Oh, and I also worry about stuff like this being enforced in the US. (I'm a US'ian) If we are more strict with our IP laws than other countries, then our companies' ability to compete on the international market could be hampered.
I don't know much about Patent law, but I always heard that patents were easy to engineer around because they have to be so specific, and a relatively minor change can be used to engineer around the patent.
How does this extend to computer programming techniques? Why don't you have to provide a specific algorithm in the patent that can be engineered around?
...beer is supposed to be warm, and that us puny US american types are the only people who drink it icy cold.
Well, they say it's supposed to be served at room temperature. "Warm" is a relative term, and "room temperature" in Europe averages 34 degrees Fahrenheit. Those Europeans just don't realize how much warmer it is in the new world.
I quote from the last page of his site, which was not at all difficult to miss if you rtfa -
Yes, but if you'd look at the pictures on the second-to-last page it appears the thermometer's sensor is on the peltier cooler itself. Even if that's a misinterpretation on my part it's clear there is no probe in the Guiness itself but outside the glass at best.
By the way, I haven't tried the Guinnes-in-a-can yet. I'm not a total snob, but other beers aren't as good in the can, so I assumed Guiness would have the same problem. How is it? And can you pour yourself a four-leaf clover in the head with the can?:-)
Re:problem solved
on
PeltierBeer
·
· Score: 4, Funny
but what about those really hot days?
Solution: drink faster.
Have you ever tried drinking Guiness quickly?
My sister did it once: she gulped the last few ounces of her Guiness because we were ready to go. I stared at her in horror as she looked at me like it was no big deal. She was feeling sick a couple of minutes later. I wasn't insensitive enough to ask how much chest hair she'd grown because of it.
No, silly. That's passe. People will be sshnuke'ing the whole internet looking for backdoors (and wearing tight leather). That's the current hip thing.
I expect Java will continue to survive in some niches (most notably, bloated web services implementations)
I was about to install Java + JBoss + Tomcat + Apache httpd (and maybe Apache Cocoon) and see what I can do with it. I'm not a seasoned developer (just had a few classes a few years ago), so I'm trying to figure out the best way to offer a scalable web platfrom mainly for database entry and reporting without much custom programming. I'm also interested in application services, but that's more into the future.
If not Java, then what?
And what tools are currently available? (Preferably open source of course, but commercial options considered.)
Re:THAT'S considered an acceptible release bug???
on
Mozilla 1.4 RC1
·
· Score: 1
(this is an RC as you pointed out)
I thought RC meant release canditate; I'm not a developer but that seems to suggest they hope to release this version if no unknown major bugs show up in beta testing. Although I don't think I've ever noticed an RC1 going straight to final.
I'm a bit surprised about the ATI thing. JRE 1.4.1_01 had the same problems, but if you disabled the DirectX stuff in the VM it would work. I wonder if there's a similar workaround for Mozilla or if the JRE is what's actually crashing.
Well, you can always change the startup graphic and title bar and release it as Kjellazilla 2.0.
I kind of agree it's a major change, but I--as you--am insterested in it as my primary browser. I'd just as soon call it Phoenix or Firebird. But the developers get to call it whatever they want on mozilla.org.
Or maybe we can release it as PostgreSQL 2.0 or MySQL 2.0!
It doesn't matter. Hell, we hate to go out in the rain for a beer run. To go to one of these other planets takes a hell of a lot of planning and a long trip through the cold vacuum of space with lots of radiation. There's no rain, but they don't let you drink beer.
Oops. I was afraid I'd mess that up. (Was typing from memory.) SQL Slammer was the big one, but there was a follow-up attack that used the other port and other protocol, but I forget how that worked.
The SQL Slammer worm (I thought it was 3 tries in a row, but not in today's logs):
Here's the TCP 1433 one. (Aha, this was Spida.) This may be the 3-in-a-row one, but I was filling my pipe today with BitTorrents of Knoppix 3.2 (deleted it; d'oh!) and ClusterKnoppix so some packets probably dropped:
May 29 07:33:46 (myhostname) kernel: Packet log: input DENY eth0 PROTO=6 200.56.97.98:5612 xxx.xxx.xxx.xxx:1433 L=48 S=0x00 I=37032 F=0x4000 T=114 SYN (#11) May 29 07:33:49 (myhostname) kernel: Packet log: input DENY eth0 PROTO=6 200.56.97.98:5612 xxx.xxx.xxx.xxx:1433 L=48 S=0x00 I=37117 F=0x4000 T=114 SYN (#11)
(These are syslog messages logged by ipchains rules.)
