it is sending too much
stuff and saturating the link. So it
sends less. But what if its assumption
is wrong? What if it is the upstream
link that is saturated and not the downstream?
Then there is no problem at all receiving
the packets from the web server, but the
web server falsely believes the downstream
link is saturated because it is missing
some of its ACKs. Therefore, it slows
down the rate it transmits stuff more than
it really needs to.
One could imagine a protocol that accounts
for this: the two endpoints could not only
send ACKs but could also keep track of
packets that were retransmitted unnecessarily.
From that, you could infer a certain amount
of information about where packets are being
dropped: if one end retransmits packets
unnecessarily, it implies that quite
possibly ACKs
are getting lost rather than data packets.
And it would be OK, in certain circumstances,
for the sender to increase the rate it sends
a little bit to compensate for this. But it
seems like there could be some hidden problems
with such a system, and to my knowledge TCP
doesn't try to be really smart like this
anyway.
So, this is where traffic shaping comes in.
If you are trying to download stuff and
the upstream end of your link is saturated,
you will have to drop ACKs, and as a result
your download will slow down.
So, the solution is this: (a) put a bandwidth
limitation so that you never saturate the
upstream link and instead queue things
on the machine you control (not the
cable modem, DSL modem, etc.), and (b)
prioritize ACKs. When you do these two
things, since you have never saturated the
link, you can always almost instantly
transmit ACKs, which means your downloads
go faster than if ACKs have to wait or are
lost.
It's not a perfect system because you
can't control the entire path, and some
other piece might be saturated. But
with the limited upstream bandwidth
available on DSL and such, it could be
really helpful in lots of cases.
I have a hard time believing that using a PC to do routing will be faster than using custom hardware.
Who on earth said anything about it being faster?
My guess is that the performance difference
between dedicated hardware and PC is quite
negligible.
All the article said was that you can do
quality-of-service queueing with regular
PC hardware pretty easily,
and that if you already have
a spare PC, that's cheaper than buying
dedicated hardware for the purpose.
As far as I know, the article didn't
claim that a PC was any better at the job
than dedicated hardware; it just claimed
that routing with quality-of-service is
better than routing with it.
Getting back to the subject of performance
for a moment, the low bandwidth involved
in most home network connections (even if
they are called "broadband") is so small
that just about any computer that can run
current software can handle it. Remember,
computers are built to move hundreds of
megabytes around in a second, and we are
only talking about fractions of a megabyte.
That leaves only the issue of latency.
But, on regular 10 megabit ethernet, a
full-size 1536 byte frame can transmit
in not much longer than 1 ms. Lots of
fancy routing hardware is built so it
can figure out how to retransmit a
packet before the entire packet has
been received, so a router could in
theory add less than 1 ms of latency
on 10 megabit ethernet.
But even if your PC receives the whole
packet and then waits a whole extra
millisecond before starting to retransmit
it, you've still only added 2 ms of latency,
which is really not that much. At least,
on my cable modem, if I ping the local
university, my latency averages about 125 ms.
How big of a deal is it if I had 2 ms to that?
For what it's worth, I just did two ping
tests to test how much latency going through
a PC does add. I first pinged the local
university from my firewall machine, which
is a 600 MHz Athlon running Solaris 8.
The lowest ping time recorded was 9 ms. Then, I pinged the same machine
from a Mac that sits inside the firewall,
so that the Solaris 8 machine is routing
the ICMP packets it was (in the previous
test) originating. The result? The lowest
ping time recorded for the Mac going through
the Solaris machine was 9.178 ms. It's
hard to say since the Solaris machine doesn't
measure in fractions of a millisecond, but
the point is that ping times were not
increased dramatically. In fact, it
appears to be less than one
millisecond difference.
A mere century ago, the usual age for marraige in most cultures was 12 to 16. Can you explain to me what has changed from that time, besides the views of society?
One very important change is that the average
life expectancy has increased by 50% or even
100% in some cases in the last century.
And the thing is, if you only expect to
live to 35 or 40 or something, then you
need to breed early so you'll be around
to take care of your kids until they're
old enough to take care of themselves.
If you scale everything
linearly, then a 50% increase in life
expectancy would make
the average age for
marriage more like 18 to 24, and
a 100% increase in life expectancy would
mean an average age for marriage of
24 to 36.
Sex is just a physical activity, the very same as sports.
Which is why if one person is mugged and
another is raped, the person who is raped
has no more lasting emotional issues than
the person who gets hit in the face, right?
Oh wait, that's not true at all. I've
known girls who have been raped, and it has
a profound effect on them years,
maybe even a decade later. I've also known
people who've been mugged, and it's more like
"bummer", and then a month later, it's
mostly forgotten.
