heh actually i work at and attend URI. im the systems manager for the computer science department. i learned about Oshean, NoX and Abilene over the course of a few days after one of those campus-wide outages during the summer...pretty interesting stuff.
my pondering at the end of my comment mainly had to do with whether I2 itself had any filtering. if i traceroute from my office to say www.gatech.edu, i can see that it goes over I2. but i have no way of knowing if the packets that make up my CounterStrike (Quake3/Tribes2/etc) game take that same route. although, now that i think about it, UNIX traceroute does use high port UDP connections just like most modern network-enabled games. hmm...
as for dialup, well, im stuck with it at home. good thing too otherwise id still be spending 4 hours a day on IRC...:)
Johnson and Wales actually does have access to I2. JWU is hooked into Oshean, which links Rhode Island's higher education and related institutions. you can find traffic statistics for various schools' gateways at their NOC Page. I believe Oshean partners with Qwest to provide Internet1 access to all of us.
Oshean, along with educational institutions from several other New England states, links to Northern Crossroads. NoX colocates at Qwest's PoP in Boston. It is there that they link to Abilene and I2 (an OC-12).
im not exactly sure what URI's link to Oshean runs at...the traffic graphs seem to suggest that we have quite a fat pipe:) i dont live in a dorm so i dont know how the net access is there, but id imagine it gets pretty saturated at night. i can sorta tell because at about 4pm on weekdays, whatever mp3 stream im listening to starts to lag. im in an office in Tyler Hall which is where most of the networking stuff is (i think)...
anyway, ive been wondering...if i play counter strike from my office on a server at Georgia Tech (another I2 school), does it go over Abilene or does certain traffic get blocked because its unlikely to be "research-related"?:)
By this logic, the mast should be more erect in heavy winds (pulling the sail flatter) and should be pulled downwards in lighter winds (allowing the sail to fill more).
well, i dont think pulling the mast of this design downward has the effect of making a more rounded airfoil. from what i can gather from the (rather small) images, it looks like lowering the mast introduces near-horizontal folds in the sail. i would think the main effect of this would be to start spilling air off the leech of the sail. this would sort of be like slackening the main halyard slightly on a normal rig without reefing it. it would also reduce the percentage of the sail area that was maintaing the airfoil shape.
first of all, the newer linux kernels support both BSD and Unix98 style pseudo-terminals. both can coexist on the same system. redhat 6.2 for example uses BSD style for text virtual consoles (/dev/tty1 -/dev/tty6 by default), and Unix98 style (/dev/pts/0 -/dev/pts/255) for X-based terminals and remote terminals.
also, native X-Windows programs (excluding things like xterm and rxvt) do not usually open pseudo-terminals for themselves (you can see this using lsof or even ps). i think they utilize some feature of the X-Windows protocol to have the X server send messages to the console on their behalf (never done any X programming so not sure how that works). so while there may be a software-imposed limit to the number of xterms i could open, there isnt one for the number of games of xbill i could open.
those of you instructing the author of this post to "read the question again" should perhaps read the title of the article again. if we assume (logically) that the title of the article is meant to be a few word summary of the theme of the article, then this poster is right on the money, IMHO.
diverting all the electricity required to run this thing to a set of alligator clamps firmly attached to your testicles, you lame-ass troll wannabe? Oh manhood...
i found this article over at the register rather interesting. among other things, they report that the OpenOffice 6 code Sun released "...seems to be an alpha version." i couldnt find anything that suggests this is true on openoffice.org...nor in the oo_605_src tarball. so what is this claim based on?
i checked this out a while back when i first saw it mentioned. someone said in a previous comment it was a pain to compile, but i dont remember it being so (though maybe i had a different version). sometimes i kinda like having that cde feel, and it is pretty relatively light weight. it seems like development is progressing steadily, too.
His problem is with the fact that the default format is TNEF and that it's entirely possible that if you don't send your mail via Exchange server, other won't be able to read it.
youre right...if someone decides to send me a message with a voting box, i wont be able to vote. we should all go to redmond with torches and sticks and demand that they use only plain text for sending voting boxes....sigh...
