Yessss!!! Knoppix-on-a-stick!!! And writable too! Well, that would mean that most of the time you won't need to carry a laptop around, as long as you know you can "parasite" someone else's machine (think conference venue, University labs, public libraries, cybercafes...) with your own OS + configs + data...
Plan 9 From Outer Space is a very cheaply made B-movie by Ed Wood. It's kinda entertaining in its own way. You can easily tell that the flying saucers in there are very cheap props hanging by a thread... Tim Burton is a big Ed Wood fan, he even did a movie bearing his name which I haven't seen, but Mars Attacks is also some sort of tribute to him I guess.
Very good advice. I have a two hour slot per week allocated to "a project with a friend". Sometimes it is work-related, sometimes it is admin stuff, sometimes it is a lesson or a hobby. The goal is that we get to work on something that we have kept postponing.
We make sure of course that it is realizable in two hours. If it is too big, we cut it in two hour chunks. Usually in that case the first two hours is all we needed to get going, and after that we don't need to keep the friend around;-)
Sometimes both of us spend the two hours on my project, sometimes it's the other way around, sometimes we work each on our own project. The trick is not to be selfish, and draw the same satisfaction no matter whose project you're working on. After all it is only two hours.
The benefit is the social interaction, the outsider view, but especially the feeling of accomplishment. I've been doing this for 4 weeks now, and it's just great! We call it "Fight Club Saturdays"...
SIP and Jabber are indeed both text-based, SIP looks like HTTP, Jabber is XML based. Both are extensible. SIP is already used in internet telephony. Jabber is open-source. SIMPLE is based on SIP and XMPP is Jabber, and both are pushing for standardization. SIMPLE has a head start, but Jabber is lobbying hard.
What's spookier is that I got the movie ref, but was puzzled by the mention of Clooney, until of course I remembered that "they" made a remake of it...
Yeah JXTA's nice, but you still need to do some programming to get something useful out of it, as it is a framework, not a network per se. It has indeed a nice pipe-based architecture.
I'm being redundant with my a previous post of mine, but you might want to check out U-P2P. You can set up a Napster-style P2P network without any programming. We're working on a Gnutella adapter too.
...or how about at least give the creator of a P2P community a framework to help him build one from scratch, and pick and choose specific protocols and other attributes based on his/her needs? Ideally not much programming knowledge should be required.
Check out our U-P2P effort, it is going in that direction.
The thing with the web though is that not everybody has the means to mirror a document that they find useful. It's due to the client-server nature of the beast. On a peer-to-peer network on the other hand, in principle anybody can share any document. As you just said, the popular ones will always be around, simply because they are of use to somebody.
If Freenet didn't have as a specific drawback the inability of someone to guarantee that a document remains hosted as long as they are willing to host it, Freenet would be a good choice for this.
There is quite a bit of academic work going on that tries indeed to preserve the persistence of documents regardless of their location. More or less like Freenet, most of the systems use the idea of distributed hashtables and some distributed routing algorithm to locate the doument based on its unique (hashed) id.
If you're interested, check out CAN, CHORD, PASTRY, OceanStore, Publius... and the Proceedings of the IPTPS workshops.
I think this is a fantastic idea! Why don't you Mozilla guys contact Google? The average Joe that does his browsing with IE and his searching with Google would probably be interested in a "Google Web Browser".
Mozilla firebird?? How many hits with a clue stick will it take to get die-hard netscape 4 users to switch to that?!?
It took a few hits, but here I am;-) The thing is, Netscape 4's email is pretty good (got all the features I need and nothing more, and it is gentle on RAM), so I was keeping it around to check mail, and I was browsing with an old IE. Yep, call me a luddite!
I tried Mozilla (and Netscape 7), but the general look and feel didn't appeal to me, and my fairly old PC didn't enjoy the memory usage.
This is until I discovered Firebird, and it was love at first sight! No need for IE anymore, but what to do to replace my email? Well, now I've converted to Thunderbird as well, but TBird + Firebird = memory hogs, and I still miss Netscape mail (is there a Netscape 4.x theme somewhere for nostalgics?;-))... Thunderbird's junk mail filter is nice, but I'm running SpamAssassin on the server, so I don't really need it.
Basically I'm waiting for Firebird and Thunderbird to replace the current Mozilla suite, and hopefully there will a drop in combined memory usage...