I don't mind if you believe that consenting
adults should be allowed to do whatever
they want sexually. I believe that too.
I also don't mind if you think the government
shouldn't get overly involved in this stuff.
I, once again, agree with that.
But let's not pretend there isn't an extra
level of psychological and emotional
complexity involved in sex, because that's
just not factual.
Why would they care about timezones? All they know about is time(). Things like cron just know how to deal with time() suddenly going back or forth an hour.
That's true. However, the implementation of
time() resides in libc, which is a
shared library, which is loaded when the
executable itself is loaded. So now we are
faced with another question: does the
implementation of time() that
resides in libc know how to deal
with a situation where the file that defines
the timezone (/usr/share/lib/zoneinfo/US/Central
in my case) changes after libc is
loaded? Will that implementation in
the library re-read the file as necessary,
or does it just read it once when it first
needs it (or when it loads)?
If it doesn't re-read the file when it
changes (and there would be little reason
to design it so that it does, since timezone
files don't change often), then every
executable that uses time()
will have to be restarted so that it can
re-read that file. It's possible that
time() will respond properly, but how will
you know the answer to that?
And that brings up another question: can
you make a list of all programs that use
time() and will therefore need to
be restarted? If you do make a list,
are you sure you got everything?
IMHO, the safe thing to do is just reboot.
There may be a way to do this without
rebooting, but it's a one-time thing, so
you don't have the opportunity to gain
the experience to know whether it's
necessary or not.
I have about 54 Unix servers to care and feed.
I estimate that we will perform zero patches to handle these specific rule changes.
I know, you only have to change the file
that defines the timezone, and there is no
need to change executable code or anything.
So, it seemingly isn't that big a deal.
But, let me ask you a question:
although you plan to patch zero
machines, how many machines will you
reboot?
The problem, for me, is this: I have
been using Unix for over 15 years now,
and for about 5 years I was a Unix
administrator. I consider myself pretty
knowledge about Unix. However, I have
NO IDEA whether the important commands
(cron and so on) re-read the
timezone description periodically or
whether a reboot is required. I've
simply never run into that issue.
So, I don't know whether a reboot is
required, so if I ran 54 Unix servers,
I suppose I would have to reboot them
all just to be on the safe side.
Now, it's not a HUGE productivity loss
to have to reboot a bunch of servers,
but it is a definite waste of time.
All in all, this modified daylight
saving time idea just seems really dumb
to me.
So, umm, how long - exactly - does it take to upload 560 GB over a broadband connection?
Actually, you'd better make that 560 + 560 GB because I may want to back up my OTHER PC as well.
Do you reformat the hard drive and install
a fresh 560GB of data every day onto it?
If not, then incremental backups could
make the requirement a couple of orders
of magnitude lower. Chances are, you
only modify a few megabytes or hundreds
of megabytes a day, so that is all that
needs to be backed up.
To look at it a different way, most new
data created on your computer probably
comes from a network. If your upstream
bandwidth is 1/10th of your downstream
bandwidth, then as long as you aren't
downloading more than 10% of the time,
you should be able to re-upload everything
you download. For many users, we can
assume you're asleep 1/3 of the time and
gone to work or otherwise away from the
computer for another 1/3 of the time. This
leaves probably about 1/3 of the time that
you're actually using the computer, so
it seems possible you're not downloading
more than 1/10th of the maximum that you
could download if you downloaded 24 hours
a day.
Of course, this probably will not apply to
everyone. If you primarily use your computer
to record music, or to record TV programs
off cable, or if you constantly leave a
bittorrent client running to download new
stuff, you may not have enough bandwidth
to upload everything you download. But
many people will, so online backups could
still be a feasible option for most people.
IT Managers get zero chicks. Lifeguards get tons of chicks. What happens when then two are combined in the same person?
You get the curious effect of a "chick ring" as the attractive and repulsive forces reach an equilibrium somewhere just out of arms reach. From there they tend to fall into a stable orbit.
Note to self: do whatever is necessary
to develop a small net positive
attractiveness to chicks, then find someone
who is both a lifeguard and an IT manager
and hang out with them a LOT. Since there
will be lots of chicks around them which
can be easily stripped (*cough*) out of
their orbit, I should get plenty of
chicks this way.
I agree. The state do-not-call lists are
redundant. And, that's a good thing. It's a
good thing for the same reason it's good when
you have redundant power supplies
or mirrored drives on an important server:
something could prevent one of the redundant
pieces from doing its job, so it's nice to
have a backup.
