| WIKI | FAQ | Tech FAQ | |
 http://csua.com/feed/ | |
| ||||||
| 2005/5/6 [Uncategorized] UID:37550 Activity:nil |
5/6 (Not a homework question) You have a horizontal gap of length L and
a rope/chain/cable with mass per unit length w, and there is vertical
gravity of acceleration g. What length of cord should you use to
bridge the gap to as to minimize the tension at the anchor points? |
| 2005/5/6-9 [Politics/Domestic/Immigration, Reference/Law/Visa] UID:37551 Activity:nil |
5/6 The last few pages of my passport say "Amendments and
Endorsements". Are these pages usable for normal
entering/exiting countries, or do I need to get extra pages
added if this is all that remains blank in my passport?
\- Well you have to consider who cares about and gives you the
visa. Typically a country will say "your passport is full"
and refuse to put the visa on the A/E page. However, when you
show up at the airport in Swaziland, I suppose you could claim
it was their embassy that put it there. I'm not saying it wont ever
work, but it is a bad idea. [what if something like a schengen
visa is used to cover up something?]. BTW, many countries have a
"jumbo" passport for drug mules er people expecting a lot of stamps.
i would be kind of curious what the us immigration would do if
some other country decided to cover up one of you A/E pages. --psb |
| 2005/5/6-9 [Computer/SW/Editors/Emacs] UID:37552 Activity:kinda low |
5/6 Is there a way to pass command-line arguments to emacs, and then detect
them in my elisp code in .emacs? I want to distinguish, in my .emacs,
whether it's started by me typing "emacs" at the command line, or by
some other program using the EDITOR environment variable. So I'm
thinking I can change EDITOR from "emacs" to "emacs -foo" and then
detect it in .emacs. Thanks.
\_ try it the other way; set up an alias or wrapper script that gets
run when you run emacs by hand, that does something like
setenv IRANEMACS yes ; emacs $@ ; unsetenv IRANEMACS.
Then you can just get it out of the environment. -tom
\- The argv is processed by startup.el and eventually will
leave a list in the variable command-line-args, e.g.
M-x lisp-interaction-mode
(eval 'command-line-args)C-j
("emacs" "/tmp/holube")
You can also look at command-switch-alist assuming unlike
holube you dont find alists scary and hard. There are probably
other ways to do what you want [e.g. with gnuclient/gnudoit]
if you want to be specific or just calling emacs -f ...
\_ I don't find them scary and hard; I find them stupid. -tom
\- you find alists "stupid" and you love perl? there you
have it folks. do you also recommend people use environment
variables "instead" of hashes?
variables "instead" of hashes? any other data structures
you find "stupid"? how about say patricia trees?
\_ I wonder why, other than emacs, no other program uses
elisp for configuration. It is not the data structure
I object to, it's the syntax and its use in an
application which should be simple. -tom
\- do you know what guile is? are you making a
serious objection to elisp as opposed to objecting
to lisps in general? lisp has much more expressive
power than say a twmrc configuration language.
"emacs people" are more interested in power than
popularity. they also dont confuse aesthetic/elegant
and simple. the rule is not "thing should be simple"
but things should be as simple as necessary" there
are certainly problems [scoping] with elisp and
limitations as well [say with continuations] but
"alists are bad" is not one anybody reasonable would
take seriously. these comments of yours are getting
increasingly retarded. keep it up. bitching about m4
however is reasonable.
\_ I am not objecting to lisp; I am objecting to
lisp used as a configuration language for a text
editor. I don't like it much as an extension
language, either, but that's more a personal
preference. The fact that you have an extension
language doesn't mean that you have to use it
for everything. It would be much more effective
to separate configuration and extension in
emacs. -tom
\_ Tom, do you know lisp, or are you objecting to
not being able to figure out lisp syntax from
randomly looking at code snippets? I don't
see any specific objections of the type
generally expected when critiquing programming
languages. If you don't know what lambas,
let*, progn etc are, you are not qualified to
let*, progn are, you are not qualified to
\- mapcar >> holube
comment on the language design. Maybe you
should just stick to comments in the form
"I do not like elisp" instead of the arrogant
"Elisp is broken". These kinds of incessant
"Elisp is broken". These of incessant
arrogant pronouncements make you look like an
asshole. But maybe you are an asshole.
