To make this exceptionally clear, RMS intended to make copyright an irrelevance by creating a bounty of free software available under the GPL.
Ok, I see where you are coming from now.
You've read me all wrong. I love the GPL. I'm not adverse to it at all. I have said nothing negative about it. Where did you get the crazy idea that I have?
Your comments had a ring of Balmer to them i.e. Balmer talks about how the GPL is a virus, and you were talking about how it destroyed copyright. The part what wasn't shining through was it destroyed it in a positive way - or to put it another way; actually provided *copy* *rights*; the right to copy the software.
PS: The GPL isn't a copyright, it's a license. Be careful when trying to explain something that you don't fully understand. I'll argue you three ways from Sunday.
In a literal sense you are right here, but since the license is used to grant rights that are not actually granted under standard copyright laws one can also think of it as a license that is used to provide copy rights.
The argument you have made is well researched but I fail to see how it makes your point. You are claiming that the GPL destroys copyright and in order for something to be destroyed it must be created first. By reasoning this means your claim is that the GPL is able to remove copyright from a product that is already protected by copyright.
Can you name one product which has been forced by the GPL to become GPL code also? Even Blender which was commercial software became GPL only because the community banded together and paid a hostage release fee for it. The company was under no duress to GPL the code, and both they and the community have now benefitted greatly from the codes release.
RMS may be a fool for wanting a world completely without copyright, but that doesn't invalidate the value of the GPL license. The GPL provides people like me (a professional full-time employed programmer) the chance to provide a gift to the community that can never be taken away.
Now, I don't expect that all software should be GPL'ed because we need different licenses to cover different needs. That's why each company is free to write it's own restrictive license, or to go with any of the popular Open Source licenses.
You forget that the GPL was designed to destroy software copyright.
The GPL is a copyright. It's not designed to destroy copyright but to enhance the rights of the people who use the software.
If I place a copy of Linux on the nightstand with a copy of Windows they will both still be there in the morning. I'm not likely to find the Windows box with it's flaps open and it's floppy hanging out.
Even the viral nature of the GPL is overated. There is nothing that compels you to copy GPL code into your software. If you don't use GPL then you don't need to GPL your software. If you do benefit from using GPL code, then all that is asked is that you enable others to benefit in the same way you did. Not a big ask really. Also, there is nothing in the GPL that says you can't sell the software and profit from that sale. You can, but you must also allow people to obtain it and the source at reasonable cost to distribute. RedHat has built it's business on this.
Since when is the whole Open Source movement able to be judged by the opinion of one man, even if he is mainly responsible for kicking it all off? Open Source *is* choice, and the people who use it have taken a harder choice and a longer path because of their firm belief in the values of Open Source software.
Also, try not confuse the Open Source movement with anti-capitolism, communism, or anti-gloalisation. Many of those people don't even have a computer let alone have an opinion on what OS to run, and many of us OS people don't have an opinion on their idealogies.
On a personal note, I really love Open Source, am lukewarm on the globalisation thing (why are there four fucking Starbucks on my street?? What happened to the independant cafes?), and am a firm supporter of basic tenets of capitolism, but tempered with socialist values. I'm mostly white, working class, an adult, and do not *force* my choice down anyones throats but will happily speak when asked.
Now, time for you to take your pills and watch some TV, you're a little excited today.
Exactly which marketplace has it failed in, Spitalfield markets...the Grand Bazaar? You need to be paying a little more attention to what's happening right now. Linux is growing in importance in the mindset of business and government. It seems nearly every week I read of another increase in mindshare and the FUD from the naysayers is not going to make it go away. A few weeks ago Oracle mentioned that it is recommending Linux for 60% of all new installations - and that is only one indicator or how Linux is growing.
Desktop acceptance may be on the horizon still, but it is only a question of time, since Linux is powering along in the server market.
There's every chance that a late version of Linux like RedHat 9.0 using tar could read those tapes and dump the files for you even now so you wouldn't even need to install an older OS.