What I mean is that if in some fit of
stupidity, the congress or the FCC or
the courts invalidate the federal do-not-call
list, the state ones still give me the option
of making my wishes known -- and respected.
And believe me, I really do not want to be
called. I work at home. I have a hard
enough time concentrating and a short
enough attention span as it is. Having
telemarketers call and interrupt my train
of thought 5 times a day or whatever was
seriously intrusive and bothersome. Having
this interruption removed is important for
the smooth functioning of my brain and thus
my business.
To say that depression is a simply physiological disorder is misleading at best. Since all mental function is grounded in the biology of the brain, any mental state can be affected through a physical intervention, but that doesn't mean the state is "purely biological" or "caused" by brain function.
I agree that it's misleading to say it's
purely physiological. But, it's misleading
to say that the cause is always thought
patterns, i.e. that it's purely psychological.
This is the way that I think of depression:
the mood that people associate with
depression comes about because of a
lack of the neurotransmitter serotonin.
That is the immediate cause of the bad
feelings. But, what causes your brain
to be low on its supply of serotonin?
It can be, fundamentally, one of two
things: either something is causing your
brain to use it up too quickly, or not
enough of it is produced in the first place.
In most cases, using it up too quickly
is going to happen because of psychological
issues. One thing that causes you to use
up serotonin is stress. You can thus
enter a depressed state of mind for
purely psychological reasons, such as
being upset about the loss of someone
you love, dealing with the stress of a
difficult situation, or even stuff like
being a dumbass and getting yourself into
bad situations frequently.
However, it's also possible that your
body just doesn't produce enough
serotonin in the first place. Thus,
you are always in short supply, and
stressful events that would affect
other people only in minor ways cause
you to become dangerously low on
serotonin because you didn't have that
much in the first place.
To make matters a little more cloudy,
a depressed state of mind can lead to
behavior that causes stressful situations,
which can then cause depletion of serotonin
and perpetuate the depression. So, you
can be stuck in a bad state that lasts
a long time even if there is nothing
really wrong with your physiology.
All in all, I think most people who
experience depression experience a
temporary state that is caused by
psychological issues, and it may linger
for a while because it's hard to break
the cycle, but it doesn't represent
a physiological abnormality. But I do think
there are people out there who have a
physiological abnormality relating to brain
chemistry. They are in the minority, but
they are out there.
On the subject of psychological therapy,
I agree that it can help. In the case of
people whose problem isn't fundamentally
caused by something physical, it is probably
the key to ending the depression (assuming
the key isn't getting them out of or changing
a bad situation they're in!). In the case
of people whose depression is caused by a
physical issue, it still can be helpful,
because the mood is basically an
equilibrium that is determined by several
factors, and things like how you deal
with stuff (psychologically) can shift
the equilibrium.
But, in neither case is it helpful to
misunderstand the fundamental cause.
For the person whose problem starts
with pscyhological issues, it isn't
helpful to just give them drugs and
forgo the counseling that they need
to solve the problems and be fine
without drugs. On the other side of
the coin, for the person whose
problem starts with biological issues,
it isn't helpful to pretend that drugs
are a crutch and that, if they only
worked harder in therapy, they would
be able to "snap out of it" and they'd
be fine when that's not true.
On the subject of drugs, I think drugs
can be useful for either class of person
suffering from depression. For the
person who can have a stable mood if
they can only climb out of the temporary
state they're in because of, basically,
a chemical feedback loop, drugs can
temporarily boost serotonin levels
enough give them more breathing room to
work on their problems. For the person
whose depression is physiological in
origin, the drugs may not correct the
exact problem but they can make the
situation better. (For example, SSRIs don't
promote production of serotonin -- they
only prevent the body from disposing of
"excess" quantities, and thus shift
the balance.)
Actually, it's kind of a sad idea. The
person who came up with it seems to be
thinking in narrow terms. They seem
to have started with the assumption that
drag and drop, as it works already, is the
best way to do things so that only
evolutionary changes need to made, and not
any revolutionary changes. Specifically,
they seem to be laboring under the assumption
that you can't temporarily let go of the
thing you're dragging by releasing the
mouse button.
As a result, they've solved a problem (that
you can't drag and drop between two locations
if one of them is obscured) by making you
do a complicated, error-prone mouse gesture
that requires fine positional movements of the
mouse all
while you're holding down the button (which
makes it that much harder because you have
to maintain force on the button while changing
force on other parts of the mouse to move it,
all with one hand).