While I'm blabbering on, here are Apache log examples of Code Red and Nimda. (Code red intentionally changed so I'm not posting virus code.)
Here's Nimda, although I haven't had it since Monday oddly enough. (Then again, Apache's been killed by the VM daily because I goofed up the swap file and didn't figure it out until today.)
Crap. I thought I could do it with wget, but wget dutifully followed the HTTP 302 redirect which sent the message anyway. (Because the message is encoded in the URL to give feedback to Caldera/SCO's web site.)
And from work, too. Great. My IP just got logged in the proxy server with all that profanity.
I'm sure there's a way to tell some www user agent to not follow the redirect, but I'm not going to tempt fate further from work.
I'm convinced this is a revolution in filesharing because it solves the two biggest Problems filesharing has, crappy downloads and getting sued.
I couldn't read the artilcle (link wouldn't resolve), but from what the others are saying you have to trust every node.
You'd really have to trust your "friends", and I don't see how finding new friends would be made easier with Waste as opposed to real life CD swapping or whatever.
Heh, they aren't guys, they're worms. The persistent ones probably share two or three octets of your IP address. The default.ida attacks are from Code Red variants and the cmd.exe attacks are from Nimda. Both attack IPs somewhat randomly but are weighted to attack 'nearby' IPs far more frequently. It was a very effective tactic, too.
Those are unpatched infected IIS servers doing that. It's really sad. Code Red almost went away for a few months but now I'm getting more Code Red hits than Nimda, but I see both every day. The SQL Slammer worm is still a daily occurrence, too.
(TCP port 1433 three SYNs in a row).
My 'intrusion detection' so far is from Apache and ipchains logs. I think I'll install Snort because I'm curious how many NetBIOS attacks I'm getting.
Uh-oh. That could be possible. I think it would be more effective to use a hacked tracker that would send out the DDOS target as a peer to many/all clients.
Then again the script kiddies have plenty of DDOS capability with thier IRCbot trojans, so they won't bother.
The problem that I see with any sort of caching or redistribution of Slashdotted sites is legal issues. Most (if not all) articles are copyrighted, and most have ads. The rest is left as an exercise for the reader.
To solve the Slashdot effect without advance permission I would think servers would have to have a failover-to-bittorrent mode built into Apache (for example) and an HTTP-by-bittrorrent concept that the previous mod 5 poster suggested.
Incidentally, this is possible to do for well-behaved browsers without altering the URL. Just change the mime type and send the.torrent file instead of the content. I don't know if IE is that smart, so you'd want a URL without an extension so IE would either use mime or autodetect or whatever it does to get the right handler for the file (instead of using the extension which I think it does by default).
extending the HTTP protocol to allow web sites to become p2p.
Very cool idea!
I'm not running a FreeNet node because I know there's likely to be some stuff I strongly object to (to the point of being physically ill) running through my node, even though it's encrypted and I'd never know for sure.
However, if BitTorrent could be used to easily deliver web sites I'd gladly host a few seeds and maybe even a tracker for community-based sites that interest me.
The publishing and seeding is doable now with some scripting glue, but the need would be to make every browser capable of obtaining and downloading the page/site/portions-of-site as easily as opening any other web page.
Perhaps what could be done now would be a newsletter-like publication where you download a directory with the HTML/JPEG or PDF file(s) and an auto-open function in the BT client for./index.html or./index.pdf or similar.
Freenet (of The FreeNet Project) is high-latency. BitTorrent constantly makes new connections. I don't see a way of merging the two.
I believe Freenet already behaves a lot like BitTorrent as far as splitting files into chunks and distributing them fairly randomly among peers. (Or perhaps you have to break a large file into pieces yourself when inserting.) Instead of a tracker you have the key.
I think the constant communication and comparing notes that BitTorrent does would be way slow if done realtime over FreeNet.
