Where I buy they set the monitor up, connect it to a PC and run a diagnostics tool and explain its output to you. Takes 15 minutes or so, but hey, if you buy a big one, you leave a lot of money there. You won't get that kind of support or expertise in all shops, but I don't mind paying a little more to get that.
as an aside, sharing files using windows file sharing? blech. Better to use HTTP on a local apache server of some sort) If they are CS students, the RIAA should put an injunction against them for bad application design.
You can either have a lot of people sharing data using crappy Windows file sharing (they know how to use it, they don't have to install anything) or a few computer-savvy people sharing data using better protocols. If I have to choose, I take the system that offers a lot more data.
In many cases, traffic passes between two dorm rooms on the same switch.
In our university network, in enough cases, p2p did slow down the various research facilities. We are not a campus university, and we are large, distributed over the city. Another problem is gaming with its many small (UDP, I think) packets.
Solution was to change the network structure. We can be grateful that they chose to do that instead of introducing internal traffic limits.
I lived for a couple of years in a very beautiful farmhouse in Germany that is about 450 years old. Thick walls which made it cool in summer inside. Unfortunately, the windows were crap so the heat went out easily in winter. They were also quite small so there wasn't enough light. Due to monumental protection regulations they could not be replaced by modern ones if those weren't very expensive special ones and we were just renting the house, so heating costs were high, as were overall maintenance costs. The house didn't have a cellar. It had a wooden structure and brick walls. I have no idea whether these properties are a must-have for a long-lived house, but it seems to have worked for this one - constructed in 1637 IIRC, in the middle of the 30 years war.
Imagine the time... Connor McLeod was a measly 100 plus change years old, and a young Strom Thurmond turned to politics.
I find it really amusing that "everyone" here thinks MS is so evil, when in reality, they are one of the best companies to work for - [...]
People here dislike MS for their policies and the way they are dealing with customers and competitors. I don't think it was ever suggested that they treat their employees badly.
Maybe a login ID factor that would tell the server to only send posts at or above your current threshold -- that shouldn't require any modification to the user's NNTP client, only to what they call themselves. So I'd be "Reziac +2" and that would tell it "login name Reziac, threshold +2"
Good idea. However, people quote rarely on/., and sometimes I find it necessary to read whatever came before an insightful comment, even if those comments aren't insightful themselves. I'd probably download everything, but again, that depends on the connection one is on.
And what about when slashdot has decided to give you mod points for the day? Kinda hard to apply via NNTP.
Right, but I didn't want to replace/. web with NNTP in its entirety. Personally I gave up on moderating when perfectly valid moderations of mine were modded down in meta moderation.
How about if the NNTP header on each post includes a link to the comment on the HTTP server, so you can go directly to it without a long hunt -- that would let you do your modding, and follow sub-threshold threads as desired. It would have to take care to preserve your login, tho -- Slashdot already has a problem with losing one's login (which kicks me from low-bandwidth to regular, and I can't use the latter.)
Also a very good idea.
Anyway, I doubt that NNTP will come. It's quite a bit of work and probably few people want it.
OTOH... how does the moderation system translate into NNTP? I *don't* want to have to fish thru all the "-1, moron" comments (when I go low-score diving, it's usually within a thread whose high-score comments have already caught my eye).
Moderation integrates badly, I agree. There could be a header X-Slashdot-Moderation: -1/5 that tells about moderation status at the time the comment is sent via NNTP as an article. For more mature stories this could work very well. However, for those comments that start well because the comment looks good but is shown to be a fake / false later, you're stuck with wrong moderation.
An extension to NNTP could allow for quick updates of comment moderation values. Newsreaders could be adjusted to mark articles (=comments) under a certain threshold read.
Non-mainstream and dupe items tend to collect nothing but smartassed posts. This is relatively new as a mass phenom here. Used to be dupes got good discussions too.
I don't notice that many dupes, but I can't quite agree on non-mainstream stories. Sometimes it's the specialized things that get those (few) people who care to have a nice, lively, insightful discussion. But maybe I was just lucky.
