Thanks for the answers so far. As I wrote in the article, the problem is not so much the hosting of video files than the front end to put on. We have to integrate it with the NGO's website, put some informations on the videos... The only suggestion in that direction sofar is getDemocracy, which I dont find really ready for production.
Anyway, I'll keep looking, and thanks for the help !
Well, the problem was to know the question... I already had the answer... Well, it took million of years to compute it, so it's probably normal that Google cant get it in a few seconds...
Wow, I almost read the result was 42, which would explain a lot... I still dont know the great question, and it doesnt seem that Google will help me to find it...
Good code comes from good programmers. Whatever the methodology, you wont get a clean design from somebody who doesnt have experience. I think quite a few managers forget that. If you hire a guy just out of a technical school because you can pay him only 50K/year than dont wonder if you get dirty code. The book talk quite a bit about the importance to have competent people at the right place (seems like an obvious advice, but...)
Not sure documenting early will save much... In the project I'm working for the moment, there has been about 200 pages of docs before we had a single line of code. Every aspect they were thinking is there. But guess what... after 3 months of developpement, only about 10% of that docs is still valid. One of the issue is to know when to document and when to keep it an informal discussion. An out of sync documentation is probably worse than no docs. And as all things, the best aproach depends of the project. In an area where the requierments change fast (like my project) spending too much time documenting them is probably waste. If the requierments are stable for a couple of years than why not. And whatever the project/methodology you will be late. As one of my teacher said : "When planing, you have to multiply by PI to get the right timeline,... whatever the number of times you already multiplied by PI". For the moment, I've never seen the Kuonen's law be wrong;-)
For the first part, I dont really think my life is in much danger by going to Rwanda... but we never know...
I dont really know what I'm going to do there. And I dont think planning too much from here (Switzerland) without knowing exactly what is needed there is a good think.
I should be working for the Presbytarian Church of Rwanda, taking care of their current computers (which dont always run well) and trying to teach them how to take care of that once I'm gone... I know, that is not a very precise definition of my job there, but that's all I have...
Hey ! Thanks for the early answers to all of you !
I dont really think I'm going there with zero knowledge... I've traveling a bit already... And I'm pretty sure I'm gonna enjoy the people as much as the country itself;-) The organisation that I'm going with has quite some experience in sending people to the south, so I'm not that much worried about traveling tips.
I am more looking for advice specifically on CS developpement projects. How to bring technology in a usefull way, or what could be the specific needs toward computers, or what kind of kind of stuff I should take with me beside my laptop and a couple of Ubuntu CDs (or any other distro).
It's not that I dont value your opinion on other matters, it just happends that I already have quite a lot of them... and there's probably another view of the situation that I can find only on Slashdot.
Well, i dont know exactly how much RAM Eclipse takes, but I work with Eclipse, MagicDraw and AndroMDA to develop Tomcat based J2EE apps. When everything is loaded (add to that firefox and postgres,...) my 1Go RAM is full and half of my 2Go swap is taken... Still, the system is pretty reactive, except during compilation...
This is one of the largest problems with copyright. It mixes copyright with authorship rights (or as Europeans call them "moral rights"). Authorship rights should remain distinct from copyright. Copyright is a limited privilege. Authorship is perpetual. The author of a document will always be its author, but that author will not always have a monopoly over it. This gives authors the misguided idea that they should have a perpetual monopoly as well.
Again, I am sure of the situation only in Switzerland, but here the copyright is define as part of the authorship rights (the full text can be found in http://www.admin.ch/ch/f/rs/c231_1.html"Loi sur le droit d'auteur", 9 october 1992, RS231.1).
As you said, the right to be the author of a work is perpetual and cannot be sold, given away,... This right goes with a few others (integrity of the work, access to the work,...). All other rights, especially the right to make copies and sell them can be sold to a third party. So, yes, the copyright is part of the authorship right, so what ?
The problem is to know who is the author, to know who can sell the work and make money ? Well, you dont need a registration system here. Simply send a letter to yourself with the work. The date on the stamp will be enough to prove you had that work earlier than anybody else in most cases. If you want a more "secure" solution, go to any usher (google translation, might be wrong), it will cost you about 100$ but you'll be safe.
The problem of the copyright law is that it is abused by everybody. A work has to be a creation, individual,... which should exclude quite a lot of things that are mainly copies of other works, or works in which there is not enough space to make the decisions that would make them clearly creations instead of mere technicall works... But as soon as there is money to make, there will be people abusing the system. (Dont get me started on patent laws...)
Well, at least in Switzerland (and probably the reset of Europe - dont know for the rest of the world), you dont have to register anything or hand a copy of your work or anything. As soon as you have a created a new work, it is copyrighted. That's the good point about copyright : you dont have to actually do anything, or pay anything. Copyright is an "automatic protection". It really should stay that way !
Maximum performance is (almost) never a good mesurement alone. At least not between AMD/Intel. If you want max perf, you would be using a Cray or a Beowulf, or whatever.
If your talking about AMD/Intel, it is already that you need either a generic CPU or a cheap one.