Another big difference between BitTorrent and FreeNet is that BitTorrent continually reports and evaluates peers and what available chunks the peers have while FreeNet is designed to let no one--not even the node owner--know what data is on the node.
It seems to me a lot of people miss the point of BitTorrent. (Or maybe I have it wrong.) Don't compare it to Kazaa or Napster or Freenet. Compare it to an HTTP download: You have a website. You want to offer a large, popular file and have no legal entanglements preventing you from doing so. With HTTP you pay for the bandwidth and if it's slashdotted (speaking generically) then everyone loses. With BitTorrent the slashdotting generally improves the download for everyone. Plus, you don't even have to seed the file yourself, and in some cases you don't have to run the tracker yourself. As far as I know you can't offer a URL for a file on Kazaa or other p2p apps, and a Freenet URL would assume to much about the end-user's setup.
I'm thrilled with BitTorrent and would like to thank Bram Cohen for this gem. It's just what Open Source/Free Software needs.
I've thought about software distribution and imaging via BitTorrent. It would take some work, but if the tracker could optimize to keep peers on local switches together or if mulitple trackers were launched (one for each switch) you could push files (software & images) very quickly and efficiently and work those GB+ backplanes.
My department uses Novell Zenworks for application distribution, and occasionally everyone needs an app pushed at once (business critical app has server upgrade requiring simultaneous client upgrade), and we have these cool switches but the bottleneck is 10MB/s and each client is a redundant TCP stream. Oh well. I can imagine using a BitTorrent push and watching those switches light up to 100% UTL. It would be worth it just for the light show.
Very good question. I wondered about this, too. If BitTorrent were to have 'preferred peers' (or seeds) then Akamai would be ideal. Think big company downloads like Microsoft Update, Real, Macromedia, Download.com, Sourcefourge, etc..
Or perhaps Akamai would be low-priority peers for those who are behind NAT or otherwise difficult to peer, since the idea would be to use the consumer's bandwidth to defray costs.
I'm curious how p2p really compares to the www, bandwidth-wise.
Don't underestimate the www. I worked at a data center for a while and was amazed at one customer we had who kept between 20 and 30MiB/s bandwidth continuously. It was a discount airline ticket site (not belonging to an airline, and not one of the big names).
Another customer had two pipes to the internet: one for their MMOG server cluster and one for their web site. The web site used at least 10x more bandwidth. They were going to upgrade one pipe to 1GiB/s....their web server. (But then they had financial problems and moved to a cheap data center.)
Slashdot Mirror
Oh, and I also worry about stuff like this being enforced in the US. (I'm a US'ian) If we are more strict with our IP laws than other countries, then our companies' ability to compete on the international market could be hampered.
I don't know much about Patent law, but I always heard that patents were easy to engineer around because they have to be so specific, and a relatively minor change can be used to engineer around the patent.
How does this extend to computer programming techniques? Why don't you have to provide a specific algorithm in the patent that can be engineered around?
...beer is supposed to be warm, and that us puny US american types are the only people who drink it icy cold.
Well, they say it's supposed to be served at room temperature. "Warm" is a relative term, and "room temperature" in Europe averages 34 degrees Fahrenheit. Those Europeans just don't realize how much warmer it is in the new world.
I quote from the last page of his site, which was not at all difficult to miss if you rtfa -
:-)
Yes, but if you'd look at the pictures on the second-to-last page it appears the thermometer's sensor is on the peltier cooler itself. Even if that's a misinterpretation on my part it's clear there is no probe in the Guiness itself but outside the glass at best.
By the way, I haven't tried the Guinnes-in-a-can yet. I'm not a total snob, but other beers aren't as good in the can, so I assumed Guiness would have the same problem. How is it? And can you pour yourself a four-leaf clover in the head with the can?
My sister did it once: she gulped the last few ounces of her Guiness because we were ready to go. I stared at her in horror as she looked at me like it was no big deal. She was feeling sick a couple of minutes later. I wasn't insensitive enough to ask how much chest hair she'd grown because of it.
No, silly. That's passe. People will be sshnuke'ing the whole internet looking for backdoors (and wearing tight leather). That's the current hip thing.
