Some of the key advantages will be: 1) decouple data, logic, and presentation 2) allow client side rules-based validation 3) spits back an XML record, maybe w/ schema validation 4) replaces a lot of javascript with markup 5) highly device independent (eg render an XForm via telephone, web browser, handheld)
Too late. The final call for modifications to the XForms proposal has passed. The GUI kit project should change its name, since it is a non-compliant implementation its own namesake. Or they can do nothing and confuse the crap out of everybody.
The Firebird RDBMS folks screamed loud, early, and often at the name collision. This standard has been around for YEARS, so it's a little late to play the change your name game.
I always get amused when open source people spout of about how slow Java supposedly is and how much they love perl/python/php in the other. Java is pretty much faster than all of those. Yet nobody ever dogs on python (say) for being slow. Why?
I think it is not unlikely that IBM will file counterclaims that SCO will have to defend. I'd expect them to pull out every patent that they've contributed to Linux under terms that require GPL distribution and to bitch-slap SCO with them. Oddly enough, one of them is the RCU patent. I would not be surprised to see them to retailiate with business libel, unfair competition, and deceptive trade practices claims.
I do like perl, but I was not think of perl when I wrote that. Actually, the language "environment" that is closest to all of my desires amond those I have worked with is the duo of Jython and Java.
I would criticize this duo in several ways: why are they such different syntax? Also, in the big picture, when you compare to C and Pascal, Java is slow. I know there are native compilers like gcj, but they are an afterthought.
I will check into OCaml. I know that on Bagley's Language Shootout, it scored comparable to C in speed, but it is obviously a much more modern language.
As to the standard libararies, I went back and forth on this. At some point you have to accept that if the language designer makes a good core language, that the community will arrive and handle ever expanding libraries. You can't "design" the community, so I left that off. Obviously it's desirable.
If OCaml would be your ideal language but for limited standard libraries, then you should treat this as an "itch" and scratch it by adding a library or two.
Cite a reference then, because the GPL does not cover compiling AT ALL from my reading
To be precise, it requires you to abide by 2b when you distribe object code or executable:
GPL: "3. You may copy and distribute the Program (or a work based on it, under Section 2) in object code or executable form under the terms of Sections 1 and 2 above provided that you also do one of the following"
This is clearly ADDING conditions beyond those of 2b, which it tacitly (and correctly) assumes would apply anyway because the object code is a derivitive of the original source form in the copyright sense.
2. covers modifications, how do you know that it covers simply copying and redistributing other people's code?
Gee, I've gone over this twice already: 1) GPL 3 requires 2b compliance to distribute executables 2) Compiling source to object creates a derivitive work under copyright law. GPL 2b applies when you distribute any derivitive work so 2b applies even without referencing 3. Three place *additional* restrictions on distribution of binaries.
If a malicious person adds code to which they DO NOT own the copyright, then even if they "place a notice" they are NOT the "copyright holder" and don't satisfy 0.
Correct as to the code that they added. I have never stated that IBM GPL'd SCO's code. However for the code that was already there, namely the code written long ago by Linus, section 0 was met and this code is still validly GPLed. Thus any work "based on it" must be GPL'd to be properly licenced. If IBM's happens to be this malicious user, that does not release SCO from its duty to respect the copyrights of Linus and the other contributors.
Consider: A = Linus's properly GPL'd kernel source code B = SCO's copyrighted unix source code
IBM allegedly created C = A+B, which would infringe BOTH A and B, for differnet reasons.
Then SCO obtained C, then they modified it by producing exectuables, which they then distributed. But C is a derivitive of A, so SCO needs a license on A to do any of these things. A As explained above, section 3 and thus 2b of the GPL apply to SCO when they distributed executables derived in part from A.
If they DID modify the code, that would make your argument more valid There was an article recently about how a SCO employee contributed to JFS and SMP. But again, in a copyright sense, changing source to exectuable is a modification. You are thinking solely in terms of "patching" in the software development sense. That is a narrower concept.