Even for very large projects (let's say NASA, NSA,...) you have some kind of constraint. It can be money, or time to put it in place, or relyability, or whatever.
Whoo ! I'm sorry, but I disagree quite a bit on that one ! First, we can reduce our power consumption without having a worst quality of life. Some ideas (just the easy ones) public transportation ; better efficiency in constructions (you can win both on eating and on climatisation) Second, yes, solar / wind /... are too expensive and not really efficient. But what is it that way ? For the solar, most of the research we have now comes from the NASA. They have almost unlimited budgets (at least if you compare them to civil budgets). So solar is expensive. It was also developped without taking ecology into account. Try to compare the work that is put into research for solar energy or for oil based energy. It is quite amazing ! Most of the solar research is done by academics almost on the personnal budget / time. Ok, the solutions arent there yet, but we need to push in the right direction and see just a step ahead...
There is another quite better approach at reducing CO2 output : consume less energy. Actually, the answer to pollution problems will come more from how much energy we consume (or water,...) than how we produce it. Now what are the good ways to reduce consumption ? I've just read in the "Monde Diplomatique" that 70% of oil conumption in France is from houses and personnal transportations... How can we do better in that area ?
Um you can easily do an hour ban on excessive hits from a given IP. Write a module for Apache that counts the hits from a given IP. If it hits a certain threshold [say > 100 hits a minute or >x KB per second] then it simply adds the ip to a firewall [ipchains, netfilter, etc].
Actually this is used in real life. I had a research project going on this at my university. Actually, what we did was comparing the current packet rate (current = last 5 minutes) to the average packet rate (let's say for the past 3 days). If the current rate is greater than average +- (alpha * std_dev) you have a problem.
If you classify traffic by destination port, then you can detect worm or DDOS attack on a specific port and route the traffic to an analyser. This can give a fairly low false positive rate if you tune alpha and the sliding average well.
We didnt actully use that to disgard traffic, but just to analyse it. As soon as you use an "active" sensor, you have to be carefull about it used against you...
I was in the Swiss Insitute of Technology. It was Ada still 2 years ago. They switched to Java now. And some scala too. Ada was the best thing I could have as a first "academical" language. I wish it stayed that way...
What developed countries are you talking about ? At least for Switzerland (which I would call a developed country) the problem is not solved. The number of infections is going up again. Knowing the danger is not enough... Of course, it is the first step...
Actually, that is the issue... If dont dont have any good Ford dealer around, you'll go for the Toyota. In most case the important difference is the human factor and not the technology. Technology by itself is never the solution...
So : start teaching Linux to everyone and you'll get the needed support in about 5 years...
Hackers are never the problem. Easily exploitable vulnerabilities in a system are.
It depends a lot on what you call hackers... Having people studying the security of the system is a good thing. Having them warn the manufacturer / buisness / developper /... is a good thing. Having them release exploit in the wild can be usefull in certain conditions to make things move if the problem is not solved.
Having them sell a service on that is quite something different !
Slashdot Mirror
Thanks for the answers so far. As I wrote in the article, the problem is not so much the hosting of video files than the front end to put on. We have to integrate it with the NGO's website, put some informations on the videos ... The only suggestion in that direction sofar is getDemocracy, which I dont find really ready for production.
Anyway, I'll keep looking, and thanks for the help !
Soviet america ... anyone ?
Well, the problem was to know the question ... I already had the answer ... ...
Well, it took million of years to compute it, so it's probably normal that Google cant get it in a few seconds
Wow, I almost read the result was 42, which would explain a lot ... ...
I still dont know the great question, and it doesnt seem that Google will help me to find it
Good code comes from good programmers. Whatever the methodology, you wont get a clean design from somebody who doesnt have experience. I think quite a few managers forget that. If you hire a guy just out of a technical school because you can pay him only 50K/year than dont wonder if you get dirty code. ...)
The book talk quite a bit about the importance to have competent people at the right place (seems like an obvious advice, but
Not sure documenting early will save much ... In the project I'm working for the moment, there has been about 200 pages of docs before we had a single line of code. Every aspect they were thinking is there. But guess what ... after 3 months of developpement, only about 10% of that docs is still valid. ... whatever the number of times you already multiplied by PI". For the moment, I've never seen the Kuonen's law be wrong ;-)
One of the issue is to know when to document and when to keep it an informal discussion. An out of sync documentation is probably worse than no docs.
And as all things, the best aproach depends of the project. In an area where the requierments change fast (like my project) spending too much time documenting them is probably waste. If the requierments are stable for a couple of years than why not.
And whatever the project/methodology you will be late. As one of my teacher said : "When planing, you have to multiply by PI to get the right timeline,
If anyone has mod points, mod the parent funny ! And thanks for the good laugh before going to bed ...
I've never met Mrs Noether personnaly, that might be why I have some problems with this concept ...
To understand recursion, we must first understand recursion ...
For the first part, I dont really think my life is in much danger by going to Rwanda ... but we never know ...
... I know, that is not a very precise definition of my job there, but that's all I have ...
I dont really know what I'm going to do there. And I dont think planning too much from here (Switzerland) without knowing exactly what is needed there is a good think.