One of the better reasons for switching to Open Source is the support of Open Standards. MS could take away a lot of complain about their office apps if they would open the standards and submit these to some standards bodies. As it is there is a huge danger of your data ending up in a data ghetto. A data ghetto is where your data becomes trapped in a single place and is unable to be sucessfully used or moved. MS Word is a prime example of data caught in a data ghetto. Imagine the hundred of millions of word processing documents a goverment must have. Now consider the governments need for openness with it's governed populace. Each and every member of the public should be given oportunity to read the government documents (unclassified ones) and should not be prevented from doing so by financial constraints.
There is a lock-in that occurs by having all this data in a format that is difficult to cleanly translate to other word processing/viewer formats. I understand that Open Office can open Word documents, but not all the formatting is cleanly cut across, and so in our business we cannot rely on this. The problem compounds as time drifts onwards. Older documents are held in older formats e.g. Word 2.0, WordPerfect 6.0. At some time it is likely that support for reading those formats will be dropped in future versions of the office suites. At that time it will be very difficult to access that data.
Open Source helps protect against this by ensuring the code to read old data (which government hordes by the kiloton) will always be available.
The Brazillian government is still able to choose and purchase MS Office...just as soon as MS Open Source it;-) That's the choice being offered to Bill. It's also important to remember that Open Source doesn't mean only one choice. There are numerous browsers, email clients, desktops and office productiivty suites to choose from. All the government is doing is saying that the software should be Open Source.
Open Source does not necessarily mean free. MS could Open Source their products and still sell them. If IBM can do it, why can't they?
10,000,000 of the hits are to an ASP page, and all our ASP pages use at least 3 components to display. Many use more components. All hook up to a back end database. At peak times we are getting 14-16 page impressions / second, which is all handled by a quad xeon processor with a dual xeon for the database server. The web server is now running at capacity, so if demand keeps increasing we will need to cluster.
This is quite an interesting point. It was games that got me hooked on PCs back in the eighties. Some games like Quake and Neverwinter Nights come with very useable scripting languages that let you either add new weapons (and stuff) or script up your own adventures. They are simplified C and thus useful as a stepping stone to "real" programming and have that instant gratification kids are looking for. Yeh, get your kids into Quake C and NWN C!
Visual Basic is still around, but I don't know of any serious programmers who really use it hard core
We use it every day here at work and we're serious:-) We use it mainly because it's fantastic for producing middle tier COM components quickly and easily. The site I develop for takes 100,000,000 hits / month, so I'd say VB is serious. Personally though, perhaps PHP, Python or some other langauge might be better. And definitely buy a small cheap machine that the kids can trash whenever they like. They'll learn better when it's their machine and they know they can destroy it.
Would you let a 12 year old child replace the alternator in your car?
It was compulsory in our family when young to help dad repair cars. So yes, he did get me replacing alternators (relatively easy) or striping parts, cleaning and then re-assembling.
and yet you'd let your 12 year old code in Perl? Aw, c'mon, show the poor little tyke some mercy and at least let him start with a structured and easy to learn language first.
Dangerous trick without some sort of accounting and quotas on the hard drive. Locked myself out of my VAX account one day with a similar trick, albeit with a Pascal program instead. I suggest you try:
By default all rights to written works (software included) default to the author. In Australia at least, all you need to do to make software copyrighted is to place a small label on the packaging and again when the software starts up. Easy. Placing said notes in code would be sufficient for it to be considered copyrighted, even if the infringing party claims they haven't read that notice, they are still held to it.
In the meantime, the copyright holders of the Linux code can request that SCO's product be removed from sale and SCO issue a recall on all the sold product until a case is heard. Now that *would* be interesting.
That said, though, I'd like to retain "compatibility" with Mutt if at all possible, so that I can use Mutt to read new or old mail without interfering with the operation of Evolution.