Can we focus on getting decent broadband to everyone first, and THEN start worrying about 12 ms of ping time? Good god, man.
You'll get it when you get it. It takes approximately one hour to convince yourself that your pre-broadband days were just a distant memory. We're working hard to make sure that one hour after you get broadband that you are still happy.
But seriously, the pingpath app linked above runs on some weird OS that I don't use. Does anybody know what the equivalent Linux functionality is?
10ms is less than one refresh interval on your monitor, so you really can't "see" the difference.
For gaming, though you often have human race conditions. The frame is drawn, the two players see each other for the first time and hit the "fire" button. Whoever gets the message to the server first kills the other. Taking a.01 second hit absolutely can make a difference even if it is less than the frame redraw time. Consider all the Olympic events and horse races etc... that have been decided by such margins. Maybe you can "see" the difference, but the difference matters given that whoever is faster on the draw wins.
I'd like a language that could run in three different modes: interactive interpreted form, highly portable runtime form, and full natvie speed optimized form. Give me common syntax for each and let them interact. For example, if you want to, you can use Jython to script java code. If you need something to be really fact, you write it in C and us the native interface. Why do I have to learn three syntaxes for this? OK, Jython is dynamically type and Java is statically typed.
Give me both OO abstraction and powerful bit-twiddling capabilities. If I want a 29 bit variable type, dammit I want a 29 bit variable type. Give it to me and let me do all kinds of low level stuff like shift three bits and mask it and treat the first 3 bytes as a unit. Then let me encapsilate that in an object oriented lock box.
The language MUST have these things built in: associate arrays, regular expressions, iterators, introspection, "here" documents, both static and dynamicly sized arrays.
The language should be database savy. This means implicit SQL cursors (like in SQLJ or PL/SQL) so that I can say something like
DBI_DRIVER.set(driver) DBI_DRIVER.login(dbname, user, passwd) for current_employee in {{
select emp_name, salary, start_date
from employee
where dept_no = $department }} do
{{update employee set salary = salary + $accounting.calulate_bonus}} end"
Section 2, which discusses modifications, does not specify source, it doesn't specify anything. Your ideas about binaries etc. are YOUR opinion.
Creating the binary from the source is explicitly covered by section 3 of the GPL, which explicitly backreferences section 2. Under copyright law, it would also be considered the preparation of a derivitive work, thus requiring a license. Neither of those is based solely on "my opinion".
On the "place the notice" issue, there are only two reasonable ways to interpret section 0. It either is either a "first release" requirement which was satisfied years ago by Linus, or it is a "per distribution" requirement, in which case the notice is "placed" by copying it intact from the original. SCO's argument makes no sense: it would seem to say that each modifier has to retype it in, and completely ignores copyright's governance over distribution too. Among the two interpretations, I think the "first release" interpretation is the correct one. Way back in 1991, Linus placed the notice. At that point the code was 100% his so there was no issue of mixing code from different authors. Once a pure GPL "program" exists, section 0 clearly tells us that the licence applies to any "derivative work under copyright law". Accordingly, is not section 0 that governs subsequent copyright uses, but the other sections. In particular, section 2b and section 3's backreferencing of section 2b govern the supposed inclusion by IBM of SCO's IP. As does section 5.
Assuming this was improperly done, IBM infringed both SCO and Linux by doing it. SCO obtained this improperly licenced mix of their IP and Linux IP, and then modified, copied, and distributed it. Unless SCO repaired the licencing problem created by IBM, they have no right either to distribute a work based on Linux. SCO as the copyright owner of the non-GPL pieces had the power to fix it and was required by section 2b to fix it. If they refused to, they are just as guilty of infringing Linux as IBM is of infringing them. I think the argument that SCO is making about their code not being GPL would be a good argument if they had not distributed the rest. Certainly IBM cannot GPL SCO's code. But SCO can do so, and is independently required to in order to *distribute* Linux.