\_ I'm not commenting on the language design
at all; I'm commenting on its use in
configuration of a text editor. I also
think it would be stupid to use java or
perl to configure a text editor. -tom
\- it might be more effecting at raising
emacs's mkt share but programmatically
i dont think it would be more effective at all.
it sounds really lame in fact to have an
.emacsrc where you do things like "set fgcolor=
blue" just to make people who dont understand
lisp a little happier. the way scope in handled
in web server configs or procmail configs
or mailagent configs is much lamer than how
it is handled in emacs. [like say if you want
to have one notion of what is a paragraph
when editing lisp vs english text].
\_ well for me, that's exactly why I hate lisp.
(set fgcolor blue) instead of just
fgcolor = blue, and it gets worse with the
nested levels of parentheses and quoted
identifiers and blah blah.
I would much prefer a Ruby-like syntax.
I always found lisp to be unpleasant.
\- setq is more powerful than simple
X=Y type assertions. also there is a
point to having setq, setq-default,
defvar, defmacro, defun, defadvise, let
let* etc. you can happily "configure"
much of emacs only knowing a little
bit about setq. however if the
documentation tells you something is
expected to be a list and you dont use
setq correctly, that's an E_TOOSHORT
problem as far as i am concerned and not
a problem with elisp. once again richer
semantics != gratuitious complexity.
\_ I'm inclined to think it is gratuitous
in the context of text editor config.
I don't want to learn all that stuff
just for emacs. I find it questionable
that all those setters are required;
there's no such complexity in Ruby and
I never felt constrained. Plus the oop
stuff would probably work nicely for
a text editor. But text editor
arguments are stupid so I'll stop.
I'm just saying I don't like lisp. I
don't mind that config files use a
rich language but it means I have to
look in a thick manual to find how
do do anything, since the gui tools
are worthless. |
| 2005/5/6-7 [Politics/Domestic, Politics/Domestic/RepublicanMedia] UID:37553 Activity:high |
5/6 Political bias at google continues. Gave $463,500 to Dems, $5,000 to
Pubs. Not good now that we know they're going to act as a news filter.
\_ short GOOG
\_ Fuck you and die.
\_ I am confused by this grammar. Did you intend to prepend
this with "I will?"
\_ "Fuck you" is a sentence by itself, and "die" is a request;
I'm just joining them together. Maybe "Fuck you; die" would
work better.
\_ May you be fucked and die.
\_ As opposed to Fox. Given that you have such a choice of news
sources out there, any number of which probably share your
political views, where on Mars do you live that you think anyone
somehow owes you news in any format/with any slant whatsoever?
You don't _have_ to read what you don't like to hear. -John
\_ Fox News is a news/opinion station. I expect bias there as well
as CNN, ABC, NBC, CBS, etc. Google is supposed to be a bunch of
algorithms. Now they're working on grading news stories (which I
hope would point out errors so I can get facts and make my own
opinions), and I'm concerned their metafilter algorithms will
have a political bias.
\_ Google is positioning itself to be an internet media company,
much like Yahoo. Very very few companies will ever exist
as 'a bunch of algorithms' and continue to be not only
profitable but capable of real growth. While I appreciate
profitable but capable of real growth. While I appeciate
your worldview, it strikes me as more than a bit naive and
ingenuous. Google isn't supposed to be anything but a
profit oriented venture targeted toward an audience that will
will bring in profits. Welcome to the new reality where the
line separating news source and news provider is increasingly
blurry (not to mention the distinctions between internet,
media, and news).
profit oriented venture targetted toward an audience that will
will bring in profits. Welcome to the new reality where
news source and news provider are an increasingly blurry line.
\_ I doubt if "just as biased as Fox" would be considered a
worthy goal to strive for.
\_ Well done, A+. Now find me an unbiased news source. -John
\_ Show me one person without sin. Does that mean one should
not strive to lead a virtuous life?
not strive to be good?
\_ Actually, I think that the Newshour is not bad since
they mostly let the people talk about the issues and
let you draw your own conclusions.
\_ Stop using it then. Stop using the Internet, too, since it
was developed at Berkeley and Stanford, both well known liberal
universities.