Just keep your mail on your mail server and use IMAP to connect to it. Then you can use Mutt, Evolution, KMail, Netscape/Mozilla or even Outlook ot get at your mail. Best of all, you can access it while you are at work too;-) I presonally use Kmail when I'm at home, Mozilla if I have to boot Windows for a bit, Pine when I shell in, and Mozilla from work (since Outlook acted all pissy about it for no good reason). IMAP means enjoying your mail no matter where you are.
You can also move all your contacts and stuff into an LDAP server for bonus points. Now I have one address book which I can use wherever I am with all my email clients on any OS. Only downside, I have yet to find a decent client to update that LDAP address book...anyone got some recomendations?
Right. And a C function is just assembler instructions that are inserted in place of the function.
It's not quite that simple since the compiler has many oportunities to apply optimisations.
Exactly. So a disassembler can't recreate them. Just like comments.A disassembler will provide equivalent labels, but it will not give them the same name as they originally had. They end up being called (for example) LABEL01, LABEL02, but they are still doing the same job, that it providing a mnenomic for a memory location.
No it's not. Branch optimization means turning a longer branch (eg. 32-bit offset) into a shorter branch (eg. 8-bit offset). Looking at the final machine code, you have no way to know which branch was in the original source code, and so the mapping is not one-to-one.
Ok, we crossed our wires here. I am talking about the branch prediction and super-pipelining abilities of modern processors whereby they can run code out of order and take code paths earlier than expected. See quote below for some info on this:
Speculations
Unlike the Pentium, the 6x86 doesn't wait until the jump conditions are determined in the execution level (something that with a long calculation such as DIV can take as much as 41 cycles), instead it decodes the expected code showing a strong confidence in its prediction ability, and speculatively carries out this code in the second pipeline - if this code is not inseparable from the results of the preceding command dependant. Thereby it can of course lead on to branching out again. The processor speculates up to four branch levels deep. With that it saves four write operations (four level write buffer) and throws them away in the case of incorrect predictions.
For the Pentium it's generally the case, that the pipelines work synchronously. For example, if pipeline Y is finished but X is still working on a command, Y twiddles its thumbs the whole time, waiting for X to finish. With the 6x86 it's different; Y carries out following commands outside of the normal order (out of order execution this is called). Therefore there can be an unforeseen sequence for reading and writing (weak ordering), something that is undesirable in exceptional cases (memory mapped I/O). That's why weak ordering can be turned off for certain address areas.
Anyway, my point is that assembler code has a very direct mapping to machine code. It translates easily both ways because of this mapping.
Actually, VBA which is the most common of the languages used for the modern variant of viruses is considered a scripting language:-) Code Red was the first decent worm I had seen for a long time since it actually did some funky stuff in assembler. All the other recent nasties were variants and improvements on "I Love You". Some added very nifty additions like execution through the preview pane, but most have held to the basic premise:
You have macros if you MASM but these will simply be assembler instructions that are inserted in the place of the macro, labels are simply memory locations and are not extra generated instructions. Branch optimisation is handled by the CPU itself and is applied to the instructions in memory at the time of execution - same with pipeline optimisation. These affect the execution of the code, not the code itself. As for syntax - choose whichever you prefer - for me that's Intel.
Ahem, since this virus was clearly written in assemlber then they are actually publishing the source code. It may have different labels for the JMP instructions, but aside from that (and working out where your data locations are) it should be exactly the same code that the cracker used. Each assembly instruction has a 1 to 1 mapping with machine code instructions.
Still, if they publish the code shown ay eEye then I suspect it won't work since it needs data segment and code segment hints and stuff to make an exe, although it could be incorporated into another project faily easily.
The code presented is pretty standard fare for viruses. Nothing to see here...move along;-) These techniques are already well known and used among h4x0rs.
There have been virus writing kits available for years now with little or no coding required. If this stuff is in assembler then even many experienced programmers wouldn't be able to deal with it. This is *not* going to teach anyone who can't already do it how to write a virus.
For reference: I can write both assembler and viruses (though I don't do the second) so I have a reasonable idea of what I am talking about. I am the only programmer out of 16 in our shop that can even write in assembler.