The interview changed my mind about C++. Before, I felt that it was only my ineptitude that made it hard for me to create classes in C++. Now, I understand that I may not be alone in my ineptitude (being part of the "rank and file" as you nicely put it), and that there is nothing wrong with using C++ simply as a way to access nice higher-level libraries and ADTs that couldn't have existed in pure C.
If a purer OO approach is needed, then just use a pure OO language instead! (insert Ruby/Smalltalk/... based on your religion)
I have a very candid question here, so please be gentle...
With improvements in DRM and better compression, there should be a new way to be present on the internet, all the while protecting copyright.
Why don't record companies (or the artists themselves in fact) just circumvent the CD business altogether and make their new productions only available for sale on the Web, using a DRM protected format? This way *nobody* can ever make a copy of the song unless the seller allows it explicitly.
Apple's iTunes has all the needed technology, they could start offering songs that aren't available anywhere else.
Everybody wins: the seller saves on packaging and distribution, the consumer gets a better price, the artist presumably gets a bigger chunk of the pie, especially if the artist IS the seller as well (it shouldn't be very expensive to implement iTunes-type technology, or am I missing something?). The losers: the HMVs of this world...
One objection could be that buyers also want to have the songs on CD, but then again we've already shifted from vinyls to cassettes to CDs, maybe the next shift is to iPod type devices?
Is this already implemented by a music label or an artist? I know of musicians like Prince who have made some of their music available on the Web only, but I don't think they have sold that music using DRM.
P2P is not only about sharing music and video. There's for example all the work on distributed hashtables (Can, Chord, Pastry...) or distributed processing (SETI@Home and others).
But even with regular file-sharing you could do much more if you could search and publish files other than MP3s. Chemists or biologists could share melocule or gene descriptions, programmers could share design patterns, without depending on a central database that they do not control... But for this to work, you need a mechanism to provide (explicitly or not) metadata about the documents, because file names won't be enough.
If you're interested, check out the project I am working on: it is called U-P2P [sourceforge.net], and it allows the explicit description of documents, as well as the description of the specialized file-sharing community itself. We also published a paper at the FREENIX track of USENIX'03 this year, it should be available on their web site. If not, drop me a message and I'll send it to you.
The problem would solve itself if the bandwidth was paid entirely for by the end which is downloading the data, rather than serving it. Then the ISPs would have to pay to download from sites, and payments to sites would become part of a customer's ISP bill.
That used to be (and still is) the business model for the Minitel system in France, which was around way before the Web became popular. Because there was a single Minitel service provider (France Telecom), someone who wanted to set up a site would register a name (just like internet domain names), and pay some fixed monthly fee to France Telecom be able to provide the service. But the end user had to pay up to a 1$/minute to access those sites, which would eventually show up on the phone bill (not sure if the sites were listed, could be embarrassing if the site you visited was pr0n...). The money was then divided between France Telecom and the service provider.
So basically back then, the end user had to pay to obtain a service on a per minute rate, and the service could actually make good money if the service was of any use. Dotcom business plans were definitely viable!
Unfortunately for the Minitel, the Web was so much cheaper and more international that it overtook it. Also France Telecom didn't market the idea with much conviction outside French borders.
So when there IS an abundance of local talent (evidenced by the posting volume of/. posts) why aren't H1B's sent home?
Maybe because you can't treat people like you'd treat old CPUs? So you bring people in when you need them, and then dispose of them when you don't? "H1Bs", like most human beings, have a family, kids who go to school, a home, friends...
Now having said that, I agree that any immigration policy has to be crafted carefully, with considerations of the impact on the long term economy, as well as cultural and social integation, so that it contributes positively to the melting-pot.
Most European countries went through the same crisis 30 years ago, when they brought in loads of blue-collar workers, who solved a punctual economic need, but who also took ages to fit in. All it did was generate racism on one side and fundamentalism on the other.
I believe there is an ongoing effort to harmonize EU diplomas. For example I heard that France will now start to offer 2 year masters degrees instead of the 1-year DEA. It will also include a full-blown thesis instead of the small 4 month "stage".
Well, supposedly the computer can't make a mistake, can it?
If anything, the computer will probably miss more mistakes than flag correct things as mistakes (ie there will be more false negatives than false positives), so it'll mark more generously than a human teacher.