If they weren't stuck in the paradigm of
traditional drag and drop, they might've
suggested a much easier to learn and
much simpler to execute solution: create
a place on the screen that can temporarily
hold whatever you're trying to drag and
drop, something that will hold onto whatever
you're dragging while you're doing whatever
you need to (rearranging windows, or maybe
something else). Basically, a second hand.
In fact, I would use a hand as the icon for it.
If you have a second hand, you can locate
the original location, drag the thing from
there to the other hand,
then locate the desired destination, and
then drag it from the other hand to
where you ultimately want it to go.
Just like you do in real life when you
want to move something but you have to
manipulate other things between picking it up
and putting it down. Nobody in the real
world flips through stuff with the same
hand they're holding things with if they
have the other hand free.
Being one of the resident curmudgeonly grammar
freaks, I feel inclined to point out that
"layoff" is a noun, and "lay off" is a verb
(phrase). Therefore, it isn't possible to
"layoff" someone; instead, you must lay them
off.
Slashdot Mirror
I sympathize with your situation because your boss clearly doesn't have enough of a clue to have any idea how to properly allocate resources. But...
I'm pretty sure you can run TCP/IP over all these things (and many more). So, you can give your boss the answer, and if necessary give him enough pricing information to get him to shut up. (Start with fibre channel! I want to see how much it costs to replace the entire ethernet network, down to the PCs on the secretaries' desks, with fibre channel!)
Not exactly. Sonar ("SOund Navigation And Ranging") uses sound but radar ("RAdio Detection And Ranging") uses radio.
I think he sort of answers that question in the essay you linked to. He says that "it is true that Christianity is ultimately based on faith".
There are many philosophical viewpoints on what are valid ways of obtaining knowledge. Some people think the only valid source of knowledge is empirical observation and rational thought. Others think that, if there is such a thing as a supernatural being, that being could impart knowledge to people through some sort of mystical revelation, i.e. God controls the universe, so if he wants, he can make you know things. (It's a pretty reasonable conclusion if you first assume God does exist and does control the universe.) Some people think empirical observation and spiritual revelation are both valid, but if the two are in conflict, revelation should take precedence.
These different viewpoints are differences in philosophy. I learned in computer science class (when I learned about diagonalization and the incompleteness theorem) that logical thought cannot give you all the answers. There are true statements that logic can't lead you to and can't support. Many people during The Enlightenment believed that the forward march of reason was inexorable, and reason could, given enough time, solve any problem. They were wrong, although they weren't even proven wrong until less than 100 years ago.
My point is, the question of what avenues for obtaining knowledge are valid is an open philosophical question. (One might even say a timeless question.) There have been relatively recent developments that have changed our views of this question.
In light of that, is it wrong thinking to believe in faith over reason? Maybe it is, or maybe it isn't. But many Christians have a pretty simple philosophy on it: they believe in faith and revelation over reason, but they do investigate Christianity in an intellectual sense enough to be sure that it's a defensible, basically consistent point of view. (It doesn't have to be perfectly consistent and complete, because none of the other views of the world are either.)
Agreed. This move is really a good idea. It relocates the books to one or more of the (literally) 17 other libraries on campus, and the FAC gets more study and gathering area. Which makes a lot of sense when you realize that the FAC is where all the freshmen (who often haven't learned about the PCL) go to "study" anyway. And by "study", I mean "bring textbooks so they can get together and chat and talk on their cell phones right next to the signs that say 'quiet please'".
It can happen easily. Ethernet works because all the cards figure out if there's a collision and back off and stop transmitting if someone else is trying to use the bus. If a card wants, it can hog the bus.
I've seen it happen twice -- once, we had a defective AUI to 10BaseT tranceiver (back when Unix machines came with Ethernet but didn't have 10BaseT plugs on the back).
The second time, it was because of a cheap ethernet hub that went nuts. (This was all back in the days before switches were cheap enough for us to afford.) That was when I was working at a NASA facility, and the robotics group downstairs had purchased the hub to hook up more computers than there were ports for in someone's office. Once the problem was found, I, being in the systems group (a/k/a system administration), was concerned that if they tossed it in the trash, someone would fish it out and plug it in again, so I asked them to destroy it. Luckily, they had a small workshop and were able to use tools such as a drill press and a few blunt objects to ensure nobody mistook it for a working piece of electronic equipment.
That's really impressive considering that any Sun 4 system would be a SPARC system, and SPARC systems weren't introduced until, what, 1989 or 1990? And I'm fairly sure the big multiprocessor SPARC systems like anything in the 4/6x0 series werent introduced until something like 1991 or even 1992.
Is it possible the system is an upgrade from some Sun 3 system that used a similar chassis and the same power supply or something?
