How is it that without a lawyer there is no way you can possibly win a case? I find that slightly disturbing.
I would *hope* that this would depend on actual facts, evidence, witnesses, etc.. not on who has the fattest purse or who can follow the exact correct "protocol" in court.
Let's say I couldn't possibly have committed the crime because I never had internet or own a PC for that matter. I sure hope I wouldn't need a lawyer to "prove" that. I would hope that the amount of lawyers one side has would not have a huge impact on the outcome (in fact, I would hope that it would only affect a very small percentage of cases) -- unless courts these days simply favor the side that has followed the correct procedures and provided the highest stack of paperwork, which would make them corrupt in my eyes as they are not really interested in the actual truth.
I bet the software is the one responsible for making the coordinates "fuzzy" by mapping a set of pretty accurate coordinates onto some predefined locations. In other words, this should be hackable to give you accurate coordinates...
Re:Java: Where Components come from
on
Java Is So 90s
·
· Score: 1
I'm a Python and Java guy, although more Java than Python. I find Python to be elegant in some ways (built-in collection types) and totally brain dead in others (classes and the 'self' parameter, lack of compile time code checks). I use Python for quick hacks, shell scripts and minor GUI work (major GUI work is an exercise in frustration in Python). I use Java for anything that would take more than 1000 lines of codes in Python because it is easier to maintain over the long run.
Why anyone would even consider using a language like Python or PHP or any other scripting language out there for larger projects is beyond me (small websites are fine, as those things are usually quick hacks anyway). Even in the smallest Python programs I often find trivial mistakes that can only be discoverd if the code is actual run. For code that is only rarely used (like exception handling code) you'll often only find that error when things actually go wrong. In Java I can atleast be sure I didn't misspell a variable or method name by accident, or that I used the wrong string concatenation operator or whatever else the Python code checker won't catch at "compile time".
As for Eclipse, you are welcome to program Java in the same editor I program Python in: VIM. Not that I would though; any decent Java Editor will offer you a host of handy functions, which are only possible with Java (and perhaps C#) because Java is a strict language without macros (#define) and other nonsense that prevents in-depth parsing. A good Java editor can tell you when variables are never read, what code calls a public method (if any at all), when you misspell a variable or when you write other nonsense. A good Java editor can create a full parse tree of your code and even rewrite parts of it to do exactly the same thing (like inverting a conditional, rewriting if-else blocks, inlining methods or globally renaming anything you want (files, methods, variables, you name it).
No Python editor could ever hope to do that, hence no Python editor will ever be a heavy weight environment like Eclipse; there just isn't much more a Python editor can offer than decent syntax highlighting and meta functions like diffing and folding.
Without somekind of external server, no schemes that check my computers time are going to provide any real protection. And I'm surprised schemes like this donot yet REQUIRE internet access (but I'm sure they soon will).
I can put the file on my computer, make a Ghost image of it, then play it. When I want it again, I restore the Ghost image, and the file will work.
Alternatively, I could figure out where the DRM program stores its data and restore backups of that.
I could set up an automatic virtual environment, just to play a single file, which would be recreated each time I wanted to play it.
Or I could simply pipe the audio somewhere else, and remove all the restrictions permanently.
I still used Eclipse though, because the milestone versions for 3.1 had full Java 5 support about 6 months ago already, and they were really quite stable. Only the advanced features like specific warnings and refactorings took a little bit longer to get added, but I was quite happy even without those:)
Considering that SWT makes Eclipse behave like a native Windows application, instead of the sluggish feel of Swing, I'd say there was a need.
I always disliked the sluggishness of Swing, and the discrepancies in the look with REAL Windows apps. Things like menu's popping up slowly, sluggish tree controls.. everything just looked slightly off, and felt like I turned off graphics acceleration.
Nowadays, Swing is somewhat improved, but there certainly was and still is a need for SWT.