It all depends on what exactly you are teaching. You could be teaching an introduction to programming, or data structures and algorithms, or Object-Oriented programming, or network computing, or... Java would probably be suitable for either of the items above, and there are textbooks covering each of them.
If you are safely assuming that your students already know how to program (for example in C) and know about data structures, then Java becomes an ideal choice to introduce object-oriented programming. If this is what you want to do, then "Object Oriented Software Development Using Java" by Xiaoping Jia is a very good bet. It introduces OO with UML, talks about patterns, and uses them to discuss the collections framework and Swing. Search for "Jia" on Amazon.com to learn more...
Slashdot Mirror
Yessss!!! Knoppix-on-a-stick!!! And writable too! Well, that would mean that most of the time you won't need to carry a laptop around, as long as you know you can "parasite" someone else's machine (think conference venue, University labs, public libraries, cybercafes...) with your own OS + configs + data...
Plan 9 From Outer Space is a very cheaply made B-movie by Ed Wood. It's kinda entertaining in its own way. You can easily tell that the flying saucers in there are very cheap props hanging by a thread... Tim Burton is a big Ed Wood fan, he even did a movie bearing his name which I haven't seen, but Mars Attacks is also some sort of tribute to him I guess.
Very good advice. I have a two hour slot per week allocated to "a project with a friend". Sometimes it is work-related, sometimes it is admin stuff, sometimes it is a lesson or a hobby. The goal is that we get to work on something that we have kept postponing.
;-)
We make sure of course that it is realizable in two hours. If it is too big, we cut it in two hour chunks. Usually in that case the first two hours is all we needed to get going, and after that we don't need to keep the friend around
Sometimes both of us spend the two hours on my project, sometimes it's the other way around, sometimes we work each on our own project. The trick is not to be selfish, and draw the same satisfaction no matter whose project you're working on. After all it is only two hours.
The benefit is the social interaction, the outsider view, but especially the feeling of accomplishment. I've been doing this for 4 weeks now, and it's just great! We call it "Fight Club Saturdays"...
AFAIK, XMPP == Jabber.
(only the AC parent got it all right!)
SIP and Jabber are indeed both text-based, SIP looks like HTTP, Jabber is XML based. Both are extensible. SIP is already used in internet telephony. Jabber is open-source. SIMPLE is based on SIP and XMPP is Jabber, and both are pushing for standardization. SIMPLE has a head start, but Jabber is lobbying hard.
What's spookier is that I got the movie ref, but was puzzled by the mention of Clooney, until of course I remembered that "they" made a remake of it...
Yeah JXTA's nice, but you still need to do some programming to get something useful out of it, as it is a framework, not a network per se. It has indeed a nice pipe-based architecture.
I'm being redundant with my a previous post of mine, but you might want to check out U-P2P. You can set up a Napster-style P2P network without any programming. We're working on a Gnutella adapter too.
...or how about at least give the creator of a P2P community a framework to help him build one from scratch, and pick and choose specific protocols and other attributes based on his/her needs? Ideally not much programming knowledge should be required.
Check out our U-P2P effort, it is going in that direction.
The thing with the web though is that not everybody has the means to mirror a document that they find useful. It's due to the client-server nature of the beast. On a peer-to-peer network on the other hand, in principle anybody can share any document. As you just said, the popular ones will always be around, simply because they are of use to somebody.
You'd think they would at least stop subjecting their paying subscribers to ads.
(BTW same thing can be said about pay TV or anything else that you don't get for free)
I think this is a fantastic idea! Why don't you Mozilla guys contact Google? The average Joe that does his browsing with IE and his searching with Google would probably be interested in a "Google Web Browser".
It took a few hits, but here I am ;-) The thing is, Netscape 4's email is pretty good (got all the features I need and nothing more, and it is gentle on RAM), so I was keeping it around to check mail, and I was browsing with an old IE. Yep, call me a luddite!
I tried Mozilla (and Netscape 7), but the general look and feel didn't appeal to me, and my fairly old PC didn't enjoy the memory usage.
This is until I discovered Firebird, and it was love at first sight! No need for IE anymore, but what to do to replace my email? Well, now I've converted to Thunderbird as well, but TBird + Firebird = memory hogs, and I still miss Netscape mail (is there a Netscape 4.x theme somewhere for nostalgics? ;-))... Thunderbird's junk mail filter is nice, but I'm running SpamAssassin on the server, so I don't really need it.