When it comes to viewing on computer, this is an exquisitely stupid idea. Why? Because right about now (2005) is when lots of people are becoming convinced to ditch their CRT and get an LCD instead. Lots of businesses have already converted, and home users are starting to do so in numbers.
So when they try to force this on people, a large number of people are going to react thusly: "Buy a new monitor to view this content?! BUT I JUST BOUGHT A NEW MONITOR!!"
On the other hand, people seem to be replacing their television sets more slowly, so maybe this technology can gain a foothold there.
Hmm, somehow I think China may remain an exception to this rule. They don't like to kowtow[1] to the US anyway, and oh by the way, they have, compared to the US, around 4 times as many young men who could be drafted into the military if necessary.
Oh yes, and they also make cheap electronics which they'd be happy to export for cash, and I don't think doing it through the black market or grey market would upset them very much.
[1] Hmm, I just learned that the English word "kowtow" comes from Chinese. How appropriate.
Any time two groups of people interact, people from one group and going to experience stupid people from the other group, and vice versa. My dad is an architect, and he has a great story about how they designed a church with a gymnasium directly above the sanctuary because of space constraints. They worked things out very carefully with some engineers and chose a specific product to isolate the gym floor so that noise from basketballs, etc. wouldn't make its way into the sanctuary while people were praying. (They designed it so that both rooms could be used at once.)
They made it very clear to the builders that no similar products should be substituted for the specific one they'd chosen. But the builder realized they could pocket some money if they specifically ignored those instructions and put a much cheaper product in instead, so they did.
A few months later, the church's gym floor started "sinking". It was a couple of inches below where it was supposed to be, and if I recall correctly, it was uneven as well. The church got very angry and complained to the local TV news, which aired a story about it. There was talk of a lawsuit. Eventually, it was found that the builder had caused the problem by substituting a cheaper product, and the builder got to replace the entire gym floor.
So, just like builders have stupid architect stores, architects have stupid builder stories. Because, in reality, both do stupid things sometimes.
That's a very good question. I think the CNN article was a bit misleading in that it didn't describe the true market value of the actual laptops. Someone from the area has stated that they were 500MHz G3 laptops. A little research on eBay of completed items that are comparable indicates they would sell for something like $300 to $325 on average, depending on configuration.
So, that's still $250 per laptop down the drain. Given 1000 laptops, that's $250,000 taxpayer dollars wasted if they could've gotten the real value for each one.
Now, having said that, it takes time and effort to list things on eBay, and flooding eBay with 1000 similar laptops is likely to drive prices down. But still, there are companies out there who buy used computers in bulk and resell them. They probably could have gotten maybe $200 per computer from such a company, with no effort at all. So they are still wasting $150,000 even if they had gone the easy route.
The worst part of all this is that $150,000 could pay, depending on salaries and the cost of benefits, for a teacher for 2 or maybe 3 years. Letting citizens get a nice break on a laptop is neat perk, but is it really worth taking $150,000 out of your school district's budget? And even if it didn't come out of the school's budget, is it really fair to the taxpayers who had to pay the money in but really don't need a laptop (or already have one, etc.)? It amounts to redistributing wealth, but in a totally arbitrary manner. If the school district is already so well-funded they really don't need the money, then what they should've done is sold the computers for a fair value and then sent rebate checks out to the taxpayers. Or, put it in a rainy day fund at the very least. Or establish a foundation, or a small scholarship.
Being a geek, I would like to point out that, while your joke is funny, there is a technical point that needs to be made. And that is this: actually, you can generate elecricity from bullshit. Here's another article about it too.
(Please note that I realize both of my sources are web sites that are mostly full of fringe nutcases. But, since they're different types of fringe nutcases, that filters out their bias and makes what they have to say perfectly valid. (Right?))
Actually, they provide better quality playback for the human-audible range, because they have much lower noise.
While it's true that CDs cut off sharply above 20 kHz and thus can't produce ultrasonics at all, it's a misconception that LPs don't also have high frequency limitations. It's tempting to believe that, because they're analog, they are producing the sound with infinite detail, but it's just not true: the higher frequency sounds require smaller features in the groove, and those small pieces are easy to wear off. After a few playings, ultrasonics, if they ever were present on the LP in the first place, are gone because the ridges that correspond to the ultrasonic frequencies are just too tiny (and therefore thin and weak) to stand up to the stress of colliding with the needle. It's much the same as the concept of keeping a really fine edge on a knife -- the finer it is, the sharper it is, but as you get finer and finer, the faster you lose the fineness of the edge you've put on there.
