Finally, can you say bloat? XML, DOM, CSS, network stack, etc, etc. -- that's alot of code just to display a random graphic. I guess the footprint of modern web browsers has numbed us to the fact that software is getting slower, not faster.
XML fights bloat by standardizing layers and re-using components. An SVG implementation need not include its own XML parser, network stack, Unicode components, or Core DOM libraries.
Quit bashing SVG. As "design by committee" goes, SVG is a shining example of success. Educate yourself before you go off.
No, that's not right. VML predates the SVG working group; it was available around 1998. They developed it as a vector format with the cooperation of HP, Autodesk, and others with the intention of creating a de facto standard, but submitted it as a primary input for the SVG standard when the SVG working group was formed.
VML sucks, *really* sucks. But that isn't terribly important, since SVG *doesn't* suck. The other major input to the SVG working group was Adobe (et al)'s PGML standard.
As far as I can tell, VML really isn't complete, at least not in terms of implementation. IE 5.5 claims to support VML but
1) Anti-aliasing is not implemented, except for fonts
2) The 1.0 specification is fraught with inconsistencies, errors, and ambiguity
3) The documentation on Microsoft's Web site is dated 1998
4) Microsoft's own VML generation tool supports tags that aren't listed in the 1.0 specification (like ).
Text works well for vector graphics (see Postscript, CGM, DXF). Text does *not* work well for rasterized graphics, except in limited use (see XPM).
If you think that vector graphics should be stored in binary, then you might as well argue that HTML should have been a binary format. The same principles apply either way.
The crux of the matter is that the XML parts are all useful and well-designed, whatever the cost. It's worth it to have full XML addressing in the SVG specification, without a doubt. Core XML libraries do most of the work, after all.
The problem is that the SVG DOM is completely head-up-ass insane. It is a nightmare to implement; I've written proof-of-concept handlers in both C++ and Python, and they were both difficult (although the Python version was far easier: 10% code size and 10% time-to-implement by comparison to my C++ version, plus I get free GC).
I suspect that SVG DOM support will be largely ignored by vendors for years to come, and even when it's available it will be wildly variant between implementations. It's a real bitch.
This is all quite true; XML works well for SVG but the DOM is just plain awful for event-driven development. CSS Events are pathetic, and the DOM is a memory pig.
I'm ambivalent at this point. You can't really afford to do without something like the XML DOM, but it leaves the programmer in the difficult position of having to do a lot of imaginative coding to balance functionality with standards-compliance.
I wish the Core DOM could be redesigned, but it's too late at this point.
MJP
Borsook is foolish, read some responses
on
Selfish Society
·
· Score: 1
Bottom line, the agenda held precious by Borsook and her colleagues is being torn apart by free thinking. They're terrified, and Borsook's awful writing confirms it.
Hell, gonna wear my shirt out to the bar tonight, always get into interesting discussions about it.
I must be doing it wrong. Every time I wear my DeCSS t-shirt I get a hundred people asking me how to pronounce DeCSS. I rarely get anything more interesting. What am I doing wrong?
This is terrific content for a discussion Website, and rather out-of-character for Slashdot (unfortunately). Congratulations are in order for jamie.
MJP
"Government has arrogated so extensive a role to itself that it's understandable that many people might imagine that nothing the government has a hand in could possibly have happened without it." -- Brian Doherty
This was exactly what I thought of when I saw the article posting. For the curious, "Tank Farm Dynamo" can be found in Brin's short story collection entitled "The River of Time".
"Please, let's go back into hiding before we actually have to confront our fuckwitted government."
Looking at the span of history, I see the present as the first time free people have ever privately possessed the technology to *stay* free. I don't plan to walk away from that for any reason, lest my children call me a coward and a hypocrite.
I'm glad to hear about your new book, and for the sake of posterity I'll read it when I can get my hands on it. A more complete single-volume effort would be great news.
I do understand the constraints of your publishing deadline, and in all fairness you did explain them in the class. In the context of the only C++ course I ever got at UTD, however, it was a pretty disappointing situation.