I should be working for the Presbytarian Church of Rwanda, taking care of their current computers (which dont always run well) and trying to teach them how to take care of that once I'm gone
Hey ! Thanks for the early answers to all of you !
... I've traveling a bit already ... And I'm pretty sure I'm gonna enjoy the people as much as the country itself ;-) The organisation that I'm going with has quite some experience in sending people to the south, so I'm not that much worried about traveling tips.
... and there's probably another view of the situation that I can find only on Slashdot.
I dont really think I'm going there with zero knowledge
I am more looking for advice specifically on CS developpement projects. How to bring technology in a usefull way, or what could be the specific needs toward computers, or what kind of kind of stuff I should take with me beside my laptop and a couple of Ubuntu CDs (or any other distro).
It's not that I dont value your opinion on other matters, it just happends that I already have quite a lot of them
There is even a web server in postscript !
Well, i dont know exactly how much RAM Eclipse takes, but I work with Eclipse, MagicDraw and AndroMDA to develop Tomcat based J2EE apps. When everything is loaded (add to that firefox and postgres, ...) my 1Go RAM is full and half of my 2Go swap is taken ... ...
Still, the system is pretty reactive, except during compilation
Again, I am sure of the situation only in Switzerland, but here the copyright is define as part of the authorship rights (the full text can be found in http://www.admin.ch/ch/f/rs/c231_1.html"Loi sur le droit d'auteur", 9 october 1992, RS231.1).
As you said, the right to be the author of a work is perpetual and cannot be sold, given away,
The problem is to know who is the author, to know who can sell the work and make money ? Well, you dont need a registration system here. Simply send a letter to yourself with the work. The date on the stamp will be enough to prove you had that work earlier than anybody else in most cases. If you want a more "secure" solution, go to any usher (google translation, might be wrong), it will cost you about 100$ but you'll be safe.
The problem of the copyright law is that it is abused by everybody. A work has to be a creation, individual,
Well, at least in Switzerland (and probably the reset of Europe - dont know for the rest of the world), you dont have to register anything or hand a copy of your work or anything. As soon as you have a created a new work, it is copyrighted. That's the good point about copyright : you dont have to actually do anything, or pay anything. Copyright is an "automatic protection". It really should stay that way !
Maximum performance is (almost) never a good mesurement alone. At least not between AMD/Intel. If you want max perf, you would be using a Cray or a Beowulf, or whatever.
...) you have some kind of constraint. It can be money, or time to put it in place, or relyability, or whatever.
...
If your talking about AMD/Intel, it is already that you need either a generic CPU or a cheap one.
Even for very large projects (let's say NASA, NSA,
You can never talk about performance alone
The question is not if it perform better for the same clock speed. The real question is to compare performance/price or performance/power consumption.
Whoo ! I'm sorry, but I disagree quite a bit on that one ! ... are too expensive and not really efficient. But what is it that way ? For the solar, most of the research we have now comes from the NASA. They have almost unlimited budgets (at least if you compare them to civil budgets). So solar is expensive. It was also developped without taking ecology into account. Try to compare the work that is put into research for solar energy or for oil based energy. It is quite amazing ! Most of the solar research is done by academics almost on the personnal budget / time. ...
First, we can reduce our power consumption without having a worst quality of life. Some ideas (just the easy ones) public transportation ; better efficiency in constructions (you can win both on eating and on climatisation)
Second, yes, solar / wind /
Ok, the solutions arent there yet, but we need to push in the right direction and see just a step ahead
There is another quite better approach at reducing CO2 output : consume less energy. Actually, the answer to pollution problems will come more from how much energy we consume (or water, ...) than how we produce it. ... How can we do better in that area ?
Now what are the good ways to reduce consumption ? I've just read in the "Monde Diplomatique" that 70% of oil conumption in France is from houses and personnal transportations
Actually this is used in real life. I had a research project going on this at my university. Actually, what we did was comparing the current packet rate (current = last 5 minutes) to the average packet rate (let's say for the past 3 days). If the current rate is greater than average +- (alpha * std_dev) you have a problem.
If you classify traffic by destination port, then you can detect worm or DDOS attack on a specific port and route the traffic to an analyser. This can give a fairly low false positive rate if you tune alpha and the sliding average well.
We didnt actully use that to disgard traffic, but just to analyse it. As soon as you use an "active" sensor, you have to be carefull about it used against you
I was in the Swiss Insitute of Technology. It was Ada still 2 years ago. They switched to Java now. And some scala too. ...
Ada was the best thing I could have as a first "academical" language. I wish it stayed that way
But is the Dalaï Lama reasonable ? I wonder
What developed countries are you talking about ? At least for Switzerland (which I would call a developed country) the problem is not solved. The number of infections is going up again. Knowing the danger is not enough ... Of course, it is the first step ...
Actually, that is the issue ... If dont dont have any good Ford dealer around, you'll go for the Toyota. In most case the important difference is the human factor and not the technology. Technology by itself is never the solution ...
So : start teaching Linux to everyone and you'll get the needed support in about 5 years ...
It depends a lot on what you call hackers
Having them sell a service on that is quite something different !