The bottom line is that CDs have LPs beat in the area of signal-to-noise ratio and they also have them beat in frequency response. While it is possible to hear the difference between LPs and CDs (because they each introduce their own kinds of distortions), it is tough to make an argument that LPs are superior unless it's based on a personal preferences for the distortions you can hear.
Music isn't passive if you actually listen to it. At least, good music isn't. Often, there are all kinds of interesting things going on, and in some cases it either requires natural talent or (as in my case personally, since I have very little natural talent) years of training just to be able to appreciate certain things the composer throws in.
Well, not necessarily. The crux of the idea is this: TCP adjusts the rate it sends based on whether (and when) it receives ACKs back for things it sent. And, two things can affect whether it receives ACKs: packet loss in the direction it's sending, or packet loss in the direction the ACKs are going.
To make this more concrete, let's imagine a scenario. You are at home, you have ADSL, and you're downloading an ISO of your favorite Linux distribution from a web server. The web server is using a TCP connection, and the TCP connection needs to reduce its rate if it's sending too many pieces of that Linux ISO too quickly. But, it doesn't know what packets you got -- it only knows what ACKs you sent back. So, if it is missing some ACKs, it assumes that this is because it is sending too much stuff and saturating the link. So it sends less. But what if its assumption is wrong? What if it is the upstream link that is saturated and not the downstream? Then there is no problem at all receiving the packets from the web server, but the web server falsely believes the downstream link is saturated because it is missing some of its ACKs. Therefore, it slows down the rate it transmits stuff more than it really needs to.
One could imagine a protocol that accounts for this: the two endpoints could not only send ACKs but could also keep track of packets that were retransmitted unnecessarily. From that, you could infer a certain amount of information about where packets are being dropped: if one end retransmits packets unnecessarily, it implies that quite possibly ACKs are getting lost rather than data packets. And it would be OK, in certain circumstances, for the sender to increase the rate it sends a little bit to compensate for this. But it seems like there could be some hidden problems with such a system, and to my knowledge TCP doesn't try to be really smart like this anyway.
So, this is where traffic shaping comes in. If you are trying to download stuff and the upstream end of your link is saturated, you will have to drop ACKs, and as a result your download will slow down. So, the solution is this: (a) put a bandwidth limitation so that you never saturate the upstream link and instead queue things on the machine you control (not the cable modem, DSL modem, etc.), and (b) prioritize ACKs. When you do these two things, since you have never saturated the link, you can always almost instantly transmit ACKs, which means your downloads go faster than if ACKs have to wait or are lost.
It's not a perfect system because you can't control the entire path, and some other piece might be saturated. But with the limited upstream bandwidth available on DSL and such, it could be really helpful in lots of cases.
Who on earth said anything about it being faster? My guess is that the performance difference between dedicated hardware and PC is quite negligible. All the article said was that you can do quality-of-service queueing with regular PC hardware pretty easily, and that if you already have a spare PC, that's cheaper than buying dedicated hardware for the purpose. As far as I know, the article didn't claim that a PC was any better at the job than dedicated hardware; it just claimed that routing with quality-of-service is better than routing with it.
Getting back to the subject of performance for a moment, the low bandwidth involved in most home network connections (even if they are called "broadband") is so small that just about any computer that can run current software can handle it. Remember, computers are built to move hundreds of megabytes around in a second, and we are only talking about fractions of a megabyte.
That leaves only the issue of latency. But, on regular 10 megabit ethernet, a full-size 1536 byte frame can transmit in not much longer than 1 ms. Lots of fancy routing hardware is built so it can figure out how to retransmit a packet before the entire packet has been received, so a router could in theory add less than 1 ms of latency on 10 megabit ethernet. But even if your PC receives the whole packet and then waits a whole extra millisecond before starting to retransmit it, you've still only added 2 ms of latency, which is really not that much. At least, on my cable modem, if I ping the local university, my latency averages about 125 ms. How big of a deal is it if I had 2 ms to that?
For what it's worth, I just did two ping tests to test how much latency going through a PC does add. I first pinged the local university from my firewall machine, which is a 600 MHz Athlon running Solaris 8. The lowest ping time recorded was 9 ms. Then, I pinged the same machine from a Mac that sits inside the firewall, so that the Solaris 8 machine is routing the ICMP packets it was (in the previous test) originating. The result? The lowest ping time recorded for the Mac going through the Solaris machine was 9.178 ms. It's hard to say since the Solaris machine doesn't measure in fractions of a millisecond, but the point is that ping times were not increased dramatically. In fact, it appears to be less than one millisecond difference.