Slashdot Mirror
Ok, I see where you are coming from now.
Your comments had a ring of Balmer to them i.e. Balmer talks about how the GPL is a virus, and you were talking about how it destroyed copyright. The part what wasn't shining through was it destroyed it in a positive way - or to put it another way; actually provided *copy* *rights*; the right to copy the software.In a literal sense you are right here, but since the license is used to grant rights that are not actually granted under standard copyright laws one can also think of it as a license that is used to provide copy rights.
Can you name one product which has been forced by the GPL to become GPL code also? Even Blender which was commercial software became GPL only because the community banded together and paid a hostage release fee for it. The company was under no duress to GPL the code, and both they and the community have now benefitted greatly from the codes release.
RMS may be a fool for wanting a world completely without copyright, but that doesn't invalidate the value of the GPL license. The GPL provides people like me (a professional full-time employed programmer) the chance to provide a gift to the community that can never be taken away.
Now, I don't expect that all software should be GPL'ed because we need different licenses to cover different needs. That's why each company is free to write it's own restrictive license, or to go with any of the popular Open Source licenses.
Why are you so averse to the GPL?
The GPL is a copyright. It's not designed to destroy copyright but to enhance the rights of the people who use the software.
If I place a copy of Linux on the nightstand with a copy of Windows they will both still be there in the morning. I'm not likely to find the Windows box with it's flaps open and it's floppy hanging out.
Even the viral nature of the GPL is overated. There is nothing that compels you to copy GPL code into your software. If you don't use GPL then you don't need to GPL your software. If you do benefit from using GPL code, then all that is asked is that you enable others to benefit in the same way you did. Not a big ask really. Also, there is nothing in the GPL that says you can't sell the software and profit from that sale. You can, but you must also allow people to obtain it and the source at reasonable cost to distribute. RedHat has built it's business on this.
Also, try not confuse the Open Source movement with anti-capitolism, communism, or anti-gloalisation. Many of those people don't even have a computer let alone have an opinion on what OS to run, and many of us OS people don't have an opinion on their idealogies.
On a personal note, I really love Open Source, am lukewarm on the globalisation thing (why are there four fucking Starbucks on my street?? What happened to the independant cafes?), and am a firm supporter of basic tenets of capitolism, but tempered with socialist values. I'm mostly white, working class, an adult, and do not *force* my choice down anyones throats but will happily speak when asked.
Now, time for you to take your pills and watch some TV, you're a little excited today.
Desktop acceptance may be on the horizon still, but it is only a question of time, since Linux is powering along in the server market.
There's every chance that a late version of Linux like RedHat 9.0 using tar could read those tapes and dump the files for you even now so you wouldn't even need to install an older OS.
There is a lock-in that occurs by having all this data in a format that is difficult to cleanly translate to other word processing/viewer formats. I understand that Open Office can open Word documents, but not all the formatting is cleanly cut across, and so in our business we cannot rely on this. The problem compounds as time drifts onwards. Older documents are held in older formats e.g. Word 2.0, WordPerfect 6.0. At some time it is likely that support for reading those formats will be dropped in future versions of the office suites. At that time it will be very difficult to access that data.
Open Source helps protect against this by ensuring the code to read old data (which government hordes by the kiloton) will always be available.
Open Source does not necessarily mean free. MS could Open Source their products and still sell them. If IBM can do it, why can't they?
10,000,000 of the hits are to an ASP page, and all our ASP pages use at least 3 components to display. Many use more components. All hook up to a back end database. At peak times we are getting 14-16 page impressions / second, which is all handled by a quad xeon processor with a dual xeon for the database server. The web server is now running at capacity, so if demand keeps increasing we will need to cluster.
This is quite an interesting point. It was games that got me hooked on PCs back in the eighties. Some games like Quake and Neverwinter Nights come with very useable scripting languages that let you either add new weapons (and stuff) or script up your own adventures. They are simplified C and thus useful as a stepping stone to "real" programming and have that instant gratification kids are looking for. Yeh, get your kids into Quake C and NWN C!
