That's a rough average of how many are online at a time, the last time I checked. That would be 10 or 100s of millions of Kazaa users. And there are many other P2P systems other that Kazaa.
Can RIAA sue them all? There is strength in numbers. People would not take them seriously, if the chance of getting caught is next to nothing (if you don't share your entire HD worth of britney sphere =b)
KISS was perhaps the first (that I know of at least) company that made a DVD console player which plays DivX. As a company that impressed me with their "cutting edge and geek" technology, it's unfortunate that it is exactly people like us, the geeks, who they are pissing off.
Respect is irrelevant. What matters is Paypal is the largest online payment service with the biggest user base, and you WILL offer Paypal as a payment option (if not the only) for your online services. The convenience of your users come first, whether you like it or not.
I say this as a webmaster who have considered various payment services.
seriously, the open sourcish license is interesting. It would work very well in an academic setting, which is analogous to actual OSS: people creating "products" with their primary goal not being money, and the GPLish license prevents it from being incorporated into proprietary, commercial products.
I don't know if you can do anything meaningful in your mom's basement with the STM kit, but interesting nonetheless.
6 "Ones"... I think there's only 1 "One", the program made to re-balance the Matrix, and is / was "re-incarnated" when needed. Basically everytime erasing his memory, and stick him into a new human body.
Curious if Neo would be in the "outer Matrix" after he died with Smith. Afterall, he fulfilled his "messianic" purpose, and his loss of purpose was actually what allowed him to "kill" Smith, as you suggested.
And on the shittiness of Matrix 2+... I don't agree. The premise is not as original as Matrix 1 and the storytelling may not be as straight forward to understand, but appreciate the depth of the script. It's rare in many movies today.
well, what matters is the technology and its development philosophy. A 3l33t name change isn't impossible.
BT does NOT work well with small files.
on
RSS & BT Together?
·
· Score: 1
The Bit Torrent protocol is designed to distribute large files, not many many small files. Torrents are partitioned into usually 128k - 1mb "pieces". So unless you embed media files into the RSS stream, which would make a case for using BT, it would be ridiculous to distribute small text streams that's not even as big as the usual lowest common denominator of the piece length.
And I think it's stupid to embed large media files into RSS streams. What's the point? Media files are more static than text content, which is what RSS is intended for. If you want to distribute big media files, just make torrents out of them and use BT as is today.
And FYI, I'm totally for using BT as a media distribution transport. I'm working on a DNS based "hub" which dynamically routes clients to trackers, in an attempt to solve the dead tracker and split swarm problems that generally plague the current BT file distribution "scene".
DRM is not a word most of us want to hear, but let's face it, it's coming whether you want it or not. So, besides the fact that "nothing is uncrackable", why not "embrace and extend" DRM?
I, for one, would welcome our open sourced DRM overlords, than the MS "trusted computing" counterpart. Besides, we all know Ogg is superior to WMA, right?
Hussein uses body doubles. How do we know it's not a fake? How much can we trust the Iran news source? How do we know it's not a conspiracy, or really, a PR move?
"... the increasing popularity of digital video recorders will actually cause a decline in ad revenues in the next few years."
why is that a surprise? Just like how the RIAA is dying (no BSD jokes here), business must adapt to technology. Technology has always changed society, adapt or you lose.
After RTFA'ing, the article's idea of using the DNS is not even about solving the tracker centralization problem, it's only a clever hack to use DNS to deliver the.torrent files.
A much better idea is to implement a P2P torrent distribution system, where the site that "post" the torrent only gives you a URL in somewhat the form of
btp://{tracker_URL}
then you connect to the swarm on that tracker, where some peers upload the originally.torrent metadata to get you started. Combine this with my DNS tracker routing idea (bthub.com as the tracker_URL), and you solve both problems of overloaded tracker and torrent web servers.
The original idea is neat, but as he said himself, should not and cannot be implemented on a mass scale.