Potential project goals / mailing list?
on
An IMDb for Books
·
· Score: 1
Here a collection of qualities / ideas that would make me want to contribute. Some of them mentioned already:
- Data should be freely distributable, in.csv format so that everyone can build applications on top of the collected data. There are Open Content licenses for this. Hopefully, lots of mirror sites will copy the data.
- People should be able to contribute using a form.
- The database design should be open for debate. Somewhere in this comments section there was a (future) librarian, I'm sure there are others who want to contribute _and_ are knowledgeable.
There could be "book connections" similar to movie connections in IMDb,...
- Take into consideration different character sets and languages. People should be able to submit reviews of The Two Towers in Finnish or Dutch.
Maybe there should be a mailing list to discuss ideas related to this project.
The IMDb.list files are text, but rather hard to parse. Is there really a.csv version somewhere as the AC indicated? If so, it would be nice if someone names the exact path to it.
I'm also not new to/., but I haven't noticed decreased quality. I just read much less stuff than I used to. Usually I get some good links, the occasional really insightful post, and most of the time some good laughs.
I don't care much about diaries, friends and foes or getting to see stories early. I can handle ads, too. I'd subscribe if I had an easy way to pay (I don't have a credit card and I'm outside of the US, at least the last time I checked that ruled out any way for me). My #1 subscription reason: NNTP access.
There is a lot of crap in/.'s comments sections, but occasionally there is really good stuff. My newsreader is very suitable for going through a lot of articles. Web interfaces to discussion forums on the other hand, although I do have a very fast connection, suck.
Thus, only downloading 0day releases within a few days of their release is only when BitTorrent works well.
True (from my limited experience), but it still is p2p, even if it is specialized.
Also, you can't search on BitTorrent. You have to find websites to download from. If the website gets shut down, so does the ability to get the files.
Yes, but again, BT is a specialized p2p tool, not flawed. It cannot (and probably never wanted to) replace protocols like those used with Freenet or Gnutella.
The design goal was to make big files available to a relatively large number of people who know exactly what they want, with everyone participating in sharing the load.
But you made me think about a definition of p2p. I would probably come up with something very general.
A slide I found suggests that there is no consensus on the term. searchNetworking has a more precise definition. Hm, I'll do some reading...
eDonkey is good for large files, albeit slow. I've been using BitTorrent for a lot of my large files (the latest buffy and anime fansubs) lately, although I don't know if this counts as P2P.
Sure, you are (potentially) downloading from several people and, at the same time, uploading to several others. Relatively centralized (but that's why it works so well, IMO).
However, due to the recent number of erroneous claims by various organisations you will need to provide us with the IP address that you used to determine this illegal fileshare. As soon as we confirm from our traffic logs a transfer of illegal material from that address to your address we will deactive the account of that student.
With Gnutella and similar systems you get query results without ever retrieving meta data (let alone content) from that system. Besides, all they claim is that someone offered the data from the IP, they never said that they themselves transfered it.
Besides, you get the usual trust problem. What keeps the MPAA from feeding 'real' users' IP addresses into that system of 'industry sniffer IPs'.
My university (in Germany) recently gets more and more similarly worded mails. The people who get them here just forward them to those responsible for the IP range where the alleged perpetrator sits, and they do not reply to the MPAA as requested in the mails, but the amount of mails they get starts to annoy them. So this kind of works, unfortunately.
They have also started to do some traffic shaping and port blocking in order to reduce bandwidth. However, they are just looking into this, some US universities seem to do a lot more already.
One of the mails I could look into contained the following phrases: The distribution of unauthorized copies of copyrighted motion pictures constitutes copyright infringement under the Copyright Act, Title 17 United States Code Section 106(3). This conduct may also violate the laws of other countries, international law, and/or treaty obligations. So, because we're not part of the USA, they basically say that the file sharing may or may not be against the law.;-)
As far as I know, uploading copyrighted material is illegal here. But they try to threaten as many people as possible with as little effort as is sufficient.