The argument that you can't inadvertantly GPL your own code doesn't really fly -- you absoluetly CAN commit unintentional and unknowing copyright infringement by not meeting a licencing term. Moreover, you have a duty to all copyright holders to assure that your actions don't infringe their work. Distribution is such an action. The GPL section 5 states "by modifying or distributing the Program (or any work based on the Program), you indicate your acceptance of this License". Such a clause is perfectly consistent with contract law -- contract acceptance may be defined as taking an action that you cannot legally take otherwise.
SCO undeniably has modified the linux kernel source code, in several ways. The most obvious is that they have participated in the kernel development process and contributed code to it. But that is not even the best argument. Note that section 2 does NOT say modification at the source code level. Simply distributing binaries or RPMs is sufficient as that was not its original form. If you read section 2.3, you'll see that distributing object code explicitly counts as a modification that subjects you to the conditions of items 1 and 2, including 2b.
Yes, I am suggesting SCO is in gross violation of the license. It's kind of bizarre for them to be griping about supposedly 1% of the Linux kernel being their IP while they ship the other 99%, which doesn't allow distribution of a mix.
I don't understand SCO's "place a notice" argument. Download the source from their site. You'll find the GPL notice, which they placed upon it when they copied it into the rpm they are distributing. There is no requirement that they redraft it by hand every time. The "place a notice" language applies to the first licensing. It was done by Mr. Torvalds. After that, anyone else is bound by the terms of the GPL or they are commiting copyright infringement against Torvalds and everyone else who contributed.
Read section 2b which requires them to licence what they ship in its entirety under the GPL in order to distribute it.
But what is the alternative here? Are you suggesting that SCO is distributing the Linux Kernel with no licnece at all? I'm sure the courts will enjoy hearing about how SCO entire Linux revenue stream was based on copyright infringement.
And don't bother with the standard "they didn't know they were infringing argument". State of mind is irrelevent. Unknowing, unintentional infringement is still infringment and all profits derived from the infringement are actual damages. And since they are still distributing the linux kernel today, they do know they are pirating it.
Converting.jpg and.png to SVG (Scalable VECTOR Graphics) would be an extremely pointless thing to do, not to mention nigh on impossible.
The point is that after doing it, you have a scalable image, with source code. Both aspects have tremendous value. If your project drew its own icons and pictures, by redrawing with SVG, you can create source code whose drawings can be modified and reused in a scalable way.
The difficulty in redrawing them in SVG clearly depends on what the image is. For some photographic types of images it may not be appropriate, but most things used in apps are hand drawn and it is rediculous to claim that it is impossible. For example, consider slashdot's penguin image: http://www.home.unix-ag.org/simon/penguin/
Actually, you get the higher of statuory damages and actual damages. Actual damages include all profits derived form the infringement. I would argue that all of SCO's past profits from Linux are based on piracy of GPL'd kernel code.
Actually, all it takes is for people to create graphics in SVG. Given that SVG creates a true "source code" for graphics, the open source movement will eventually embrace SVG and everybody else will be behind.
I would encourage all open source projects that have images to start producing their.jpg and.png by using SVG source code and adding a conversion step to their build code.
Well, it looks like the examples at http://www.croczilla.com/svg/ work fine -- but they are all.xml files. How do I tell Mozilla to go ahead and try to render the.svg files. It seems to want to force me to use a handler app.
OK, I got the new Mozilla SVG enabled browser and I go to this page and click on the logos and up pops the dialog saying I've got an SVG mime-type file and it asks if I want to say or use an applicaiton.
Where's the part where you get to render the SVG in Mozilla?
Would people please stop repeating this mindless knock on XML!
XML compresses very well, and in most situations compressed xml is smaller than the custom binary formats it is replacing. In the context of SVG, the.svgz file format already exists to compress the raw SVG xml, and the result is quite good.