I've circulated the idea of using DNS around the BitTorrent mail list for some time now, but it uses the DNS in a different way from distributing the torrent files themselves through DNS caches. It's doubtful putting data inside caches would "bog down the DNS system", but to turn DNS into essentially web servers is:
1. An ugly, ugly hack, and a wrong tool for the job (tm)
2. Wrong using others' resources in a way that is not intended (serving binary data)
My goal of using the DNS is however the same: solve the dead trackers problem. But to serve data directly from DNS, my idea to use what DNS is for: route traffic. You request the torrent's 40 byte info_hash as a subdomain of the DNS server's domain, and it returns you the optimal tracker's IP. For example:
{torrent_hash}.bthub.com
and bthub.com will be where i'll host the DNS service. It's basically a dynamic DNS server, updated frequently, of all known torrent hashes from different trackers in its database (my estimation is around 15000 - 20000 torrents, or subdomains. I know because I run a BT search engine, see my sig). This way, you don't "hack" the DNS, but you get the benefit of:
1. Dynamic routing to the tracker with the most ppl downloading your file, or the biggest "swarm", and hence the highest speed
2. Re-routing to a SINGLE new tracker when one dies
The 2nd benefit is important, as the case today is ppl just randomly pick a new tracker when the original dies, without knowing how big the swarm is on the new tracker. This splits the original torrent swarm across multiple trackers, which don't talk with each other (yet).
There are some discussion on the mail list of adding connectivity between the trackers, similar to how you link IRC servers together into a network. But that's no where near a proposal, and using DNS to act as a "hub" for the trackers would need the least effort and changes to the BT protocol and trackers.
1. List things to integrate and improve Linux UI on desktop (KDE, Gnome, etc. UI frameworks)
2. Get IBM, Redhat, etc. corporate support and funding
3. Attach weights to TODO items, distribute bounty $ by weights and total fundings available
4. ??
5. Uber Linux desktop (no, not profits this time.)
My rationale is that, as pointed out by some earlier comments, Linux UI's inconsistencies and often failure to work at all without hacking things around is due to no central direction and vision for UI design. IBM provides IT services in deploying OSS in companies while commercial distros like Redhat do QA on OSS packages, but besides KDE and Gnome, there needs to be a more coherent vision and a mean to make it happen.
If you are interested, I've created a temporary forum where we can discuss this further. Excuse the rest that's going on at the site ^^
ok, you understand RDBMS more than i do. Can you give some benchmarks of when Innodb would be slower than Postgresql? Is the scalability jokes you refered to, uses innodb? Can you give some example of what makes Postresql more scalable, beside being more sql compliant or closer to a "read rdbms"?
you see, i'm not a mysql zealot. Really. I only want to see facts in when is one faster than the other, what level of scalability one can reach over the other, etc. I'm not interested in theoretical debates of RDBMS or SQL compliance. I don't need the kitchen sink either, i just want something fast and reliable.
personally i don't care about sql compliance (as long as it's not wildly different), but yes reliablility is important. But, again, Innodb supports transaction and referential key constraints (although not very well). I have lost data (non important ones) in MyISAM when the server constantly crashed (hardware problems), but never when I used Innodb with some speed penalty.
and, just to join the bashing fun, vacuum in postgresql is gay =b
I run a search engine with peak 100 concurrent fulltext searches using Mysql, see my sig.
Historically, Innodb didn't exist until "recently". Mysql and Postgresql have their own strengths, but saying one is superior because it has more features is equally lame.
Feature-wise, I agree PostgreSQL is way more loaded. And full, native transaction support prob. makes it more reliable too. So use it if you are building a banking system (or anything that deals heavily with money, where reliablity is paramount). But Mysql is designed for speed and average use on websites, where performance is important. So stop saying Mysql sucks because it doesn't have feature X or feature Y.
For losing data and poor performance with updates and inserts, Innodb gives you transaction and row level locking. These problems apply only to the default MyISAM table type. I use MyISAM for heavy SELECT tables, and Innodb for more updated and inserted tables, and in ones I don't want to lose data in.
that is a good point. Connection to Mysql for example in PHP has the Pconnect option, which allows for persistent database connections. It would be nice if PHP's IMAP module implement a similar persistent connection as well, strictly for performance sake.
Speaking of webmail, IMP seems to be the most advanced of all the webmail apps written in PHP. Don't know how it solves the non-persistent imap connection problem, but there it says U of Penn is using it after evaluation of different systems. It didn't quite put performance into consideration however..
I agree that PHP is a template engine that's good for database / application abstraction only to a certain degree, and building a banking system using it is not using the right tool for the job. But Yahoo is fine using it for what it does, and it is an enterprise.