Considering that the last stable version of Eclipse, 3.0, didn't yet support Java 5 and all the new stuff in it (generics, autoboxing, enhanced for construct, enumerations), I'd say that Java 5.0 support is a pretty good indication what Eclipse 3.1 can do for you now.
This means it can refactor code using Java 5 specifics, it has specific warnings/errors for Java 5 code, quick fixes, code formatting/highlighting, and so on.
Is java suffering from the same problems that every other previous attempt to make an interpreted language work suffered from? Is it just too slow?
No, because the "bytecode" is not actually interpreted all the time, it is in fact interpreted and then compiled into native code and re-used from then on. The code thus produced is specific to your platform, so it can produce code specifically for your Pentium 4 (or whatever you are using) while normal binaries for your platform are often compiled for 80386 (for example).
The price you pay is a slightly increased startup time when new code gets loaded because it still needs to finish compilation.
Remember when Java was once a viable "plugin" for your browser? When was the last time you used it?
Today I think. My bank uses it for secure authentication.
Remember when Java was supposed to be next foundation for desk top applications? Where are they?
I don't ever remember hearing that, I believe Java was to be the next foundation for server side applications. It certainly delivered on that promise.
However, there are a few desktop applications written in Java that are very useable and very nice. Azureus and Eclipse are excellent examples. Most people don't even realize they're written in Java because they look and feel just like normal Windows applications thanks to SWT.
Java's not the new C++ , it's the new Pascal.
I think Java is neither. The programming model is a lot less complex than either of those, yet very powerful. C++/Pascal programmers often have problems letting go of the concept of pointers and have trouble accomplishing things in Java without them. Takes a bit of getting used to.
Funny you should mention UltraEdit:) Just 3 weeks ago I converted a good friend of mine from UltraEdit to Eclipse. After I showed him how he could develop Tomcat applications with HotCode replacement, and debug them live inside Eclipse, he wonders how he ever did without.
He's just messaged me asking if I already managed to download Eclipse 3.1 -- I didn't even know it was out yet:)
Eclipse eats as much as you give it. I think it defaults to 256 MB ram, and given the chance it will fill that with cached lists of classes, methods, and so on in order to speed up search and overview functions.
Well, I can tell you that all the CD's I've written are no longer error free. And that was after 5 years, while stored in a closed box. They were still PARTIALLY readable, but the outer tracks were all damaged beyond recovery.
These discs (about 100 of them) were put in that box and stored away immediately after writing AND verifying them. When I read them 5 years later (after an important HD crashed) some of the discs had noticable defects (circular areas on the disc which had different colors)
Fast forward 10 years to the DVD era. I've written 15 of those 2 years ago. They are already showing signs of degrading (they can still be read, but the outer tracks are read slowly, as if a lot of error correction is happening).
One thing I've noticed however: rewritable CD's seem to last a lot longer. Even after 5 years I can read those without problem.
If Moore's law manages to hold another 100 years, computers will be fast enough to calculate all moves in a chess game in about an hour orso.
However, that wouldn't mean they would automatically win. It's more likely that they could always force a draw, and only win if the opponent makes a mistake.
Of course, Moore's law is highly unlikely to last for another century, as it is already showing signs of breaking down.
Why should you stop providing your "own" solution ? You didn't COPY it ! You didnt even know it existed !
Granted this is a problem, but it is also the reason why penalties are smaller in this situation than when someone knowningly infringes.
It's not A problem, it's THE problem. The software industry consists of many hundreds of thousands small companies and millions of individuals, unlike any other industry which makes use of patents.
The software they produce often solves very similar problems, like setting up a website for some client, or building a customer management system, or converting data from one format to another and so on. A lot of these processes do something unique, as they handle only certain data, or provide a certain look and feel, but are complex enough to be covered by dozens of patents.
If patents are allowed to go through, a lot of these programs could no longer be created. It simple would be too expensive (checking for patents, defending against patents) or too risky (getting sued because you inadvertently stepped on someone's patent).
The problem is that the software industry is huge, and every day, two programmers think of the same things on different continents... if one of them patents their idea, the other cannot do his/her job anymore, or worse, becomes a criminal.