Chris' article seems to suggest that a TNEF-encoded message, when viewed with a non-TNEF-aware mail program, will be unreadable. well, that may be true for a heavily formatted message that includes things like tables, different font styles, and, god forbid, voting boxes, but those are things that are rather difficult to produce using plain text anyway. personally i never send anything other than plain text, so if i wanted to do something complicated, i would put it format that could handle complicated things (word, ps, html, etc) and just attach it to a plaintext message or stick it on the web and mail a url. for simple "non-formatted" messages, however, this line from the KB article clarifies that there wont be any problems:
A TNEF-encoded message contains a plain text version of the message, and a binary attachment that "packages" various other parts of the original message.
what exactly is the problem with that? i find it irritating when people send me simple messages formatted with html since i then have to mentally read around it. so it would seem to me that its actually more convenient for someone like me to read a simple message because all the formatting info is stored in an attachment rather than scattered among the text itself.
so id like to know, exactly how does this mean that MS is trying to "supplant ASCII"? i think this article is just going to appeal to people looking for yet another way to bash MS. i mean, MS has done lots of things that warrent bashing, but this is not one of them IMO.
hmm...surely you dont mean to imply that the minute fraction of the linux community participating in this discussion represent 90% of their peers? surely you dont believe that the only people that use linux are those who have "switched" over from a Microsoft OS? surely you dont think that in order to benefit from the linux port of the AOL client, one would have to delve deep into the bowels of linux and subject themselves to the nitty-gritty details of compiling kernels and configuring pppd from scratch using only the CLI?
er, yes but...they were talking specifically about ip_masq (which ive yet to see out of the context of linux)...and i always thought ip_masq was somehow functionally different from NAT. though now it seems as i start looking for an explanation of the differences, that there are implimentations of NAT that function similarly (so called "1:many NAT").
eh, not completely true. almost all outgoing IP connections on any platform will use a local port between 1024 and 65535. an ip_masq box typically will use local ports in the 60000 range for masq'd outgoing connections. i have two machines behind a masq box and there are currently 40 entries in/proc/net/ip_masquerade, all with ports above 60000 (of course most are just waiting to be expired, but anyway). it wouldnt be 100% proof, but i can see @home watching customer traffic to see if anyone has mostly local ports >60k, and then bitching at them in the same way they bitch when they find an open service port (21, 25, 80, etc). i wonder how long itll be until they start monitoring the actual content of traffic to figure out who is running irc bots and whatnot...
yes, i think this is quite important to the question and it seems a lot of people are missing it. when you consider the freedoms that companies have to monitor and police their internal networks, i dont see where the problem lies with respect to the whole 'censorware vs freedom of speech' issue.
i recently setup a client's mailserver to silently copy every email that passed through it to a specific account so that the bosses could monitor people's usage. and iirc, they had inquired about software to block emails that were deemed non work-related.
when you think about it, its sort of similar to owning a house and being able to choose who to allow over as dinner guests...is it not?
But the pesky 13-year-old who's seen his 31337 friends "hack" their way into porn sites... they'll be able to get around it. But if you make
him sign an agreement that he won't, then you can kick him off when he does.
ok, IANAL, however, i do know for certain that in the US, no one under 18 can enter into a binding contract (ok, its more like no one can attempt to enforce a binding contract on someone under 18). now that doesnt say that you cant kick a kid off cause he's violating your usage policy, but what if the parents of some of his underage friends who are also looking at the site over his shoulder find out and go looking for someone to blame? i can certainly imagine them going after the company for 'allowing' it to happen...negligence is one of those words that gets assigned easily when dealing with minors in the legal system it seems...
i just wanted to point out another reason why keyword filtering is utterly pointless. this example comes from yesterday's quickies section. as long as the net is around and people have an interest in programming, there will be sites like l33t surfer. this site uses a CGI to reformat all the printed text on a page into "l33t speak," thus nullifying any attempts to block certain words. here is a live example to prove my point:
WARNING: the following are actual links to a real live porn site. if you are offended by graphic nudity or are underage, dont click on them. i hereby disclaim any responsibility for the consequences of following the links.
before and after filtering through the l33t surfer.
as you can see, all of the naughty words (and all the non-naughty ones as well) are pretty unreadable, but the pictures are still there. any links you subsequently follow from that page are also filtered by the CGI. observant folks will notice that the META tags used by search engines arent touched by the l33t surfer's CGI program, but im sure it would be a trivial modification to make them unreadable as well.