Slashdot Mirror
That's a rough average of how many are online at a time, the last time I checked. That would be 10 or 100s of millions of Kazaa users. And there are many other P2P systems other that Kazaa.
Can RIAA sue them all? There is strength in numbers. People would not take them seriously, if the chance of getting caught is next to nothing (if you don't share your entire HD worth of britney sphere =b)
KISS was perhaps the first (that I know of at least) company that made a DVD console player which plays DivX. As a company that impressed me with their "cutting edge and geek" technology, it's unfortunate that it is exactly people like us, the geeks, who they are pissing off.
They can "kiss" their mindshare goodbye =b
Respect is irrelevant. What matters is Paypal is the largest online payment service with the biggest user base, and you WILL offer Paypal as a payment option (if not the only) for your online services. The convenience of your users come first, whether you like it or not.
I say this as a webmaster who have considered various payment services.
^^
seriously, the open sourcish license is interesting. It would work very well in an academic setting, which is analogous to actual OSS: people creating "products" with their primary goal not being money, and the GPLish license prevents it from being incorporated into proprietary, commercial products.
I don't know if you can do anything meaningful in your mom's basement with the STM kit, but interesting nonetheless.
6 "Ones"... I think there's only 1 "One", the program made to re-balance the Matrix, and is / was "re-incarnated" when needed. Basically everytime erasing his memory, and stick him into a new human body.
And this time it happens to be Keanu's body ^.^
I wrote a little piece myself here.
Curious if Neo would be in the "outer Matrix" after he died with Smith. Afterall, he fulfilled his "messianic" purpose, and his loss of purpose was actually what allowed him to "kill" Smith, as you suggested.
And on the shittiness of Matrix 2+... I don't agree. The premise is not as original as Matrix 1 and the storytelling may not be as straight forward to understand, but appreciate the depth of the script. It's rare in many movies today.
well, what matters is the technology and its development philosophy. A 3l33t name change isn't impossible.
The Bit Torrent protocol is designed to distribute large files, not many many small files. Torrents are partitioned into usually 128k - 1mb "pieces". So unless you embed media files into the RSS stream, which would make a case for using BT, it would be ridiculous to distribute small text streams that's not even as big as the usual lowest common denominator of the piece length.
And I think it's stupid to embed large media files into RSS streams. What's the point? Media files are more static than text content, which is what RSS is intended for. If you want to distribute big media files, just make torrents out of them and use BT as is today.
And FYI, I'm totally for using BT as a media distribution transport. I'm working on a DNS based "hub" which dynamically routes clients to trackers, in an attempt to solve the dead tracker and split swarm problems that generally plague the current BT file distribution "scene".
DRM is not a word most of us want to hear, but let's face it, it's coming whether you want it or not. So, besides the fact that "nothing is uncrackable", why not "embrace and extend" DRM?
I, for one, would welcome our open sourced DRM overlords, than the MS "trusted computing" counterpart. Besides, we all know Ogg is superior to WMA, right?
Hussein uses body doubles. How do we know it's not a fake? How much can we trust the Iran news source? How do we know it's not a conspiracy, or really, a PR move?
i want to know
IE bug exploit link demo page with some promo for alternatives
load times is a wrong benchmark to pick, it doesn't show proccessing power well. Load times are mostly disk-bound.
Load times more than 3x faster IS a LOT faster. And no doubt they will have many many times faster actual data crunching speed.
funny, that's what I got a few months ago. Athlon XP (soon Athlon 64) + nForce 2 mobo offers the best bang for the buck.
why is that a surprise? Just like how the RIAA is dying (no BSD jokes here), business must adapt to technology. Technology has always changed society, adapt or you lose.
After RTFA'ing, the article's idea of using the DNS is not even about solving the tracker centralization problem, it's only a clever hack to use DNS to deliver the .torrent files.
.torrent metadata to get you started. Combine this with my DNS tracker routing idea (bthub.com as the tracker_URL), and you solve both problems of overloaded tracker and torrent web servers.
A much better idea is to implement a P2P torrent distribution system, where the site that "post" the torrent only gives you a URL in somewhat the form of
btp://{tracker_URL}
then you connect to the swarm on that tracker, where some peers upload the originally
The original idea is neat, but as he said himself, should not and cannot be implemented on a mass scale.