If you have a robots.txt entry and somebody ignores it, then you should have a good case to sue them. If you don't have a robots.txt entry and somebody relies on it, you should have no case to sue.
Hopefully, the legal standard will be that when you choose a particular technolgy to disseminate your works, that the documents that define that technology become legally binding. In other words, the law should reinforce the technology.
I think that if a site requires the referring page to come from one its own domain and you spoof that somehow to make a deep link, then you would be guilty of copyright infringement, for unauthorized derivitive work. Deep-linking should be legal as an explicit authorization rule, because the web server authorizes you to use the work in the way your HTTP GET requests. If your HTTP GET header is fraudulent, then you don't have a grant to use it.
If a site requires it and you use a proxy that strips it off, then you will simply be denied access. That's the proxie's fault for purposefully thwarting the HTTP standard, so I have no sympathy for them.
Sympathy for McVoy is irrelevent. Free software is trying to defeat the proprietary software model by being more efficient. If you think its important to feed his family, send him a check, but don't perpetuate a software model you believe is inferior based on sympathy. His customers need the money they would save just as much as his company does.
The solution to this is quite simple. Free software needs to build a better SCM tool. It should not copy BitKeeper, though not for the reason McVoy gives. His reasoning is wrong -- nothing forces a particular deployment to upgrade to the new protocols, so if you become compatible with any particular version, you've achieved what you need to switch. The real reason not to copy BitKeeper is because you cannot be the best by being second. Although, if McVoy spends all his resources on gimmicks like breaking API's, maybe you can catch and pass.
I agree completely that browsers need to support SVG, and until this is more advanced, tools like the one in this article are getting ahead of things (not that that's necessarily bad). People should vote for Mozilla bug 122092: "Enable SVG support". Ultimately, this needs to be done without plugins -- its really just another image format.
The Konqueror browser seems to have a push to get SVG going too: KSVG, but it has a way to go ("Release 0.1 pending").
There are a good set of SVG resources for Linux. The Apache Jakarta projects java SVG viewer, Batik is probably the farthest along.
Slashdot Mirror
Some of the key advantages will be:
1) decouple data, logic, and presentation
2) allow client side rules-based validation
3) spits back an XML record, maybe w/ schema validation
4) replaces a lot of javascript with markup
5) highly device independent (eg render an XForm via telephone, web browser, handheld)
Too late. The final call for modifications to the XForms proposal has passed. The GUI kit project should change its name, since it is a non-compliant implementation its own namesake. Or they can do nothing and confuse the crap out of everybody.
The Firebird RDBMS folks screamed loud, early, and often at the name collision. This standard has been around for YEARS, so it's a little late to play the change your name game.
I always get amused when open source people spout of about how slow Java supposedly is and how much they love perl/python/php in the other. Java is pretty much faster than all of those. Yet nobody ever dogs on python (say) for being slow. Why?
I think it is not unlikely that IBM will file counterclaims that SCO will have to defend. I'd expect them to pull out every patent that they've contributed to Linux under terms that require GPL distribution and to bitch-slap SCO with them. Oddly enough, one of them is the RCU patent. I would not be surprised to see them to retailiate with business libel, unfair competition, and deceptive trade practices claims.
I do like perl, but I was not think of perl when I wrote that. Actually, the language "environment" that is closest to all of my desires amond those I have worked with is the duo of Jython and Java.
I would criticize this duo in several ways: why are they such different syntax? Also, in the big picture, when you compare to C and Pascal, Java is slow. I know there are native compilers like gcj, but they are an afterthought.
I will check into OCaml. I know that on Bagley's Language Shootout, it scored comparable to C in speed, but it is obviously a much more modern language.
As to the standard libararies, I went back and forth on this. At some point you have to accept that if the language designer makes a good core language, that the community will arrive and handle ever expanding libraries. You can't "design" the community, so I left that off. Obviously it's desirable.