I am probably not the typical C++ student, and I understand that you can't please everybody. I offered my remarks only as a single datum for the feedback forum, not as an unbiased reviewer, myself.
As a student of Mr. Heller at the University of Texas at Dallas, taking his C++ course in a CS track, I ended up hating this book.
First of all, this book should be given to people who don't know how to program anything beyond Visual Basic or Perl. It is fundamentally inappropriate for CS majors, especially since it covers -- as the reviewer accurately reports -- about half of the necessary material.
Mr. Heller would have done better introducing the C Programming Language, because he really fails to communicate more than the utter basics of object-oriented programming. It's not that he doesn't mean well, but the fact remains that C++ is a fascinatingly complex language, and anyone who is fooled into thinking that he/she can gain useful C++ skills from this book is in for a sorry time.
I realize that many people will need an extremely gradual introduction to the language, but at the very least Mr. Heller could have finished the job. I mean, really: there's no discussion of how to use polymorphism, no discussion of class modelling, no genuine engagement of inheritance issues, and -- except for the very basics -- almost no discussion of C++ class syntax whatsoever.
I know I'm being hard on the book and on Mr. Heller. It's not out of animosity toward Heller in the least; I just feel that those who are ready to tackle C++ should grab a good introduction -- "Practical C++ Programming" from O'Reilly comes to mind -- and just *learn* the language once and for all. Otherwise you'll end up learning it ten or more times over the next four years (the way I did) piecemeal, in a haphazard fashion that is likely to thoroughly confuse even the most brilliant student.
My apologies to Mr. Heller for harsh remarks; he's a fair and intelligent man, personally.
Wrong. Illegal copying is not "stealing". Go ahead, look it up in the dictionary. There are no illegal copying scenarios that I know of which involve deprivation of property.
(1) makes legitimate (2).
Given (1)'s inherent flaw, by your own admission this puts (2) on shaky ground.
(2) can't be enforced for all.
This isn't about enforcement. This is about the moral status of data copying. If Metallica's copyright infringement was forgivable, you still fail to justify the Napster lawsuit. If it was not forgivable, then we might have a different conversation about this lawsuit.
(3) makes legit (4)
No, it doesn't, lackwit. Metallica copied The Misfits' album, apparently without fear of law-enforcement. I still can't see how the enforcement issue has anything to do with the scale of the transgression. They are independently variable.
and gives an easy attack point for (2), which is covered under law.
This is your conclusion? The justification for law is its ease of use within the context of specific circumstance?
This is simple. If you have any further questions I will write in simple terms.
A real logical argument, rather than this valueless exposition of nonsense, would be a worthwhile follow-up.
In other words, the recording industry made money in exactly the way everyone else in a free market makes money: by exploiting a scarcity. They took advantage of the scarcity of recorded media and built a massive business around it.
The scarcity is coming to an end, and so is the business model of the recording industry. The question for defenders of the Napster lawsuits is: what makes you think that we should give the recording industry special treatment so that it can sustain its business model?
The copyright laws are explicitly designed to create an artificial scarcity. The justification for this legislation is that it specially motivates artists to create music. I think that the burden of proof is upon those who demand special treatment; do we have a shortage of musicians?
Using Napster cannot be wrong because it's easy. That is an idiotic justification for law. If it really matters that the scale of bootlegging has changed, has it ever occurred to anyone that it might not be wrong in the first place?
As a programmer, my natural inclination is to reject management forces. But I have come to the inescapable conclusion that software development is a process that must be managed, on some level, as an organized task.
The current systems and processes for software development are revolting, but that doesn't change the need for organization. Existing processes are modelled on old engineering practices and driven by corporate machines. Open Source development is changing the underlying models, and we can expect processes to follow suit as the effectiveness of OS organizational models gains recognition.
The important thing is to continue developing and practicing those new organizational models, and to continue pointing out those models which bear fruit. The Linux kernel and the Apache Web server stand as monuments to efficiency and quality, and they are destined to be among the forefathers of whatever processes we use in the next decade.
I just hope that we don't reject corporate development models by rejecting the entire notion of organizational intelligence. Next-generation management models would be a giant step forward for software development; management-annihilation would be a step backward.