One very important change is that the average life expectancy has increased by 50% or even 100% in some cases in the last century. And the thing is, if you only expect to live to 35 or 40 or something, then you need to breed early so you'll be around to take care of your kids until they're old enough to take care of themselves. If you scale everything linearly, then a 50% increase in life expectancy would make the average age for marriage more like 18 to 24, and a 100% increase in life expectancy would mean an average age for marriage of 24 to 36.
Which is why if one person is mugged and another is raped, the person who is raped has no more lasting emotional issues than the person who gets hit in the face, right? Oh wait, that's not true at all. I've known girls who have been raped, and it has a profound effect on them years, maybe even a decade later. I've also known people who've been mugged, and it's more like "bummer", and then a month later, it's mostly forgotten.
I don't mind if you believe that consenting adults should be allowed to do whatever they want sexually. I believe that too. I also don't mind if you think the government shouldn't get overly involved in this stuff. I, once again, agree with that. But let's not pretend there isn't an extra level of psychological and emotional complexity involved in sex, because that's just not factual.
Maybe you were a teenager a long time before I was, but back in the 1980's when I was teenager, telephones were electronic.
That's true. However, the implementation of time() resides in libc, which is a shared library, which is loaded when the executable itself is loaded. So now we are faced with another question: does the implementation of time() that resides in libc know how to deal with a situation where the file that defines the timezone (/usr/share/lib/zoneinfo/US/Central in my case) changes after libc is loaded? Will that implementation in the library re-read the file as necessary, or does it just read it once when it first needs it (or when it loads)?
If it doesn't re-read the file when it changes (and there would be little reason to design it so that it does, since timezone files don't change often), then every executable that uses time() will have to be restarted so that it can re-read that file. It's possible that time() will respond properly, but how will you know the answer to that?
And that brings up another question: can you make a list of all programs that use time() and will therefore need to be restarted? If you do make a list, are you sure you got everything?
IMHO, the safe thing to do is just reboot. There may be a way to do this without rebooting, but it's a one-time thing, so you don't have the opportunity to gain the experience to know whether it's necessary or not.
I know, you only have to change the file that defines the timezone, and there is no need to change executable code or anything. So, it seemingly isn't that big a deal.
But, let me ask you a question: although you plan to patch zero machines, how many machines will you reboot?
The problem, for me, is this: I have been using Unix for over 15 years now, and for about 5 years I was a Unix administrator. I consider myself pretty knowledge about Unix. However, I have NO IDEA whether the important commands (cron and so on) re-read the timezone description periodically or whether a reboot is required. I've simply never run into that issue. So, I don't know whether a reboot is required, so if I ran 54 Unix servers, I suppose I would have to reboot them all just to be on the safe side.
Now, it's not a HUGE productivity loss to have to reboot a bunch of servers, but it is a definite waste of time. All in all, this modified daylight saving time idea just seems really dumb to me.
Do you reformat the hard drive and install a fresh 560GB of data every day onto it? If not, then incremental backups could make the requirement a couple of orders of magnitude lower. Chances are, you only modify a few megabytes or hundreds of megabytes a day, so that is all that needs to be backed up.
To look at it a different way, most new data created on your computer probably comes from a network. If your upstream bandwidth is 1/10th of your downstream bandwidth, then as long as you aren't downloading more than 10% of the time, you should be able to re-upload everything you download. For many users, we can assume you're asleep 1/3 of the time and gone to work or otherwise away from the computer for another 1/3 of the time. This leaves probably about 1/3 of the time that you're actually using the computer, so it seems possible you're not downloading more than 1/10th of the maximum that you could download if you downloaded 24 hours a day.
Of course, this probably will not apply to everyone. If you primarily use your computer to record music, or to record TV programs off cable, or if you constantly leave a bittorrent client running to download new stuff, you may not have enough bandwidth to upload everything you download. But many people will, so online backups could still be a feasible option for most people.
Note to self: do whatever is necessary to develop a small net positive attractiveness to chicks, then find someone who is both a lifeguard and an IT manager and hang out with them a LOT. Since there will be lots of chicks around them which can be easily stripped (*cough*) out of their orbit, I should get plenty of chicks this way.
I agree. The state do-not-call lists are redundant. And, that's a good thing. It's a good thing for the same reason it's good when you have redundant power supplies or mirrored drives on an important server: something could prevent one of the redundant pieces from doing its job, so it's nice to have a backup.
What I mean is that if in some fit of stupidity, the congress or the FCC or the courts invalidate the federal do-not-call list, the state ones still give me the option of making my wishes known -- and respected.