If OCaml would be your ideal language but for limited standard libraries, then you should treat this as an "itch" and scratch it by adding a library or two.
Cite a reference then, because the GPL does not cover compiling AT ALL from my reading
To be precise, it requires you to abide by 2b when you distribe object code or executable:
GPL: "3. You may copy and distribute the Program (or a work based on it, under Section 2) in object code or executable form under the terms of Sections 1 and 2 above provided that you also do one of the following"
This is clearly ADDING conditions beyond those of 2b, which it tacitly (and correctly) assumes would apply anyway because the object code is a derivitive of the original source form in the copyright sense.
2. covers modifications, how do you know that it covers simply copying and redistributing other people's code?
Gee, I've gone over this twice already:
1) GPL 3 requires 2b compliance to distribute executables
2) Compiling source to object creates a derivitive work under copyright law. GPL 2b applies when you distribute any derivitive work so 2b applies even without referencing 3. Three place *additional* restrictions on distribution of binaries.
If a malicious person adds code to which they DO NOT own the copyright, then even if they "place a notice" they are NOT the "copyright holder" and don't satisfy 0.
Correct as to the code that they added. I have never stated that IBM GPL'd SCO's code. However for the code that was already there, namely the code written long ago by Linus, section 0 was met and this code is still validly GPLed. Thus any work "based on it" must be GPL'd to be properly licenced. If IBM's happens to be this malicious user, that does not release SCO from its duty to respect the copyrights of Linus and the other contributors.
Consider:
A = Linus's properly GPL'd kernel source code
B = SCO's copyrighted unix source code
IBM allegedly created C = A+B, which would infringe BOTH A and B, for differnet reasons.
Then SCO obtained C, then they modified it by producing exectuables, which they then distributed. But C is a derivitive of A, so SCO needs a license on A to do any of these things. A As explained above, section 3 and thus 2b of the GPL apply to SCO when they distributed executables derived in part from A.
If they DID modify the code, that would make your argument more valid
There was an article recently about how a SCO employee contributed to JFS and SMP. But again, in a copyright sense, changing source to exectuable is a modification. You are thinking solely in terms of "patching" in the software development sense. That is a narrower concept.
Can we focus on getting decent broadband to everyone first, and THEN start worrying about 12 ms of ping time? Good god, man.
You'll get it when you get it. It takes approximately one hour to convince yourself that your pre-broadband days were just a distant memory. We're working hard to make sure that one hour after you get broadband that you are still happy.
But seriously, the pingpath app linked above runs on some weird OS that I don't use. Does anybody know what the equivalent Linux functionality is?
10ms is less than one refresh interval on your monitor, so you really can't "see" the difference.
.01 second hit absolutely can make a difference even if it is less than the frame redraw time. Consider all the Olympic events and horse races etc... that have been decided by such margins. Maybe you can "see" the difference, but the difference matters given that whoever is faster on the draw wins.
For gaming, though you often have human race conditions. The frame is drawn, the two players see each other for the first time and hit the "fire" button. Whoever gets the message to the server first kills the other. Taking a
Give me both OO abstraction and powerful bit-twiddling capabilities. If I want a 29 bit variable type, dammit I want a 29 bit variable type. Give it to me and let me do all kinds of low level stuff like shift three bits and mask it and treat the first 3 bytes as a unit. Then let me encapsilate that in an object oriented lock box.
The language MUST have these things built in: associate arrays, regular expressions, iterators, introspection, "here" documents, both static and dynamicly sized arrays.
The language should be database savy. This means implicit SQL cursors (like in SQLJ or PL/SQL) so that I can say something like
Sorry that should be section 3, not 2.3. My bad.
Section 2, which discusses modifications, does not specify source, it doesn't specify anything. Your ideas about binaries etc. are YOUR opinion.