We also have an internal DC hub, but although there are quite a few users, and although there is quite a bit of exchanging going on, there is no reduction of external traffic.
People just seem to use the hub as an additional source of data, not as the exclusive one.
Notice how we go for years with Sun assuring us that the language is just fine, trust us, we know what's good for you [...]
Actually I've heard quite often how Sun's people would admit that some language features were cut out from 1.0 to make the language less complex and have implementations finished earlier.
The rsync algorithm docs say that [...] the much more expensive strong checksum must be calculated for each block where the weak checksum matches.
So the strong checksum (MD5) must be cached for a lot of blocks (typical application: files are nearly identical), and block size is usually 500 - 1000 bytes. Is 1/30th the size of the original (500 bytes vs. 128 bits) acceptable for the checksum cache?
Als die Nazis die Kommunisten holten, habe ich geschwiegen; Ich war ja kein Kommunist. Als sie die Sozialdemokraten einsperrten, habe ich geschwiegen; ich war ja kein Sozialdemokrat. Als sie die Gewerkschafter holten, habe ich geschwiegen; ich war ja kein Gewerkschafter. Als sie die Juden holten, habe ich geschwiegen, ich war ja kein Jude. Als sie mich holten, gab es keinen mehr, der protestieren konnte.
At the bottom there also is a link to the corresponding Spiegel TV video. It's called Die Rückkehr der Kartoffelkanone (Return of the potato gun / cannon), so that indicates already that this kind of weapon isn't exactly new. But that shall not keep everyone here from making fun of Germans!;-)
Slashdot Mirror
Where I buy they set the monitor up, connect it to a PC and run a diagnostics tool and explain its output to you. Takes 15 minutes or so, but hey, if you buy a big one, you leave a lot of money there. You won't get that kind of support or expertise in all shops, but I don't mind paying a little more to get that.
as an aside, sharing files using windows file sharing? blech. Better to use HTTP on a local apache server of some sort) If they are CS students, the RIAA should put an injunction against them for bad application design.
You can either have a lot of people sharing data using crappy Windows file sharing (they know how to use it, they don't have to install anything) or a few computer-savvy people sharing data using better protocols. If I have to choose, I take the system that offers a lot more data.
In many cases, traffic passes between two dorm rooms on the same switch.
In our university network, in enough cases, p2p did slow down the various research facilities. We are not a campus university, and we are large, distributed over the city. Another problem is gaming with its many small (UDP, I think) packets.
Solution was to change the network structure. We can be grateful that they chose to do that instead of introducing internal traffic limits.
At least in his last presidential election he had 100 percent of the vote. That's more than twice of what George W. got!
I lived for a couple of years in a very beautiful farmhouse in Germany that is about 450 years old. Thick walls which made it cool in summer inside. Unfortunately, the windows were crap so the heat went out easily in winter. They were also quite small so there wasn't enough light. Due to monumental protection regulations they could not be replaced by modern ones if those weren't very expensive special ones and we were just renting the house, so heating costs were high, as were overall maintenance costs. The house didn't have a cellar. It had a wooden structure and brick walls. I have no idea whether these properties are a must-have for a long-lived house, but it seems to have worked for this one - constructed in 1637 IIRC, in the middle of the 30 years war.
Imagine the time... Connor McLeod was a measly 100 plus change years old, and a young Strom Thurmond turned to politics.
...to terrorism over a couple of edges. Watch out, Kevin Bacon!
I find it really amusing that "everyone" here thinks MS is so evil, when in reality, they are one of the best companies to work for - [...]
People here dislike MS for their policies and the way they are dealing with customers and competitors. I don't think it was ever suggested that they treat their employees badly.
Maybe a login ID factor that would tell the server to only send posts at or above your current threshold -- that shouldn't require any modification to the user's NNTP client, only to what they call themselves. So I'd be "Reziac +2" and that would tell it "login name Reziac, threshold +2"
/., and sometimes I find it necessary to read whatever came before an insightful comment, even if those comments aren't insightful themselves. I'd probably download everything, but again, that depends on the connection one is on.