my point is, as long as someone can sit down and search out a filtering site like this one (babelfish anyone? try filtering porn in bad translations of several languages), keyword filtering will be futile.
ive seen and read about dogs being used to sniff out skin cancer. theyre also being used by people with epilepsy. they are trained to alert their owners of an oncoming attack so that they can get to a safe place in time. they are also trained to look after them during and after the attack. i believe they can predict usually around 15 minutes beforehand, and possibly up to 30.
granted this kind of training isnt cheap (over $10,000 for a 'seizure-dog' i think), but im sure its well worth it to people who suffer from the disease.
i (and probably many of the debian maintainers) would much rather see a kernel that has had almost a year and a half to mature (ie 2.2.x) used in potato than one which is just entering the.0 testing phase. i think trying to cram 2.4.x into potato even if the freeze hadnt happened yet would sacrifice some of debian's well known stability.
this site was posted on slashdot like about a year ago. it has pictures and basic specs on many old-school machines. they also have separate sections devoted to video game hardware and scans of old computer ads.
Slashdot Mirror
heh actually i work at and attend URI. im the systems manager for the computer science department. i learned about Oshean, NoX and Abilene over the course of a few days after one of those campus-wide outages during the summer...pretty interesting stuff.
:)
my pondering at the end of my comment mainly had to do with whether I2 itself had any filtering. if i traceroute from my office to say www.gatech.edu, i can see that it goes over I2. but i have no way of knowing if the packets that make up my CounterStrike (Quake3/Tribes2/etc) game take that same route. although, now that i think about it, UNIX traceroute does use high port UDP connections just like most modern network-enabled games. hmm...
as for dialup, well, im stuck with it at home. good thing too otherwise id still be spending 4 hours a day on IRC...
--Siva
Johnson and Wales actually does have access to I2. JWU is hooked into Oshean, which links Rhode Island's higher education and related institutions. you can find traffic statistics for various schools' gateways at their NOC Page. I believe Oshean partners with Qwest to provide Internet1 access to all of us.
:) i dont live in a dorm so i dont know how the net access is there, but id imagine it gets pretty saturated at night. i can sorta tell because at about 4pm on weekdays, whatever mp3 stream im listening to starts to lag. im in an office in Tyler Hall which is where most of the networking stuff is (i think)...
:)
Oshean, along with educational institutions from several other New England states, links to Northern Crossroads. NoX colocates at Qwest's PoP in Boston. It is there that they link to Abilene and I2 (an OC-12).
im not exactly sure what URI's link to Oshean runs at...the traffic graphs seem to suggest that we have quite a fat pipe
anyway, ive been wondering...if i play counter strike from my office on a server at Georgia Tech (another I2 school), does it go over Abilene or does certain traffic get blocked because its unlikely to be "research-related"?
--Siva
By this logic, the mast should be more erect in heavy winds (pulling the sail flatter) and should be pulled downwards in lighter winds (allowing the sail to fill more).
well, i dont think pulling the mast of this design downward has the effect of making a more rounded airfoil. from what i can gather from the (rather small) images, it looks like lowering the mast introduces near-horizontal folds in the sail. i would think the main effect of this would be to start spilling air off the leech of the sail. this would sort of be like slackening the main halyard slightly on a normal rig without reefing it. it would also reduce the percentage of the sail area that was maintaing the airfoil shape.
--Siva
BTW what the hell do you need to open that many windows for?
:P
for porn, silly
--Siva
first of all, the newer linux kernels support both BSD and Unix98 style pseudo-terminals. both can coexist on the same system. redhat 6.2 for example uses BSD style for text virtual consoles (/dev/tty1 - /dev/tty6 by default), and Unix98 style (/dev/pts/0 - /dev/pts/255) for X-based terminals and remote terminals.