Creating the binary from the source is explicitly covered by section 3 of the GPL, which explicitly backreferences section 2. Under copyright law, it would also be considered the preparation of a derivitive work, thus requiring a license. Neither of those is based solely on "my opinion".
On the "place the notice" issue, there are only two reasonable ways to interpret section 0. It either is either a "first release" requirement which was satisfied years ago by Linus, or it is a "per distribution" requirement, in which case the notice is "placed" by copying it intact from the original. SCO's argument makes no sense: it would seem to say that each modifier has to retype it in, and completely ignores copyright's governance over distribution too. Among the two interpretations, I think the "first release" interpretation is the correct one. Way back in 1991, Linus placed the notice. At that point the code was 100% his so there was no issue of mixing code from different authors. Once a pure GPL "program" exists, section 0 clearly tells us that the licence applies to any "derivative work under copyright law". Accordingly, is not section 0 that governs subsequent copyright uses, but the other sections. In particular, section 2b and section 3's backreferencing of section 2b govern the supposed inclusion by IBM of SCO's IP. As does section 5.
Assuming this was improperly done, IBM infringed both SCO and Linux by doing it. SCO obtained this improperly licenced mix of their IP and Linux IP, and then modified, copied, and distributed it. Unless SCO repaired the licencing problem created by IBM, they have no right either to distribute a work based on Linux. SCO as the copyright owner of the non-GPL pieces had the power to fix it and was required by section 2b to fix it. If they refused to, they are just as guilty of infringing Linux as IBM is of infringing them. I think the argument that SCO is making about their code not being GPL would be a good argument if they had not distributed the rest. Certainly IBM cannot GPL SCO's code. But SCO can do so, and is independently required to in order to *distribute* Linux.
The argument that you can't inadvertantly GPL your own code doesn't really fly -- you absoluetly CAN commit unintentional and unknowing copyright infringement by not meeting a licencing term. Moreover, you have a duty to all copyright holders to assure that your actions don't infringe their work. Distribution is such an action. The GPL section 5 states "by modifying or distributing the Program (or any work based on the Program), you indicate your acceptance of this License". Such a clause is perfectly consistent with contract law -- contract acceptance may be defined as taking an action that you cannot legally take otherwise.
SCO undeniably has modified the linux kernel source code, in several ways. The most obvious is that they have participated in the kernel development process and contributed code to it. But that is not even the best argument. Note that section 2 does NOT say modification at the source code level. Simply distributing binaries or RPMs is sufficient as that was not its original form. If you read section 2.3, you'll see that distributing object code explicitly counts as a modification that subjects you to the conditions of items 1 and 2, including 2b.
Yes, I am suggesting SCO is in gross violation of the license. It's kind of bizarre for them to be griping about supposedly 1% of the Linux kernel being their IP while they ship the other 99%, which doesn't allow distribution of a mix.
I don't understand SCO's "place a notice" argument. Download the source from their site. You'll find the GPL notice, which they placed upon it when they copied it into the rpm they are distributing. There is no requirement that they redraft it by hand every time. The "place a notice" language applies to the first licensing. It was done by Mr. Torvalds. After that, anyone else is bound by the terms of the GPL or they are commiting copyright infringement against Torvalds and everyone else who contributed.
Read section 2b which requires them to licence what they ship in its entirety under the GPL in order to distribute it.
But what is the alternative here? Are you suggesting that SCO is distributing the Linux Kernel with no licnece at all? I'm sure the courts will enjoy hearing about how SCO entire Linux revenue stream was based on copyright infringement.
And don't bother with the standard "they didn't know they were infringing argument". State of mind is irrelevent. Unknowing, unintentional infringement is still infringment and all profits derived from the infringement are actual damages. And since they are still distributing the linux kernel today, they do know they are pirating it.
Converting .jpg and .png to SVG (Scalable VECTOR Graphics) would be an extremely pointless thing to do, not to mention nigh on impossible.