And believe me, I really do not want to be called. I work at home. I have a hard enough time concentrating and a short enough attention span as it is. Having telemarketers call and interrupt my train of thought 5 times a day or whatever was seriously intrusive and bothersome. Having this interruption removed is important for the smooth functioning of my brain and thus my business.
I agree that it's misleading to say it's purely physiological. But, it's misleading to say that the cause is always thought patterns, i.e. that it's purely psychological.
This is the way that I think of depression: the mood that people associate with depression comes about because of a lack of the neurotransmitter serotonin. That is the immediate cause of the bad feelings. But, what causes your brain to be low on its supply of serotonin? It can be, fundamentally, one of two things: either something is causing your brain to use it up too quickly, or not enough of it is produced in the first place.
In most cases, using it up too quickly is going to happen because of psychological issues. One thing that causes you to use up serotonin is stress. You can thus enter a depressed state of mind for purely psychological reasons, such as being upset about the loss of someone you love, dealing with the stress of a difficult situation, or even stuff like being a dumbass and getting yourself into bad situations frequently.
However, it's also possible that your body just doesn't produce enough serotonin in the first place. Thus, you are always in short supply, and stressful events that would affect other people only in minor ways cause you to become dangerously low on serotonin because you didn't have that much in the first place.
To make matters a little more cloudy, a depressed state of mind can lead to behavior that causes stressful situations, which can then cause depletion of serotonin and perpetuate the depression. So, you can be stuck in a bad state that lasts a long time even if there is nothing really wrong with your physiology.
All in all, I think most people who experience depression experience a temporary state that is caused by psychological issues, and it may linger for a while because it's hard to break the cycle, but it doesn't represent a physiological abnormality. But I do think there are people out there who have a physiological abnormality relating to brain chemistry. They are in the minority, but they are out there.
On the subject of psychological therapy, I agree that it can help. In the case of people whose problem isn't fundamentally caused by something physical, it is probably the key to ending the depression (assuming the key isn't getting them out of or changing a bad situation they're in!). In the case of people whose depression is caused by a physical issue, it still can be helpful, because the mood is basically an equilibrium that is determined by several factors, and things like how you deal with stuff (psychologically) can shift the equilibrium.
But, in neither case is it helpful to misunderstand the fundamental cause. For the person whose problem starts with pscyhological issues, it isn't helpful to just give them drugs and forgo the counseling that they need to solve the problems and be fine without drugs. On the other side of the coin, for the person whose problem starts with biological issues, it isn't helpful to pretend that drugs are a crutch and that, if they only worked harder in therapy, they would be able to "snap out of it" and they'd be fine when that's not true.
On the subject of drugs, I think drugs can be useful for either class of person suffering from depression. For the person who can have a stable mood if they can only climb out of the temporary state they're in because of, basically, a chemical feedback loop, drugs can temporarily boost serotonin levels enough give them more breathing room to work on their problems. For the person whose depression is physiological in origin, the drugs may not correct the exact problem but they can make the situation better. (For example, SSRIs don't promote production of serotonin -- they only prevent the body from disposing of "excess" quantities, and thus shift the balance.)
Yes.
Nope.
Actually, it's kind of a sad idea. The person who came up with it seems to be thinking in narrow terms. They seem to have started with the assumption that drag and drop, as it works already, is the best way to do things so that only evolutionary changes need to made, and not any revolutionary changes. Specifically, they seem to be laboring under the assumption that you can't temporarily let go of the thing you're dragging by releasing the mouse button.
As a result, they've solved a problem (that you can't drag and drop between two locations if one of them is obscured) by making you do a complicated, error-prone mouse gesture that requires fine positional movements of the mouse all while you're holding down the button (which makes it that much harder because you have to maintain force on the button while changing force on other parts of the mouse to move it, all with one hand).
If they weren't stuck in the paradigm of traditional drag and drop, they might've suggested a much easier to learn and much simpler to execute solution: create a place on the screen that can temporarily hold whatever you're trying to drag and drop, something that will hold onto whatever you're dragging while you're doing whatever you need to (rearranging windows, or maybe something else). Basically, a second hand. In fact, I would use a hand as the icon for it.
If you have a second hand, you can locate the original location, drag the thing from there to the other hand, then locate the desired destination, and then drag it from the other hand to where you ultimately want it to go. Just like you do in real life when you want to move something but you have to manipulate other things between picking it up and putting it down. Nobody in the real world flips through stuff with the same hand they're holding things with if they have the other hand free.
Being one of the resident curmudgeonly grammar freaks, I feel inclined to point out that "layoff" is a noun, and "lay off" is a verb (phrase). Therefore, it isn't possible to "layoff" someone; instead, you must lay them off.