\_ Who cares? No one is forcing you to get your news from google.
You can still go and get your news from some other site. If
this bothers you so much why don't you set up your own news
agregation site that is "fair and balanced"? |
| 2005/5/6-9 [Uncategorized] UID:37554 Activity:nil |
5/6 Does the latest Dave Matthew's Band Video have Berkeley in the
background?
\_ "So what I felt when I was drunk was right?"
"Yes, but the Dave Matthews Band does not rock." |
| 2005/5/6-8 [Computer/SW/Security, Computer/SW/Unix] UID:37555 Activity:nil |
5/6 A lot of web sites now have a login snippet on their main page,
which forefox does not display a SSL icon
(http://www.bankofamerica.com Are those logins safe? You can
usually find a specific login page within the website that
have the SSL icon. I assume bank sites are usually safe in
their design, but what about sites like
http://www.officedepot.com Some sites's login page
(http://www.bookpool.com/ac does not have a SSL icon, but
their login button specifically says "secure login", how does
it work? As an end user, how can one be sure the login/pw
information is encrypted while in transit?
\_ It's usually good practice to put the login page under SSL to
preempt concerns like yours. Many places don't have a login box
on their front page, and make you click through to an https link
to get a login box. Others put the login box on their front
page to save you that step, but the load of putting their front
page under SSL is prohibitive. If they say it's a secure login,
the HTTP Post that sends your information will be under ssl. If
you want to test this, put in a bogus login/password and watch it
jump to SSL when you click "login".
\_ For verification:
http://www.bankofamerica.com/signin/security_details_popup.cfm
\_ So you have to 'observe' the flashing by of the SSL icon
to distinguish these sites from sites that indeed uses
no security. I guess a better question is, how do you
tell if the HTTP post used to send your login
information is under SSL?
\_ Best course of action: don't worry about it. if someone's
really intent on stealing your info, there are easier ways
to do it. There are non-technical ways to protect yourself
better. keep an eye on your account activity. get your
annual credit check (or more frequently if you're worried).
SSL is no guarantee no matter how Verisign wants to package
it.
\_ I find security policy varies significantly
between sites. Your password can be as strong as
you like, but often times the "I lost my password"
feature is typically implemented with very little
security in mind. Better sites will allow you to
reset your password after you verified who you are
(via secret questions, etc), never revealing what
your actual password was. But some no so security
conscious sites will simply email your password in
plain text, and sometimes all you have to do is to
provide your email address. Some sites will also
reset your password with only the email address.
You can only guess how careful those sites will
treat your data (such as credit card info).. I am
trying to sort out the sites that have my login
information so that the lesser secure sites do not
share the same password as the more
secure/important sites...
\_ The guy I spoke to said it used to be configurable but was
taken out. If I turn any of my URLs into https, it stays
https, including turning all the links into ssl, but I know
of several people where it redirects to http. No clue why
it varies. -John
\_ The only way to be sure is to look at the source and see
how it's posting the login. But even then, you won't know
for sure that the authentication server is using weak
encryption.
\_ What's pretty funny is that gmail defaults back to http when you've
logged in, and they seem to have removed the setting the security
guy I mentioned which lets you set ssl for all mail access. -John
\_ My gmail still stays https and always has. I know yahoo
switches back to http after login.
\_ The guy I spoke to said it used to be configurable but was
taken out. If I turn any of my URLs into https, it stays
https, including turning all the links into ssl, but I know
of several people where it redirects to http. No clue why
it varies. -John
\_ You're right. I just never noticed it, because my
bookmark specified https. Thanks for the tip. |
| 2005/5/6-7 [Health/Disease/General] UID:37556 Activity:kinda low |
5/6 So infuriating it deserves to be posted again. US Religious groups
opposing HPV (genital warts) vaccination on moral grounds. Thousands
of womens lives will be saved from cervical cancer by these vaccines.
http://www.newscientist.com/channel/sex/mg18624954.500
More details from CDC on HPV:
http://www.cdc.gov/std/HPV/STDFact-HPV.htm#common
Over 50 percent of sexually active men and women contract it in their
lifetimes, and by the age of 50, 80 percent of women will have
contracted the virus. While many cases of HPV disappear of their own
accord, it is the main risk factor in contracting cervical cancer for
women. The vaccines nearing approval prevented infection in over
90 percent of cases...