/. web with NNTP in its entirety. Personally I gave up on moderating when perfectly valid moderations of mine were modded down in meta moderation.
Good idea. However, people quote rarely on
And what about when slashdot has decided to give you mod points for the day? Kinda hard to apply via NNTP.
Right, but I didn't want to replace
How about if the NNTP header on each post includes a link to the comment on the HTTP server, so you can go directly to it without a long hunt -- that would let you do your modding, and follow sub-threshold threads as desired. It would have to take care to preserve your login, tho -- Slashdot already has a problem with losing one's login (which kicks me from low-bandwidth to regular, and I can't use the latter.)
Also a very good idea.
Anyway, I doubt that NNTP will come. It's quite a bit of work and probably few people want it.
JMDB: I recently searched for "imdb data" and related tools, but I didn't find that one. Seems to be exactly what I was looking for, thanks.
OTOH... how does the moderation system translate into NNTP? I *don't* want to have to fish thru all the "-1, moron" comments (when I go low-score diving, it's usually within a thread whose high-score comments have already caught my eye).
Moderation integrates badly, I agree. There could be a header X-Slashdot-Moderation: -1/5 that tells about moderation status at the time the comment is sent via NNTP as an article. For more mature stories this could work very well. However, for those comments that start well because the comment looks good but is shown to be a fake / false later, you're stuck with wrong moderation.
An extension to NNTP could allow for quick updates of comment moderation values. Newsreaders could be adjusted to mark articles (=comments) under a certain threshold read.
Non-mainstream and dupe items tend to collect nothing but smartassed posts. This is relatively new as a mass phenom here. Used to be dupes got good discussions too.
I don't notice that many dupes, but I can't quite agree on non-mainstream stories. Sometimes it's the specialized things that get those (few) people who care to have a nice, lively, insightful discussion. But maybe I was just lucky.
Here a collection of qualities / ideas that would make me want to contribute. Some of them mentioned already:
.csv format so that everyone can build applications on top of the collected data. There are Open Content licenses for this. Hopefully, lots of mirror sites will copy the data.
...
- Data should be freely distributable, in
- People should be able to contribute using a form.
- The database design should be open for debate. Somewhere in this comments section there was a (future) librarian, I'm sure there are others who want to contribute _and_ are knowledgeable.
There could be "book connections" similar to movie connections in IMDb,
- Take into consideration different character sets and languages. People should be able to submit reviews of The Two Towers in Finnish or Dutch.
Maybe there should be a mailing list to discuss ideas related to this project.
The IMDb .list files are text, but rather hard to parse. Is there really a .csv version somewhere as the AC indicated? If so, it would be nice if someone names the exact path to it.
I'm also not new to /., but I haven't noticed decreased quality. I just read much less stuff than I used to. Usually I get some good links, the occasional really insightful post, and most of the time some good laughs.
/.'s comments sections, but occasionally there is really good stuff. My newsreader is very suitable for going through a lot of articles. Web interfaces to discussion forums on the other hand, although I do have a very fast connection, suck.
I don't care much about diaries, friends and foes or getting to see stories early. I can handle ads, too. I'd subscribe if I had an easy way to pay (I don't have a credit card and I'm outside of the US, at least the last time I checked that ruled out any way for me). My #1 subscription reason: NNTP access.
There is a lot of crap in
Thus, only downloading 0day releases within a few days of their release is only when BitTorrent works well.
True (from my limited experience), but it still is p2p, even if it is specialized.
Also, you can't search on BitTorrent. You have to find websites to download from. If the website gets shut down, so does the ability to get the files.
Yes, but again, BT is a specialized p2p tool, not flawed. It cannot (and probably never wanted to) replace protocols like those used with Freenet or Gnutella.
The design goal was to make big files available to a relatively large number of people who know exactly what they want, with everyone participating in sharing the load.