Patents simply donot scale, the problem gets worse with more people involved. If we only had 1000 programmers, patenting would make sense as it would encourage some of those people to think of novell ideas since it would make them some money.
Patents work fine for "smaller" industries, where there simply is a lot LESS innovation going on, and where innovation is far more expensive to achieve. It doesn't work for software, where programmers are being innovative almost every day (at little costs) just to comply with a customers demands.
Why should you stop providing your "own" solution ? You didn't COPY it ! You didnt even know it existed !
Granted this is a problem, but it is also the reason why penalties are smaller in this situation than when someone knowningly infringes.
It's not A problem, it's THE problem. The software industry consists of many hundreds of thousands small companies and millions of individuals, unlike any other industry which makes use of patents.
The software they produce often solves very similar problems, like setting up a website for some client, or building a customer management system, or converting data from one format to another and so on. A lot of these processes do something unique, as they handle only certain data, or provide a certain look and feel, but are complex enough to be covered by dozens of patents.
If patents are allowed to go through, a lot of these programs could no longer be created. It simple would be too expensive (checking for patents, defending against patents) or too risky (getting sued because you inadvertently stepped on someone's patent).
The problem is that the software industry is huge, and every day, two programmers think of the same things on different continents... if one of them patents their idea, the other cannot do his/her job anymore, or worse, becomes a criminal.
Patents simply donot scale, the problem gets worse with more people involved. If we only had 1000 programmers, patenting would make sense as it would encourage some of those people to think of novell ideas since it would make them some money.
Patents work fine for "smaller" industries, where there simply is a lot LESS innovation going on, and where innovation is far more expensive to achieve. It doesn't work for software, where programmers are being innovative almost every day (at little costs) just to comply with a customers demands.
I don't see how Mac could be a threat. It seem they don't want it to run on "standard" Windows hardware, so it's not like people can give it try as easy as they can try Linux.
I don't know what Microsoft is doing with their protocol, admittedly, and I haven't read the article either. My idea was more in response to the parent.
However, it is hard for me to see how this could possibly work. Peers could only generate "new blocks" if they have the complete file (or enough blocks to reconstruct it). A partial file would be useless, so only seeds would be able to do this.
That's however based on how PAR2 works, which can only compute recovery blocks based on the complete original input.
If you make the data you wish to share twice as large by adding 50% PAR2 blocks you are doing exactly the same thing.
Any n blocks would do, as long as n > 50%. Some of those will be "normal" data blocks, and some will be the blocks calculated with PAR2 -- it doesn't matter, as PAR2 can calculate the rest as soon as you have 50% of the blocks, where 50% would equal the ORIGINAL filesize, before adding the PAR2 blocks.
As a matter a fact, I haven't. I use unpatched Windows boxes; furthermore I donot use a firewall or virusscanner under Windows. Although I do admit to running Ad-Aware every now and then to reassure myself no crap is installed.
The trick is to put your box behind a Linux firewall, and refusing to use crap like IE and Outlook. Furthermore, I donot click on everything I see in my e-mail.
The only time I ever had something odd on my box was somekind of program trying to make connection to some server... its fatal flaw was that it aroused my suspicion by scanning the system32 folder every second and seeing if some special files were added to it. I noticed this when I ran filemon, tracked it down and killed it. It more than likely was installed because I downloaded software using BitTorrent -- something I haven't done since. It seems some release groups don't mind adding spyware/trojans to their stuff.
I don't see why you would need more data (atleast not a significant amount). Let's say I have a large file of 600 MB. I run PAR2 over it and tell it to create 600 MB worth of PAR2 blocks using a chunksize exactly equal to what I will use when creating the Torrent (let's say 1 MB).
I now fire up BitTorrent and share this torrent.
Someone manages to download 50% this file (ie, that person has 600 completed chunks of 1 MB). S/he can now stop BitTorrent, and run PAR2 over the data to reconstruct the rest.
In what way was there more data transferred?