\_ Only sentence I found in the article supporting your claim is this:
"In the US, for instance, religious groups are gearing up to oppose
vaccination ..."
I think parents, not government should decide whether or not to
I think parents should have the right to decide whether or not to
vaccinate their pre-legal age daughters.
\_ Idiotic. Would you make measles, mumps and rubella vaccination
"optional" as well?
\_ Seriously, idiotic. If there were a safe HIV vaccine which
you could administer to children, would you say the same?
Also, another sentence is "Abstinence is the best way to
prevent HPV." This is only true if abstinence moved from a
personal choice to a perfectly executed legal requirement.
When half of the population are carriers, and carriers are most
often completely asymptomatic, the only effective way to prevent
infection, even if you remain abstinent until marriage, is this
vaccine for now. |
| 2005/5/6-9 [Uncategorized] UID:37557 Activity:nil |
5/6 Does anyone know of an inexpensive place to print posters
(~3'x4')? Kinkos charges $8-10/square foot. Thanks.
\_ I've seen some really nice posters made through http://zazzle.com.
I think they're based in Menlo Park so shipping is really fast.
not sure what their prices are like, though. |
| 2005/5/6-8 [Transportation/Airplane, Reference/Military] UID:37558 Activity:low |
5/6 Does any part of a hollow point bullet liquify during its flight?
\_ No.
\_ so how does it end up looking like this?
http://www.landro.no/images/Ref_nr64.jpg
I just got this off of images.google, and there are lots like it.
That sure looks to me like it's melted. Are you saying the
solid just deforms in flight?
\_ Lead is very, very soft. The reason for using a hollow
point bullet is because it expands when it hits flesh, making
a much larger wound cavity than the initial caliber of the
bullet.
\_ it didnt' take that shape in flight, but rather upon impact. |
| 2005/5/6-8 [Computer/Companies/Google] UID:37559 Activity:nil |
5/6 craigslist housing + google maps. http://paulrademacher.com/housing \_ holy shit, that's pretty nice. \_ How long before Google copies that and claim as their own? |
| 2005/5/6-7 [Finance/Investment] UID:37560 Activity:high |
5/6 Damn this economy! Only 274K jobs added in April, and unemployment is
at 5.2%.
\_ the question is, how many of those jobs are relevant to YOU (are
they mostly blue collar jobs?) Yes, they are. The economy is good
for the Red States.
\_ http://csua.org/u/bz4
Hourly wages are up, average workweek is longer. Your knee-jerk
is showing.
\_ Hourly wages have been alternatively just keeping pace w/
inflation or falling behind, according to the report in
question.
http://bls.gov/news.release/empsit.t17.htm
-- ulysses
\_ Real wages are down, you know the inflation adjusted kind.
And down even more if you subtract for the rising cost
of health care. So this news is great if you are an
investor, pretty cruddy if you are an employee.
\_ I'm thinking you missed the sarcasm.
\_ So, you know computer nerds who are having trouble finding a
job?
\_ Yes, "nerds" being the key word. -John
\_ I haven't checked for this recent report, but the last time I saw
numbers like this, the average salary was quite good--40-50K
IIRC.
\_ Average hourly pay is $16/hr. That is alot of overtime to
get 40k. You recall incorrectly, I think.
\_ I meant the average of the new jobs, not all jobs.
\_ http://www.uswa.org/uswa/program/content/1663.php
New jobs pay an average of $9/hr.
\_ Yeah, it's all Bush's fault! Damn him! |
| 2005/5/6-9 [Computer/HW/Drives] UID:37561 Activity:nil |
5/6 What do you people think of sticking four 250GB IDE Hitachi drives
in a single PC and calling that the backup server for 8 people (each
person gets 125GB for 2 sets of notebook images and other assorted
files).
I do know that SCSI is built for this and IDE is a hack-job /
consumer-grade.
I know you lose throughput by sticking two drives on the same IDE cable
because of the master/slave issue, especially if they are both being
accessed at the same time.
I am assuming if an individual drive gets toasted we just replace the
drive and tell everyone to re-backup (I've heard too many stories of
both drives in a RAID getting toasted anyway). I am assuming if the
computer fries then we just get a new computer and stick in all the
drives.