/devices.txt and console(4) for more info...
also, native X-Windows programs (excluding things like xterm and rxvt) do not usually open pseudo-terminals for themselves (you can see this using lsof or even ps). i think they utilize some feature of the X-Windows protocol to have the X server send messages to the console on their behalf (never done any X programming so not sure how that works). so while there may be a software-imposed limit to the number of xterms i could open, there isnt one for the number of games of xbill i could open.
mmm....hundreds of millions of dead Bill's...
see
--Siva
those of you instructing the author of this post to "read the question again" should perhaps read the title of the article again. if we assume (logically) that the title of the article is meant to be a few word summary of the theme of the article, then this poster is right on the money, IMHO.
--Siva
diverting all the electricity required to run this thing to a set of alligator clamps firmly attached to your testicles, you lame-ass troll wannabe? Oh manhood...
--Siva
CTRL-V 248 in insert mode seemed to work just fine for me (using vim) :P granted, i had to look it up, but thats what help is for :)
yes im a vi whore yadda yadda
--Siva
i found this article over at the register rather interesting. among other things, they report that the OpenOffice 6 code Sun released "...seems to be an alpha version." i couldnt find anything that suggests this is true on openoffice.org...nor in the oo_605_src tarball. so what is this claim based on?
--Siva
Keyboard not found.
im sure people are getting tired of hearing comments like this, but i just have to point out that XFce has been mentioned on /. before. see:
/99 /11/18/0957218.shtml
http://slashdot.org/asksl ash dot/00/06/24/2253237.shtml
http://slashdot.org/articles
i checked this out a while back when i first saw it mentioned. someone said in a previous comment it was a pain to compile, but i dont remember it being so (though maybe i had a different version). sometimes i kinda like having that cde feel, and it is pretty relatively light weight. it seems like development is progressing steadily, too.
ok...commence flaming
--Siva
Keyboard not found.
yep...plus, the tnef program that has been referred to several times in other comments has been on freshmeat for almost a year...
--Siva
Keyboard not found.
actually, "can you say 'foo' boys and girls? i knew you could" was used by Fred Rogers of Mr. Rogers' Neighborhood.
wont you be my neighbor?
--Siva
Keyboard not found.
His problem is with the fact that the default format is TNEF and that it's entirely possible that if you don't send your mail via Exchange server, other won't be able to read it.
youre right...if someone decides to send me a message with a voting box, i wont be able to vote. we should all go to redmond with torches and sticks and demand that they use only plain text for sending voting boxes....sigh...
--Siva
Keyboard not found.
Chris' article seems to suggest that a TNEF-encoded message, when viewed with a non-TNEF-aware mail program, will be unreadable. well, that may be true for a heavily formatted message that includes things like tables, different font styles, and, god forbid, voting boxes, but those are things that are rather difficult to produce using plain text anyway. personally i never send anything other than plain text, so if i wanted to do something complicated, i would put it format that could handle complicated things (word, ps, html, etc) and just attach it to a plaintext message or stick it on the web and mail a url. for simple "non-formatted" messages, however, this line from the KB article clarifies that there wont be any problems:
A TNEF-encoded message contains a plain text version of the message, and a binary attachment that "packages" various other parts of the original message.
what exactly is the problem with that? i find it irritating when people send me simple messages formatted with html since i then have to mentally read around it. so it would seem to me that its actually more convenient for someone like me to read a simple message because all the formatting info is stored in an attachment rather than scattered among the text itself.
so id like to know, exactly how does this mean that MS is trying to "supplant ASCII"? i think this article is just going to appeal to people looking for yet another way to bash MS. i mean, MS has done lots of things that warrent bashing, but this is not one of them IMO.
--Siva
Keyboard not found.
hmm...surely you dont mean to imply that the minute fraction of the linux community participating in this discussion represent 90% of their peers? surely you dont believe that the only people that use linux are those who have "switched" over from a Microsoft OS? surely you dont think that in order to benefit from the linux port of the AOL client, one would have to delve deep into the bowels of linux and subject themselves to the nitty-gritty details of compiling kernels and configuring pppd from scratch using only the CLI?
--Siva
Keyboard not found.
er, yes but...they were talking specifically about ip_masq (which ive yet to see out of the context of linux)...and i always thought ip_masq was somehow functionally different from NAT. though now it seems as i start looking for an explanation of the differences, that there are implimentations of NAT that function similarly (so called "1:many NAT").
so...yeah...or something
--Siva
Keyboard not found.
They can't possibly detect ip-masq.