But you made me think about a definition of p2p. I would probably come up with something very general.
A slide I found suggests that there is no consensus on the term. searchNetworking has a more precise definition. Hm, I'll do some reading...
eDonkey is good for large files, albeit slow. I've been using BitTorrent for a lot of my large files (the latest buffy and anime fansubs) lately, although I don't know if this counts as P2P.
Sure, you are (potentially) downloading from several people and, at the same time, uploading to several others. Relatively centralized (but that's why it works so well, IMO).
Damn those computer-confiscating Nazis!
However, due to the recent number of erroneous claims by various organisations you will need to provide us with the IP address that you used to determine this illegal fileshare. As soon as we confirm from our traffic logs a transfer of illegal material from that address to your address we will deactive the account of that student.
With Gnutella and similar systems you get query results without ever retrieving meta data (let alone content) from that system. Besides, all they claim is that someone offered the data from the IP, they never said that they themselves transfered it.
Besides, you get the usual trust problem. What keeps the MPAA from feeding 'real' users' IP addresses into that system of 'industry sniffer IPs'.
My university (in Germany) recently gets more and more similarly worded mails. The people who get them here just forward them to those responsible for the IP range where the alleged perpetrator sits, and they do not reply to the MPAA as requested in the mails, but the amount of mails they get starts to annoy them. So this kind of works, unfortunately.
;-)
They have also started to do some traffic shaping and port blocking in order to reduce bandwidth. However, they are just looking into this, some US universities seem to do a lot more already.
One of the mails I could look into contained the following phrases: The distribution of unauthorized copies of copyrighted motion pictures constitutes copyright infringement under the Copyright Act, Title 17 United States Code Section 106(3). This conduct may also violate the laws of other countries, international law, and/or treaty obligations. So, because we're not part of the USA, they basically say that the file sharing may or may not be against the law.
As far as I know, uploading copyrighted material is illegal here. But they try to threaten as many people as possible with as little effort as is sufficient.
Reminds me of that free FTP server Warftp. Obviously the author dislikes government agencies so much he released it under a license that allows everybody except governmental institutions to use his software.
So if someone doesn't like the US military, they can restrict the license of their software.
We also have an internal DC hub, but although there are quite a few users, and although there is quite a bit of exchanging going on, there is no reduction of external traffic.
People just seem to use the hub as an additional source of data, not as the exclusive one.
Notice how we go for years with Sun assuring us that the language is just fine, trust us, we know what's good for you [...]
Actually I've heard quite often how Sun's people would admit that some language features were cut out from 1.0 to make the language less complex and have implementations finished earlier.
The rsync algorithm docs say that [...] the much more expensive strong checksum must be calculated for each block where the weak checksum matches.
So the strong checksum (MD5) must be cached for a lot of blocks (typical application: files are nearly identical), and block size is usually 500 - 1000 bytes. Is 1/30th the size of the original (500 bytes vs. 128 bits) acceptable for the checksum cache?
Available at gnuwin32.sf.net, unxutils.sf.net and other places.
The quote is by Martin Niemöller:
Als die Nazis die Kommunisten holten,
habe ich geschwiegen;
Ich war ja kein Kommunist.
Als sie die Sozialdemokraten einsperrten,
habe ich geschwiegen;
ich war ja kein Sozialdemokrat.
Als sie die Gewerkschafter holten,
habe ich geschwiegen;
ich war ja kein Gewerkschafter.
Als sie die Juden holten, habe ich geschwiegen,
ich war ja kein Jude.
Als sie mich holten, gab es keinen mehr,
der protestieren konnte.
Here's a page with an English translation.
More precisely, this article: Durchschlagende Wucht.
;-)
At the bottom there also is a link to the corresponding Spiegel TV video. It's called Die Rückkehr der Kartoffelkanone (Return of the potato gun / cannon), so that indicates already that this kind of weapon isn't exactly new. But that shall not keep everyone here from making fun of Germans!