My original plan was to give a 160GB external drive to every employee
with a notebook, but the boss is inquiring about the server option
because he wants anyone who steals information from the company to
commit more obviously illegal acts, and to avoid the "My external
drive was stolen! What's an encryption option?" problems.
Yes, I know: YOU GET WHAT YOU PAY FOR. The question is:
Is what I get "enough" / workable? Thanks.
\_ Get SATA with a decent raid controller. If it's not relevant for
real time data access your main priority is being able to lose
one drive and not lose data. Do hot standby if you need to be
able to continue without rebooting, although $+ -John
\_ just make sure you are not running Solaris !!!
\_ Fair enough--it sounded like he wanted something reasonably
elegantly improvised, hence my assumption of Linux/*BSD.
\- what i have done because i dont trust the institutional backup
people is to write a little program wich basically reads from
an ssh inbound and dd's into a file. access is controled via
ssk keys. so you can do something like
"tar,ufsdump,cat importantfile | ssh locker@foo dump <label>"
which then creates a file at
server:/locker/hosts/client/<datestamp>label" ... so basically you
are providing a bitstore which is agnostic about the dump format
so people can write their own "backup clients" and have their
own backup policies. it has worked pretty well for say 50-60
machines and netapps. quite a cheep solution with a 3ware raid.--psb
\_ I agree with John, go with SATA, significantly more bang for the
buck performance-wise vs. IDE, easier to expand if you ever want
more than eight drives, and still way cheaper than SCSI. I
actually set up something similar for a company I used to work for,
and its still humming along happily several years later. For this
sort of thing, I actually prefer to use Linux or BSD software raid,
since it's one less piece of hardware in the mix that can fail, and
cpu cycles are cheap. -dans |
| 2005/5/6-11 [Computer/SW/Languages/C_Cplusplus] UID:37562 Activity:moderate |
5/6 (Not a homework question) You have a horizontal gap of length L and
\_ bullshit.
\_ double-bullshit. I made a paperclip chain and am curious how
to most easily drape it (and I'm a big nerd).
a rope/chain/cable with mass per unit length w, and there is vertical
gravity of acceleration g. What length of cord should you use to
bridge the gap to as to minimize the tension at the anchor points?
\_ A suspension like this is described by a hyperbolic cosine, the
least amount of tension is produced when the rope or chain is exactly
the length of L. Let me know if you need more proof. -scottyg
\- re: hyperbolic cos: also and probably more commonly known as
Cantenary arc/arch [like the st. louis object]. you can
google from "cantenary" ... somebody probably has the
solved integral that gives you a closed form for arc length.
this name actually comes from the idea of "hanging chain",
so appropos to you underlying physical model. and now lets
talk about the feynman sprinkler.
\_ And when you can't find the answer, search for
"catenary."
\_ Perhaps you misunderstood. The gap width is L. Trying to bridge
that with a cord onle L long would require infinite tension when
under gravity.
\_ it is a simple optimization problem that reduces down to the
more string you have, the greater the weight, and therefore the
greater the tension. L is the minimum amount of string one
would need to span the gap, therefore it is also the optimized
length for minimum tension. See the equations below for tens-
ion, T, as they are correct. BTW, I have a degree in Physics,
and not CS, if that gives me any more credit. -scottyg
\_ "assume the horse is a sphere"... sorry, nope. it doesn't.
\_ no it wouldn't.
\_ assuming an infinite tensile-strength cord, yes it would.
Take a reasonably heavy rope or chain and try to pull it
to be absolutely straight while not supported in the
middle. You can't do it.
\_ but that's not what we're trying to do. you can
straighten it by pushing in the middle. in any case
the answer is "as straight as you can".
\_ What I'm trying to do is find what length of paper
clip chain will bridge a gap L with minimum tension.
The answer is NOT L, because that has extremely
high tension which bends the paper clips all out of
shape. But the answer is not 100L either, because
that is way too much extra weight.
\_ you don't have to pull the ends to straighten
the chain. you can straighten it theoretically
without bending.
\_ Based on what little I remember about CE, scottyg is right.