/proc/net/ip_masquerade, all with ports above 60000 (of course most are just waiting to be expired, but anyway). it wouldnt be 100% proof, but i can see @home watching customer traffic to see if anyone has mostly local ports >60k, and then bitching at them in the same way they bitch when they find an open service port (21, 25, 80, etc). i wonder how long itll be until they start monitoring the actual content of traffic to figure out who is running irc bots and whatnot...
eh, not completely true. almost all outgoing IP connections on any platform will use a local port between 1024 and 65535. an ip_masq box typically will use local ports in the 60000 range for masq'd outgoing connections. i have two machines behind a masq box and there are currently 40 entries in
--Siva
Keyboard not found.
Your tragic misspelling of the word "Gödel" has.....has.....*sob* i can't do it...please forgive me :~(
--Siva the lost
Keyboard not found.
yes, i think this is quite important to the question and it seems a lot of people are missing it. when you consider the freedoms that companies have to monitor and police their internal networks, i dont see where the problem lies with respect to the whole 'censorware vs freedom of speech' issue.
i recently setup a client's mailserver to silently copy every email that passed through it to a specific account so that the bosses could monitor people's usage. and iirc, they had inquired about software to block emails that were deemed non work-related.
when you think about it, its sort of similar to owning a house and being able to choose who to allow over as dinner guests...is it not?
--Siva
Keyboard not found.
But the pesky 13-year-old who's seen his 31337 friends "hack" their way into porn sites... they'll be able to get around it. But if you make
him sign an agreement that he won't, then you can kick him off when he does.
ok, IANAL, however, i do know for certain that in the US, no one under 18 can enter into a binding contract (ok, its more like no one can attempt to enforce a binding contract on someone under 18). now that doesnt say that you cant kick a kid off cause he's violating your usage policy, but what if the parents of some of his underage friends who are also looking at the site over his shoulder find out and go looking for someone to blame? i can certainly imagine them going after the company for 'allowing' it to happen...negligence is one of those words that gets assigned easily when dealing with minors in the legal system it seems...
--Siva
Keyboard not found.
i just wanted to point out another reason why keyword filtering is utterly pointless. this example comes from yesterday's quickies section. as long as the net is around and people have an interest in programming, there will be sites like l33t surfer. this site uses a CGI to reformat all the printed text on a page into "l33t speak," thus nullifying any attempts to block certain words. here is a live example to prove my point:
WARNING: the following are actual links to a real live porn site. if you are offended by graphic nudity or are underage, dont click on them. i hereby disclaim any responsibility for the consequences of following the links.
before and after filtering through the l33t surfer.
as you can see, all of the naughty words (and all the non-naughty ones as well) are pretty unreadable, but the pictures are still there. any links you subsequently follow from that page are also filtered by the CGI. observant folks will notice that the META tags used by search engines arent touched by the l33t surfer's CGI program, but im sure it would be a trivial modification to make them unreadable as well.
my point is, as long as someone can sit down and search out a filtering site like this one (babelfish anyone? try filtering porn in bad translations of several languages), keyword filtering will be futile.
--Siva
Keyboard not found.
ive seen and read about dogs being used to sniff out skin cancer. theyre also being used by people with epilepsy. they are trained to alert their owners of an oncoming attack so that they can get to a safe place in time. they are also trained to look after them during and after the attack. i believe they can predict usually around 15 minutes beforehand, and possibly up to 30.
granted this kind of training isnt cheap (over $10,000 for a 'seizure-dog' i think), but im sure its well worth it to people who suffer from the disease.
--Siva
Keyboard not found.
i (and probably many of the debian maintainers) would much rather see a kernel that has had almost a year and a half to mature (ie 2.2.x) used in potato than one which is just entering the .0 testing phase. i think trying to cram 2.4.x into potato even if the freeze hadnt happened yet would sacrifice some of debian's well known stability.
--Siva
Keyboard not found.
this site was posted on slashdot like about a year ago. it has pictures and basic specs on many old-school machines. they also have separate sections devoted to video game hardware and scans of old computer ads.
http://www.geocities.com/~compcloset/
--Siva
Keyboard not found.
I miss gopher. ;(
remember when schools had gopher sites, and no web sites?
remember ftp-by-mail?
Keyboard not found.