1. An ugly, ugly hack, and a wrong tool for the job (tm)
2. Wrong using others' resources in a way that is not intended (serving binary data)
My goal of using the DNS is however the same: solve the dead trackers problem. But to serve data directly from DNS, my idea to use what DNS is for: route traffic. You request the torrent's 40 byte info_hash as a subdomain of the DNS server's domain, and it returns you the optimal tracker's IP. For example:
{torrent_hash}.bthub.com
and bthub.com will be where i'll host the DNS service. It's basically a dynamic DNS server, updated frequently, of all known torrent hashes from different trackers in its database (my estimation is around 15000 - 20000 torrents, or subdomains. I know because I run a BT search engine, see my sig). This way, you don't "hack" the DNS, but you get the benefit of:
1. Dynamic routing to the tracker with the most ppl downloading your file, or the biggest "swarm", and hence the highest speed
2. Re-routing to a SINGLE new tracker when one dies
The 2nd benefit is important, as the case today is ppl just randomly pick a new tracker when the original dies, without knowing how big the swarm is on the new tracker. This splits the original torrent swarm across multiple trackers, which don't talk with each other (yet).
There are some discussion on the mail list of adding connectivity between the trackers, similar to how you link IRC servers together into a network. But that's no where near a proposal, and using DNS to act as a "hub" for the trackers would need the least effort and changes to the BT protocol and trackers.
I discussed this idea in detail on this thread "Decentralizing trackers: Use hubs", feel free to take a look.
2. Get IBM, Redhat, etc. corporate support and funding
3. Attach weights to TODO items, distribute bounty $ by weights and total fundings available
4. ??
5. Uber Linux desktop (no, not profits this time.)
My rationale is that, as pointed out by some earlier comments, Linux UI's inconsistencies and often failure to work at all without hacking things around is due to no central direction and vision for UI design. IBM provides IT services in deploying OSS in companies while commercial distros like Redhat do QA on OSS packages, but besides KDE and Gnome, there needs to be a more coherent vision and a mean to make it happen.
If you are interested, I've created a temporary forum where we can discuss this further. Excuse the rest that's going on at the site ^^
Linux UI bounties project
ok, you understand RDBMS more than i do. Can you give some benchmarks of when Innodb would be slower than Postgresql? Is the scalability jokes you refered to, uses innodb? Can you give some example of what makes Postresql more scalable, beside being more sql compliant or closer to a "read rdbms"?
you see, i'm not a mysql zealot. Really. I only want to see facts in when is one faster than the other, what level of scalability one can reach over the other, etc. I'm not interested in theoretical debates of RDBMS or SQL compliance. I don't need the kitchen sink either, i just want something fast and reliable.
personally i don't care about sql compliance (as long as it's not wildly different), but yes reliablility is important. But, again, Innodb supports transaction and referential key constraints (although not very well). I have lost data (non important ones) in MyISAM when the server constantly crashed (hardware problems), but never when I used Innodb with some speed penalty.
and, just to join the bashing fun, vacuum in postgresql is gay =b
Slashdot runs on Mysql last time I checked.
I run a search engine with peak 100 concurrent fulltext searches using Mysql, see my sig.
Historically, Innodb didn't exist until "recently". Mysql and Postgresql have their own strengths, but saying one is superior because it has more features is equally lame.
Feature-wise, I agree PostgreSQL is way more loaded. And full, native transaction support prob. makes it more reliable too. So use it if you are building a banking system (or anything that deals heavily with money, where reliablity is paramount). But Mysql is designed for speed and average use on websites, where performance is important. So stop saying Mysql sucks because it doesn't have feature X or feature Y.
For losing data and poor performance with updates and inserts, Innodb gives you transaction and row level locking. These problems apply only to the default MyISAM table type. I use MyISAM for heavy SELECT tables, and Innodb for more updated and inserted tables, and in ones I don't want to lose data in.
Speaking of webmail, IMP seems to be the most advanced of all the webmail apps written in PHP. Don't know how it solves the non-persistent imap connection problem, but there it says U of Penn is using it after evaluation of different systems. It didn't quite put performance into consideration however..
forgot to preview.
I agree that PHP is a template engine that's good for database / application abstraction only to a certain degree, and building a banking system using it is not using the right tool for the job. But Yahoo is fine using it for what it does, and it is an enterprise.