I expect Java will continue to survive in some niches (most notably, bloated web services implementations)
I was about to install Java + JBoss + Tomcat + Apache httpd (and maybe Apache Cocoon) and see what I can do with it. I'm not a seasoned developer (just had a few classes a few years ago), so I'm trying to figure out the best way to offer a scalable web platfrom mainly for database entry and reporting without much custom programming. I'm also interested in application services, but that's more into the future.
If not Java, then what?
And what tools are currently available? (Preferably open source of course, but commercial options considered.)
(this is an RC as you pointed out)
I thought RC meant release canditate; I'm not a developer but that seems to suggest they hope to release this version if no unknown major bugs show up in beta testing. Although I don't think I've ever noticed an RC1 going straight to final.
I'm a bit surprised about the ATI thing. JRE 1.4.1_01 had the same problems, but if you disabled the DirectX stuff in the VM it would work. I wonder if there's a similar workaround for Mozilla or if the JRE is what's actually crashing.
Well, you can always change the startup graphic and title bar and release it as Kjellazilla 2.0.
I kind of agree it's a major change, but I--as you--am insterested in it as my primary browser. I'd just as soon call it Phoenix or Firebird. But the developers get to call it whatever they want on mozilla.org.
Or maybe we can release it as PostgreSQL 2.0 or MySQL 2.0!
It doesn't matter. Hell, we hate to go out in the rain for a beer run. To go to one of these other planets takes a hell of a lot of planning and a long trip through the cold vacuum of space with lots of radiation. There's no rain, but they don't let you drink beer.
Forget it.
Oops. I was afraid I'd mess that up. (Was typing from memory.) SQL Slammer was the big one, but there was a follow-up attack that used the other port and other protocol, but I forget how that worked.
/default.ida?XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX%ublah-blah-b lah-url-encoded-code-red-worm---%u0000%u00=a HTTP/1.0" 404 205
/scripts/root.exe?/c+dir HTTP/1.0" 404 210 /MSADC/root.exe?/c+dir HTTP/1.0" 404 208 /c/winnt/system32/cmd.exe?/c+dir HTTP/1.0" 404 218 /d/winnt/system32/cmd.exe?/c+dir HTTP/1.0" 404 218 /scripts/..%255c../winnt/system32/cmd.exe?/c+dir HTTP/1.0" 404 232 /_vti_bin/..%255c../..%255c../..%255c../winnt/syst em32/cmd.exe?/c+dir HTTP/1.0" 404 249 /_mem_bin/..%255c../..%255c../..%255c../winnt/syst em32/cmd.exe?/c+dir HTTP/1.0" 404 249 /msadc/..%255c../..%255c../..%255c/..%c1%1c../..%c 1%1c../..%c1%1c../winnt/system32/cmd.exe?/c+dir HTTP/1.0" 404 265 /scripts/..%c1%1c../winnt/system32/cmd.exe?/c+dir HTTP/1.0" 404 231 /scripts/..%c0%2f../winnt/system32/cmd.exe?/c+dir HTTP/1.0" 404 231 /scripts/..%c0%af../winnt/system32/cmd.exe?/c+dir HTTP/1.0" 404 231 /scripts/..%c1%9c../winnt/system32/cmd.exe?/c+dir HTTP/1.0" 404 231 /scripts/..%%35%63../winnt/system32/cmd.exe?/c+dir HTTP/1.0" 400 215 /scripts/..%%35c../winnt/system3
The SQL Slammer worm (I thought it was 3 tries in a row, but not in today's logs):
May 29 16:03:10 (myhostname) kernel: Packet log: input DENY eth0 PROTO=17 66.108.147.93:1339 xxx.xxx.xxx.xxx:1434 L=404 S=0x00 I=62109 F=0x0000 T=110 (#12)
Here's the TCP 1433 one. (Aha, this was Spida.) This may be the 3-in-a-row one, but I was filling my pipe today with BitTorrents of Knoppix 3.2 (deleted it; d'oh!) and ClusterKnoppix so some packets probably dropped:
May 29 07:33:46 (myhostname) kernel: Packet log: input DENY eth0 PROTO=6 200.56.97.98:5612 xxx.xxx.xxx.xxx:1433 L=48 S=0x00 I=37032 F=0x4000 T=114 SYN (#11)
May 29 07:33:49 (myhostname) kernel: Packet log: input DENY eth0 PROTO=6 200.56.97.98:5612 xxx.xxx.xxx.xxx:1433 L=48 S=0x00 I=37117 F=0x4000 T=114 SYN (#11)
(These are syslog messages logged by ipchains rules.)