Basically I'm waiting for Firebird and Thunderbird to replace the current Mozilla suite, and hopefully there will a drop in combined memory usage...
My thoughts exactly!
The interview changed my mind about C++. Before, I felt that it was only my ineptitude that made it hard for me to create classes in C++. Now, I understand that I may not be alone in my ineptitude (being part of the "rank and file" as you nicely put it), and that there is nothing wrong with using C++ simply as a way to access nice higher-level libraries and ADTs that couldn't have existed in pure C.
If a purer OO approach is needed, then just use a pure OO language instead! (insert Ruby/Smalltalk/... based on your religion)
Simply spreading a rumor that such a thing happened might do the trick...
I have a very candid question here, so please be gentle...
With improvements in DRM and better compression, there should be a new way to be present on the internet, all the while protecting copyright.
Why don't record companies (or the artists themselves in fact) just circumvent the CD business altogether and make their new productions only available for sale on the Web, using a DRM protected format? This way *nobody* can ever make a copy of the song unless the seller allows it explicitly.
Apple's iTunes has all the needed technology, they could start offering songs that aren't available anywhere else.
Everybody wins: the seller saves on packaging and distribution, the consumer gets a better price, the artist presumably gets a bigger chunk of the pie, especially if the artist IS the seller as well (it shouldn't be very expensive to implement iTunes-type technology, or am I missing something?). The losers: the HMVs of this world...
One objection could be that buyers also want to have the songs on CD, but then again we've already shifted from vinyls to cassettes to CDs, maybe the next shift is to iPod type devices?
Is this already implemented by a music label or an artist? I know of musicians like Prince who have made some of their music available on the Web only, but I don't think they have sold that music using DRM.
really neat... of course though, it only works well as long as the dictionary itself is quite small...
I find that the "I'm tired of repetitive jokes" thing is getting repetitive as well... ;-)
P2P is not only about sharing music and video. There's for example all the work on distributed hashtables (Can, Chord, Pastry...) or distributed processing (SETI@Home and others).
But even with regular file-sharing you could do much more if you could search and publish files other than MP3s. Chemists or biologists could share melocule or gene descriptions, programmers could share design patterns, without depending on a central database that they do not control... But for this to work, you need a mechanism to provide (explicitly or not) metadata about the documents, because file names won't be enough.
If you're interested, check out the project I am working on: it is called U-P2P [sourceforge.net], and it allows the explicit description of documents, as well as the description of the specialized file-sharing community itself. We also published a paper at the FREENIX track of USENIX'03 this year, it should be available on their web site. If not, drop me a message and I'll send it to you.
Maybe because you can't treat people like you'd treat old CPUs? So you bring people in when you need them, and then dispose of them when you don't? "H1Bs", like most human beings, have a family, kids who go to school, a home, friends...
Now having said that, I agree that any immigration policy has to be crafted carefully, with considerations of the impact on the long term economy, as well as cultural and social integation, so that it contributes positively to the melting-pot.
Most European countries went through the same crisis 30 years ago, when they brought in loads of blue-collar workers, who solved a punctual economic need, but who also took ages to fit in. All it did was generate racism on one side and fundamentalism on the other.
I believe there is an ongoing effort to harmonize EU diplomas. For example I heard that France will now start to offer 2 year masters degrees instead of the 1-year DEA. It will also include a full-blown thesis instead of the small 4 month "stage".
Well, supposedly the computer can't make a mistake, can it?
If anything, the computer will probably miss more mistakes than flag correct things as mistakes (ie there will be more false negatives than false positives), so it'll mark more generously than a human teacher.
It all depends on what exactly you are teaching. You could be teaching an introduction to programming, or data structures and algorithms, or Object-Oriented programming, or network computing, or... Java would probably be suitable for either of the items above, and there are textbooks covering each of them.
If you are safely assuming that your students already know how to program (for example in C) and know about data structures, then Java becomes an ideal choice to introduce object-oriented programming. If this is what you want to do, then "Object Oriented Software Development Using Java" by Xiaoping Jia is a very good bet. It introduces OO with UML, talks about patterns, and uses them to discuss the collections framework and Swing. Search for "Jia" on Amazon.com to learn more...