A cable w/ uniform weight per unit length is described as
a cantenary. You need two equations to determine the length,
L, of the cable. Assuming that you know the separation between
the endpoints (X) and the "sag" (S) (the distance from the
horizontal passing through the end points and the lowest point
of the cable), you can use the following equations:
EQ1.1 The half length, L/2, equation:
EQ 1.1 The half length, L/2, equation:
y^2 - (L/2)^2 = c^2
EQ2.1 The parameter, c, equation:
EQ 2.1 The parameter, c, equation:
y = c cosh (X/c)
Since y = S + c you can simplify to:
EQ 1.2: L = 2(2c^2 - 2Sc - S^2)^(1/2)
EQ 2.2: S/c + 1 = cosh (X/c)
You can use excel or some such to solve for c in EQ 2.2 and
then it is simple to substitute into EQ 1.2 to get L. iirc,
in most cases it is simpler to just think of the thing as a
parabola.
then it is simple to substitute into EQ 1.2 to get L.
Now that I re-read your question, it seems like you want to
minimze S as well as L right? I don't remember enough calc
to help you out there, but I think that you probably need
to do this by minimizing the tension T at the end points.
The equation for T at the end points is:
EQ 3.1: T = w(c^2 + (L^2)/4)^(1/2)
Good luck!
\_ Hmm, what scottyg claims above is that tension T is minimum
when the string length L equals to the gap width X. Now,
as L approaches X, S approaches 0. Using your equations,
as X approaches L, S approaches 0. Using your equations,
we get this:
limit(L->X) S = 0;
limit(X->L) S = 0;
EQ 2.2: S/c + 1 = cosh (X/c)
limit(L->X) (S/c + 1) = limit(L->X) cosh (X/c)
1 = limit(L->X) cosh (X/c)
0 = limit(L->X) X/c
thus limit(L->X) c = infinity
limit(X->L) (S/c + 1) = limit(X->L) cosh (X/c)
1 = limit(X->L) cosh (X/c)
0 = limit(X->L) X/c
thus limit(X->L) c = infinity
EQ 3.1: T = w(c^2 + (L^2)/4)^(1/2)
limit(L->X) T = limit(L->X) (w(c^2 + (L^2)/4)
limit(X->L) T = limit(X->L) (w(c^2 + (L^2)/4)
^(1/2))
= infinity
I don't see how the tension is minimum in this case, using
your presumably correct equations. --- L&S CS major
your equations. --- L&S CS major
your presumably correct equations. So scottyg is wrong.
--- L&S CS major
\_ I think that what scottyg is getting at is that you
have to balance the tension at the end points against
the tensile strength of the material that makes up
the cable in order to find the min sag and length.
\_ Sorry, but this "L is the minimum amount of
string one would need to span the gap, therefore
it is also the optimized length for minimum
tension" seems to contradict that.
\_ Okay, I see your point. When I originally
posted the equations I was probably wrong
about the minimizing the tension at the
end points. Anyway, do you see anything
wrong w/ balancing the tensile strength
of the material w/ the tension at the
end points to get the optimal length
and sag?
\_ Not at all. That seems like a very reasonable
way to think about it. |
| 2005/5/6-9 [Uncategorized] UID:37563 Activity:nil |
5/6 Broadcast Flag Order Vacated:
http://pacer.cadc.uscourts.gov/docs/common/opinions/200505/04-1037b.pdf
\_ Just wanted to add that all written opinions are available for free
from now on. |
| 2005/5/6-9 [Computer/HW/Scanner, Computer/SW/Apps] UID:37564 Activity:nil |
5/6 I have a 45-page typed manuscript that needs to be turned into
a digital text document. Any recommendations for an OCR service
that would do that with high reliability and low cost?
\_ If you have a scanner, photoshop and acrobat (full version)
you can do it yourself. Just scan each page in at 300dpi or
higher and then use acrobat to OCR it into a pdf. I did this
for lots of old papers that I lost the soft copies of and
it worked great.
\_ How many errors per page did acrobat OCR make?
\_ My originals weren't particularly good, but acrobat
didn't make many mistakes on main text, maybe 2 or
3 per page not counting missing spaces. It had a
harder time w/ figures, but overall it wasn't too
bad. |
| 4/15 |