While I'm blabbering on, here are Apache log examples of Code Red and Nimda. (Code red intentionally changed so I'm not posting virus code.)
66.73.162.229 - - [29/May/2003:16:38:29 -0500] "GET
Here's Nimda, although I haven't had it since Monday oddly enough. (Then again, Apache's been killed by the VM daily because I goofed up the swap file and didn't figure it out until today.)
66.36.142.101 - - [26/May/2003:17:14:20 -0500] "GET
66.36.142.101 - - [26/May/2003:17:14:21 -0500] "GET
66.36.142.101 - - [26/May/2003:17:14:21 -0500] "GET
66.36.142.101 - - [26/May/2003:17:14:22 -0500] "GET
66.36.142.101 - - [26/May/2003:17:14:22 -0500] "GET
66.36.142.101 - - [26/May/2003:17:14:22 -0500] "GET
66.36.142.101 - - [26/May/2003:17:14:23 -0500] "GET
66.36.142.101 - - [26/May/2003:17:14:23 -0500] "GET
66.36.142.101 - - [26/May/2003:17:14:23 -0500] "GET
66.36.142.101 - - [26/May/2003:17:14:24 -0500] "GET
66.36.142.101 - - [26/May/2003:17:14:24 -0500] "GET
66.36.142.101 - - [26/May/2003:17:14:28 -0500] "GET
66.36.142.101 - - [26/May/2003:17:14:28 -0500] "GET
66.36.142.101 - - [26/May/2003:17:14:29 -0500] "GET
Crap. I thought I could do it with wget, but wget dutifully followed the HTTP 302 redirect which sent the message anyway. (Because the message is encoded in the URL to give feedback to Caldera/SCO's web site.)
And from work, too. Great. My IP just got logged in the proxy server with all that profanity.
I'm sure there's a way to tell some www user agent to not follow the redirect, but I'm not going to tempt fate further from work.
Funny, I suddenly have very amorous feelings for Linux after reading that.
I'm convinced this is a revolution in filesharing because it solves the two biggest Problems filesharing has, crappy downloads and getting sued.
I couldn't read the artilcle (link wouldn't resolve), but from what the others are saying you have to trust every node.
You'd really have to trust your "friends", and I don't see how finding new friends would be made easier with Waste as opposed to real life CD swapping or whatever.
...some of these guys are dammed persistant...
Heh, they aren't guys, they're worms. The persistent ones probably share two or three octets of your IP address. The default.ida attacks are from Code Red variants and the cmd.exe attacks are from Nimda. Both attack IPs somewhat randomly but are weighted to attack 'nearby' IPs far more frequently. It was a very effective tactic, too.
Those are unpatched infected IIS servers doing that. It's really sad. Code Red almost went away for a few months but now I'm getting more Code Red hits than Nimda, but I see both every day. The SQL Slammer worm is still a daily occurrence, too.
(TCP port 1433 three SYNs in a row).
My 'intrusion detection' so far is from Apache and ipchains logs. I think I'll install Snort because I'm curious how many NetBIOS attacks I'm getting.
Uh-oh. That could be possible. I think it would be more effective to use a hacked tracker that would send out the DDOS target as a peer to many/all clients.
Then again the script kiddies have plenty of DDOS capability with thier IRCbot trojans, so they won't bother.
The problem that I see with any sort of caching or redistribution of Slashdotted sites is legal issues. Most (if not all) articles are copyrighted, and most have ads. The rest is left as an exercise for the reader.
.torrent file instead of the content. I don't know if IE is that smart, so you'd want a URL without an extension so IE would either use mime or autodetect or whatever it does to get the right handler for the file (instead of using the extension which I think it does by default).