A few more checksums in the.torrent file I admit, but that's all...
Yes there is. When you are downloading the last segment you need (#500), then in the current BitTorrent implementation that segment will only be available from a smaller subset of peers (which makes sense, since the more commonly available segments are far more likely to have been downloaded already).
However, if you have 499 segments out of 1500, of which 1000 are PAR2 segments, then you have a pool of 1001 segments to choose from -- you don't care which one specifically, any of them will do. It's likely that almost all of the peers will have one of those segments you donot have yet, and you just need one to reconstruct the rest.
This would be totally trivial to add to BitTorrent right now. When you create your torrent, create it from the original file, but create it from a file that's twice as large (50% real data, the other 50% PAR2 blocks) -- trivial to make this automatic when creating Torrents.
As soon as you downloaded 50%, you can run a tool to recover the other 50%.
All it would require is PAR2 integration, in say, Azureus, and a way of detecting whether the file is a file with PAR2 blocks (maybe there's room for some extra information in the protocol to mark such files).
The way I see it this would improve BitTorrent in two important ways:
1) As mentioned, BitTorrent wouldn't stall near the 'end' of the file, since "the end" is now at 50% -- the point where you can recover the rest.
2) Because there are more blocks, there will be a greater diversity of blocks spread over the peers. Since you only need 50% of the blocks to reconstruct the file, there's less chance of running into torrents where all peers have 99.9% of the blocks and there's no seeds -- someone among those is likely to have a PAR2 block you don't have yet, and then everyone can reconstruct the original data.
It would be funny if someone added this to Azureus or some other client and have it out before Microsoft:)
[quote]I don't know what I'm going to do if they stop supporting 2000[/quote]
Why not just keep on using it? It's not like Win2k will stop working simply because M$ doesn't support it anymore. In fact, I cannot remember ever needing their support, so I'm sure I can do without it.
Slashdot Mirror
I would *hope* that this would depend on actual facts, evidence, witnesses, etc.. not on who has the fattest purse or who can follow the exact correct "protocol" in court.
Let's say I couldn't possibly have committed the crime because I never had internet or own a PC for that matter. I sure hope I wouldn't need a lawyer to "prove" that. I would hope that the amount of lawyers one side has would not have a huge impact on the outcome (in fact, I would hope that it would only affect a very small percentage of cases) -- unless courts these days simply favor the side that has followed the correct procedures and provided the highest stack of paperwork, which would make them corrupt in my eyes as they are not really interested in the actual truth.
I bet the software is the one responsible for making the coordinates "fuzzy" by mapping a set of pretty accurate coordinates onto some predefined locations. In other words, this should be hackable to give you accurate coordinates...
Why anyone would even consider using a language like Python or PHP or any other scripting language out there for larger projects is beyond me (small websites are fine, as those things are usually quick hacks anyway). Even in the smallest Python programs I often find trivial mistakes that can only be discoverd if the code is actual run. For code that is only rarely used (like exception handling code) you'll often only find that error when things actually go wrong. In Java I can atleast be sure I didn't misspell a variable or method name by accident, or that I used the wrong string concatenation operator or whatever else the Python code checker won't catch at "compile time".
As for Eclipse, you are welcome to program Java in the same editor I program Python in: VIM. Not that I would though; any decent Java Editor will offer you a host of handy functions, which are only possible with Java (and perhaps C#) because Java is a strict language without macros (#define) and other nonsense that prevents in-depth parsing. A good Java editor can tell you when variables are never read, what code calls a public method (if any at all), when you misspell a variable or when you write other nonsense. A good Java editor can create a full parse tree of your code and even rewrite parts of it to do exactly the same thing (like inverting a conditional, rewriting if-else blocks, inlining methods or globally renaming anything you want (files, methods, variables, you name it).
No Python editor could ever hope to do that, hence no Python editor will ever be a heavy weight environment like Eclipse; there just isn't much more a Python editor can offer than decent syntax highlighting and meta functions like diffing and folding.