/
The point is that after doing it, you have a scalable image, with source code. Both aspects have tremendous value. If your project drew its own icons and pictures, by redrawing with SVG, you can create source code whose drawings can be modified and reused in a scalable way.
The difficulty in redrawing them in SVG clearly depends on what the image is. For some photographic types of images it may not be appropriate, but most things used in apps are hand drawn and it is rediculous to claim that it is impossible. For example, consider slashdot's penguin image:
http://www.home.unix-ag.org/simon/penguin
Actually, you get the higher of statuory damages and actual damages. Actual damages include all profits derived form the infringement. I would argue that all of SCO's past profits from Linux are based on piracy of GPL'd kernel code.
Actually, all it takes is for people to create graphics in SVG. Given that SVG creates a true "source code" for graphics, the open source movement will eventually embrace SVG and everybody else will be behind.
.jpg and .png by using SVG source code and adding a conversion step to their build code.
I would encourage all open source projects that have images to start producing their
Well, it looks like the examples at http://www.croczilla.com/svg/ work fine -- but they are all .xml files. How do I tell Mozilla to go ahead and try to render the .svg files. It seems to want to force me to use a handler app.
OK, I got the new Mozilla SVG enabled browser and I go to this page and click on the logos and up pops the dialog saying I've got an SVG mime-type file and it asks if I want to say or use an applicaiton.
Where's the part where you get to render the SVG in Mozilla?
I would expect that the option to turn off javascript for HTML pages would work exactly the same way for SVG, so I don't really see this as a concern.
Yep. I was one of them. This is absolutely great news. Hopefully the folks over at ksvg will see this and buckle down.
Would people please stop repeating this mindless knock on XML!
.svgz file format already exists to compress the raw SVG xml, and the result is quite good.
XML compresses very well, and in most situations compressed xml is smaller than the custom binary formats it is replacing. In the context of SVG, the
If you have a robots.txt entry and somebody ignores it, then you should have a good case to sue them. If you don't have a robots.txt entry and somebody relies on it, you should have no case to sue.
Hopefully, the legal standard will be that when you choose a particular technolgy to disseminate your works, that the documents that define that technology become legally binding. In other words, the law should reinforce the technology.
I think that if a site requires the referring page to come from one its own domain and you spoof that somehow to make a deep link, then you would be guilty of copyright infringement, for unauthorized derivitive work. Deep-linking should be legal as an explicit authorization rule, because the web server authorizes you to use the work in the way your HTTP GET requests. If your HTTP GET header is fraudulent, then you don't have a grant to use it.
If a site requires it and you use a proxy that strips it off, then you will simply be denied access. That's the proxie's fault for purposefully thwarting the HTTP standard, so I have no sympathy for them.
Sympathy for McVoy is irrelevent. Free software is trying to defeat the proprietary software model by being more efficient. If you think its important to feed his family, send him a check, but don't perpetuate a software model you believe is inferior based on sympathy. His customers need the money they would save just as much as his company does.
The solution to this is quite simple. Free software needs to build a better SCM tool. It should not copy BitKeeper, though not for the reason McVoy gives. His reasoning is wrong -- nothing forces a particular deployment to upgrade to the new protocols, so if you become compatible with any particular version, you've achieved what you need to switch. The real reason not to copy BitKeeper is because you cannot be the best by being second. Although, if McVoy spends all his resources on gimmicks like breaking API's, maybe you can catch and pass.
I agree completely that browsers need to support SVG, and until this is more advanced, tools like the one in this article are getting ahead of things (not that that's necessarily bad). People should vote for Mozilla bug 122092: "Enable SVG support". Ultimately, this needs to be done without plugins -- its really just another image format.
The Konqueror browser seems to have a push to get SVG going too: KSVG, but it has a way to go ("Release 0.1 pending").
There are a good set of SVG resources for Linux. The Apache Jakarta projects java SVG viewer, Batik is probably the farthest along.