To solve the Slashdot effect without advance permission I would think servers would have to have a failover-to-bittorrent mode built into Apache (for example) and an HTTP-by-bittrorrent concept that the previous mod 5 poster suggested.
Incidentally, this is possible to do for well-behaved browsers without altering the URL. Just change the mime type and send the
extending the HTTP protocol to allow web sites to become p2p.
./index.html or ./index.pdf or similar.
Very cool idea!
I'm not running a FreeNet node because I know there's likely to be some stuff I strongly object to (to the point of being physically ill) running through my node, even though it's encrypted and I'd never know for sure.
However, if BitTorrent could be used to easily deliver web sites I'd gladly host a few seeds and maybe even a tracker for community-based sites that interest me.
The publishing and seeding is doable now with some scripting glue, but the need would be to make every browser capable of obtaining and downloading the page/site/portions-of-site as easily as opening any other web page.
Perhaps what could be done now would be a newsletter-like publication where you download a directory with the HTML/JPEG or PDF file(s) and an auto-open function in the BT client for
Freenet (of The FreeNet Project) is high-latency. BitTorrent constantly makes new connections. I don't see a way of merging the two.
I believe Freenet already behaves a lot like BitTorrent as far as splitting files into chunks and distributing them fairly randomly among peers. (Or perhaps you have to break a large file into pieces yourself when inserting.) Instead of a tracker you have the key.
I think the constant communication and comparing notes that BitTorrent does would be way slow if done realtime over FreeNet.
Another big difference between BitTorrent and FreeNet is that BitTorrent continually reports and evaluates peers and what available chunks the peers have while FreeNet is designed to let no one--not even the node owner--know what data is on the node.
It seems to me a lot of people miss the point of BitTorrent. (Or maybe I have it wrong.) Don't compare it to Kazaa or Napster or Freenet. Compare it to an HTTP download: You have a website. You want to offer a large, popular file and have no legal entanglements preventing you from doing so. With HTTP you pay for the bandwidth and if it's slashdotted (speaking generically) then everyone loses. With BitTorrent the slashdotting generally improves the download for everyone. Plus, you don't even have to seed the file yourself, and in some cases you don't have to run the tracker yourself. As far as I know you can't offer a URL for a file on Kazaa or other p2p apps, and a Freenet URL would assume to much about the end-user's setup.
I'm thrilled with BitTorrent and would like to thank Bram Cohen for this gem. It's just what Open Source/Free Software needs.
I've thought about software distribution and imaging via BitTorrent. It would take some work, but if the tracker could optimize to keep peers on local switches together or if mulitple trackers were launched (one for each switch) you could push files (software & images) very quickly and efficiently and work those GB+ backplanes.
My department uses Novell Zenworks for application distribution, and occasionally everyone needs an app pushed at once (business critical app has server upgrade requiring simultaneous client upgrade), and we have these cool switches but the bottleneck is 10MB/s and each client is a redundant TCP stream. Oh well. I can imagine using a BitTorrent push and watching those switches light up to 100% UTL. It would be worth it just for the light show.
Very good question. I wondered about this, too. If BitTorrent were to have 'preferred peers' (or seeds) then Akamai would be ideal. Think big company downloads like Microsoft Update, Real, Macromedia, Download.com, Sourcefourge, etc..
Or perhaps Akamai would be low-priority peers for those who are behind NAT or otherwise difficult to peer, since the idea would be to use the consumer's bandwidth to defray costs.
Are you running a firewall? Does it reject or just ignore requests on closed ports? Does it restrict ICMP in any way?
I'm surprised the software would keep trying if it's getting an ICMP "connection refused" message back.
I'm curious how p2p really compares to the www, bandwidth-wise.
Don't underestimate the www. I worked at a data center for a while and was amazed at one customer we had who kept between 20 and 30MiB/s bandwidth continuously. It was a discount airline ticket site (not belonging to an airline, and not one of the big names).
Another customer had two pipes to the internet: one for their MMOG server cluster and one for their web site. The web site used at least 10x more bandwidth. They were going to upgrade one pipe to 1GiB/s....their web server. (But then they had financial problems and moved to a cheap data center.)
Whoah.
Unfortunately, that standard is patent-encumbered.