There's also music I bought instead of downloading, didn't like, but did pay for.
Usually about half the tracks of the album they came on.
In the end, we're even.
I can put the file on my computer, make a Ghost image of it, then play it. When I want it again, I restore the Ghost image, and the file will work.
Alternatively, I could figure out where the DRM program stores its data and restore backups of that.
I could set up an automatic virtual environment, just to play a single file, which would be recreated each time I wanted to play it.
Or I could simply pipe the audio somewhere else, and remove all the restrictions permanently.
I still used Eclipse though, because the milestone versions for 3.1 had full Java 5 support about 6 months ago already, and they were really quite stable. Only the advanced features like specific warnings and refactorings took a little bit longer to get added, but I was quite happy even without those :)
I always disliked the sluggishness of Swing, and the discrepancies in the look with REAL Windows apps. Things like menu's popping up slowly, sluggish tree controls.. everything just looked slightly off, and felt like I turned off graphics acceleration.
Nowadays, Swing is somewhat improved, but there certainly was and still is a need for SWT.
This means it can refactor code using Java 5 specifics, it has specific warnings/errors for Java 5 code, quick fixes, code formatting/highlighting, and so on.
The price you pay is a slightly increased startup time when new code gets loaded because it still needs to finish compilation.
Today I think. My bank uses it for secure authentication. I don't ever remember hearing that, I believe Java was to be the next foundation for server side applications. It certainly delivered on that promise.However, there are a few desktop applications written in Java that are very useable and very nice. Azureus and Eclipse are excellent examples. Most people don't even realize they're written in Java because they look and feel just like normal Windows applications thanks to SWT.
I think Java is neither. The programming model is a lot less complex than either of those, yet very powerful. C++/Pascal programmers often have problems letting go of the concept of pointers and have trouble accomplishing things in Java without them. Takes a bit of getting used to.He's just messaged me asking if I already managed to download Eclipse 3.1 -- I didn't even know it was out yet :)
Try starting it with -vmargs -Xmx128M
These discs (about 100 of them) were put in that box and stored away immediately after writing AND verifying them. When I read them 5 years later (after an important HD crashed) some of the discs had noticable defects (circular areas on the disc which had different colors)
Fast forward 10 years to the DVD era. I've written 15 of those 2 years ago. They are already showing signs of degrading (they can still be read, but the outer tracks are read slowly, as if a lot of error correction is happening).
One thing I've noticed however: rewritable CD's seem to last a lot longer. Even after 5 years I can read those without problem.
However, that wouldn't mean they would automatically win. It's more likely that they could always force a draw, and only win if the opponent makes a mistake.
Of course, Moore's law is highly unlikely to last for another century, as it is already showing signs of breaking down.
Do it in bytes, that's even more impressive. I have about 2598455214080 bytes of storage in my Linux box, all stuffed in a midi tower (8 drives total)
Shame, it should have been upheld. Then we might get to see how stupid such patents really are.
The software they produce often solves very similar problems, like setting up a website for some client, or building a customer management system, or converting data from one format to another and so on. A lot of these processes do something unique, as they handle only certain data, or provide a certain look and feel, but are complex enough to be covered by dozens of patents.
If patents are allowed to go through, a lot of these programs could no longer be created. It simple would be too expensive (checking for patents, defending against patents) or too risky (getting sued because you inadvertently stepped on someone's patent).
The problem is that the software industry is huge, and every day, two programmers think of the same things on different continents... if one of them patents their idea, the other cannot do his/her job anymore, or worse, becomes a criminal.
Patents simply donot scale, the problem gets worse with more people involved. If we only had 1000 programmers, patenting would make sense as it would encourage some of those people to think of novell ideas since it would make them some money.
Patents work fine for "smaller" industries, where there simply is a lot LESS innovation going on, and where innovation is far more expensive to achieve. It doesn't work for software, where programmers are being innovative almost every day (at little costs) just to comply with a customers demands.