Visual Basic is still around, but I don't know of any serious programmers who really use it hard core
We use it every day here at work and we're serious :-) We use it mainly because it's fantastic for producing middle tier COM components quickly and easily. The site I develop for takes 100,000,000 hits / month, so I'd say VB is serious. Personally though, perhaps PHP, Python or some other langauge might be better. And definitely buy a small cheap machine that the kids can trash whenever they like. They'll learn better when it's their machine and they know they can destroy it.
It was compulsory in our family when young to help dad repair cars. So yes, he did get me replacing alternators (relatively easy) or striping parts, cleaning and then re-assembling.
and yet you'd let your 12 year old code in Perl? Aw, c'mon, show the poor little tyke some mercy and at least let him start with a structured and easy to learn language first.
dd if=/dev/urandom of=winsock.dll count=8749
so you don't run out of drive space ;-)
He already did, thanks to Redhat generously giving him a million dollars in shares ;-) It's in his book "Just for Fun".
By default all rights to written works (software included) default to the author. In Australia at least, all you need to do to make software copyrighted is to place a small label on the packaging and again when the software starts up. Easy. Placing said notes in code would be sufficient for it to be considered copyrighted, even if the infringing party claims they haven't read that notice, they are still held to it.
In the meantime, the copyright holders of the Linux code can request that SCO's product be removed from sale and SCO issue a recall on all the sold product until a case is heard. Now that *would* be interesting.
Just keep your mail on your mail server and use IMAP to connect to it. Then you can use Mutt, Evolution, KMail, Netscape/Mozilla or even Outlook ot get at your mail. Best of all, you can access it while you are at work too ;-) I presonally use Kmail when I'm at home, Mozilla if I have to boot Windows for a bit, Pine when I shell in, and Mozilla from work (since Outlook acted all pissy about it for no good reason). IMAP means enjoying your mail no matter where you are.
You can also move all your contacts and stuff into an LDAP server for bonus points. Now I have one address book which I can use wherever I am with all my email clients on any OS. Only downside, I have yet to find a decent client to update that LDAP address book...anyone got some recomendations?
It's not quite that simple since the compiler has many oportunities to apply optimisations.
Exactly. So a disassembler can't recreate them. Just like comments.A disassembler will provide equivalent labels, but it will not give them the same name as they originally had. They end up being called (for example) LABEL01, LABEL02, but they are still doing the same job, that it providing a mnenomic for a memory location.
No it's not. Branch optimization means turning a longer branch (eg. 32-bit offset) into a shorter branch (eg. 8-bit offset). Looking at the final machine code, you have no way to know which branch was in the original source code, and so the mapping is not one-to-one.
Ok, we crossed our wires here. I am talking about the branch prediction and super-pipelining abilities of modern processors whereby they can run code out of order and take code paths earlier than expected. See quote below for some info on this:
Anyway, my point is that assembler code has a very direct mapping to machine code. It translates easily both ways because of this mapping.
Get executed
Send yourself to entire address book
Profit!
You have macros if you MASM but these will simply be assembler instructions that are inserted in the place of the macro, labels are simply memory locations and are not extra generated instructions. Branch optimisation is handled by the CPU itself and is applied to the instructions in memory at the time of execution - same with pipeline optimisation. These affect the execution of the code, not the code itself. As for syntax - choose whichever you prefer - for me that's Intel.
Why, because I can write assembler? Did you not read the bit where I said I choose not to write viruses?
Still, if they publish the code shown ay eEye then I suspect it won't work since it needs data segment and code segment hints and stuff to make an exe, although it could be incorporated into another project faily easily.
The code presented is pretty standard fare for viruses. Nothing to see here...move along ;-) These techniques are already well known and used among h4x0rs.
For reference: I can write both assembler and viruses (though I don't do the second) so I have a reasonable idea of what I am talking about. I am the only programmer out of 16 in our shop that can even write in assembler.