The software they produce often solves very similar problems, like setting up a website for some client, or building a customer management system, or converting data from one format to another and so on. A lot of these processes do something unique, as they handle only certain data, or provide a certain look and feel, but are complex enough to be covered by dozens of patents.
If patents are allowed to go through, a lot of these programs could no longer be created. It simple would be too expensive (checking for patents, defending against patents) or too risky (getting sued because you inadvertently stepped on someone's patent).
The problem is that the software industry is huge, and every day, two programmers think of the same things on different continents... if one of them patents their idea, the other cannot do his/her job anymore, or worse, becomes a criminal.
Patents simply donot scale, the problem gets worse with more people involved. If we only had 1000 programmers, patenting would make sense as it would encourage some of those people to think of novell ideas since it would make them some money.
Patents work fine for "smaller" industries, where there simply is a lot LESS innovation going on, and where innovation is far more expensive to achieve. It doesn't work for software, where programmers are being innovative almost every day (at little costs) just to comply with a customers demands.
I don't see how Mac could be a threat. It seem they don't want it to run on "standard" Windows hardware, so it's not like people can give it try as easy as they can try Linux.
However, it is hard for me to see how this could possibly work. Peers could only generate "new blocks" if they have the complete file (or enough blocks to reconstruct it). A partial file would be useless, so only seeds would be able to do this.
That's however based on how PAR2 works, which can only compute recovery blocks based on the complete original input.
If you make the data you wish to share twice as large by adding 50% PAR2 blocks you are doing exactly the same thing.
Any n blocks would do, as long as n > 50%. Some of those will be "normal" data blocks, and some will be the blocks calculated with PAR2 -- it doesn't matter, as PAR2 can calculate the rest as soon as you have 50% of the blocks, where 50% would equal the ORIGINAL filesize, before adding the PAR2 blocks.
The trick is to put your box behind a Linux firewall, and refusing to use crap like IE and Outlook. Furthermore, I donot click on everything I see in my e-mail.
The only time I ever had something odd on my box was somekind of program trying to make connection to some server... its fatal flaw was that it aroused my suspicion by scanning the system32 folder every second and seeing if some special files were added to it. I noticed this when I ran filemon, tracked it down and killed it. It more than likely was installed because I downloaded software using BitTorrent -- something I haven't done since. It seems some release groups don't mind adding spyware/trojans to their stuff.
I now fire up BitTorrent and share this torrent.
Someone manages to download 50% this file (ie, that person has 600 completed chunks of 1 MB). S/he can now stop BitTorrent, and run PAR2 over the data to reconstruct the rest.
In what way was there more data transferred?
A few more checksums in the .torrent file I admit, but that's all...
However, if you have 499 segments out of 1500, of which 1000 are PAR2 segments, then you have a pool of 1001 segments to choose from -- you don't care which one specifically, any of them will do. It's likely that almost all of the peers will have one of those segments you donot have yet, and you just need one to reconstruct the rest.
As soon as you downloaded 50%, you can run a tool to recover the other 50%.
All it would require is PAR2 integration, in say, Azureus, and a way of detecting whether the file is a file with PAR2 blocks (maybe there's room for some extra information in the protocol to mark such files).
The way I see it this would improve BitTorrent in two important ways:
1) As mentioned, BitTorrent wouldn't stall near the 'end' of the file, since "the end" is now at 50% -- the point where you can recover the rest.
2) Because there are more blocks, there will be a greater diversity of blocks spread over the peers. Since you only need 50% of the blocks to reconstruct the file, there's less chance of running into torrents where all peers have 99.9% of the blocks and there's no seeds -- someone among those is likely to have a PAR2 block you don't have yet, and then everyone can reconstruct the original data.
It would be funny if someone added this to Azureus or some other client and have it out before Microsoft :)
[quote]I don't know what I'm going to do if they stop supporting 2000[/quote] Why not just keep on using it? It's not like Win2k will stop working simply because M$ doesn't support it anymore. In fact, I cannot remember ever needing their support, so I'm sure I can do without it.