MJP
Re:The refusal to understand was funny *once*.
on
Product Placement
·
· Score: 1
Dude, the reason you're getting laughed off the stage is for saying things like "Malda has a massive influence in Open Source" and "his followers blindly, uncritically accept his every word". Gross overstatement is a recipe for disaster in rhetoric. If you want to be an effective speaker, you need to pick up on that.
In this specific case, it's worse than usual. Malda almost never offers editorial opinions. Mostly he offers them on Geeks in Space, where they're offered 1) on demand, and 2) mildly, to say the least. Malda is not a demagogue, plain and simple. He's a personality, mostly for entertainment purposes.
Anyway, your analysis is appreciated, but you're really making a mountain out of a molehill. That's the reason so many people are dismissing you out of hand.
You sound like a little kid who thinks that they're smarter than their parents and should be able to do whatever they want.
If you wanted an ad hominem debate, why didn't you say so from the beginning? You sound like a desperate curmudgeon who wants to paint the world with a single color to save his preconceived notions.
All of which is beside the point of logical argument, of course. But you already knew that, didn't you?
In case you didn't notice, the law routinely disagrees with your position. And after reading your post in toto, I'm way more inclined to trust their intelligence than yours.
If I hear you right, this issue has nothing to do with what you think for yourself; you're more inclined to leave the thinking to someone else, the one whose "intelligence" you trust the most.
Silly me, arguing with a muppet.
No shit, Einstein. Now here's a clue for you: Distributing it freely isn't using it legally.
my.mp3.com isn't distributing MP3 files freely. They're distributing them to owners of the IP in question. What, the finer points of the issue escaped you? Which part of the my.mp3.com business model did you miss?
Yeah, viva la revolucion, dude. Death to the infidels, and stuff.
So this is what passes for debate with you. What a privilege to match wits...
Slashdot Mirror
Finally, can you say bloat? XML, DOM, CSS, network stack, etc, etc. -- that's alot of code just to display a random graphic. I guess the footprint of modern web browsers has numbed us to the fact that software is getting slower, not faster.
XML fights bloat by standardizing layers and re-using components. An SVG implementation need not include its own XML parser, network stack, Unicode components, or Core DOM libraries.
Quit bashing SVG. As "design by committee" goes, SVG is a shining example of success. Educate yourself before you go off.
MJP
Amen to that. There should be a Slashdot area for developers.
MJP
Actually, I beat you to the joke. :-)
http://savage.sourceforge.net
MJP
No, that's not right. VML predates the SVG working group; it was available around 1998. They developed it as a vector format with the cooperation of HP, Autodesk, and others with the intention of creating a de facto standard, but submitted it as a primary input for the SVG standard when the SVG working group was formed.
VML sucks, *really* sucks. But that isn't terribly important, since SVG *doesn't* suck. The other major input to the SVG working group was Adobe (et al)'s PGML standard.
As far as I can tell, VML really isn't complete, at least not in terms of implementation. IE 5.5 claims to support VML but
1) Anti-aliasing is not implemented, except for fonts
2) The 1.0 specification is fraught with inconsistencies, errors, and ambiguity
3) The documentation on Microsoft's Web site is dated 1998
4) Microsoft's own VML generation tool supports tags that aren't listed in the 1.0 specification (like ).
MJP
Text works well for vector graphics (see Postscript, CGM, DXF). Text does *not* work well for rasterized graphics, except in limited use (see XPM).
If you think that vector graphics should be stored in binary, then you might as well argue that HTML should have been a binary format. The same principles apply either way.
MJP
The crux of the matter is that the XML parts are all useful and well-designed, whatever the cost. It's worth it to have full XML addressing in the SVG specification, without a doubt. Core XML libraries do most of the work, after all.
The problem is that the SVG DOM is completely head-up-ass insane. It is a nightmare to implement; I've written proof-of-concept handlers in both C++ and Python, and they were both difficult (although the Python version was far easier: 10% code size and 10% time-to-implement by comparison to my C++ version, plus I get free GC).
I suspect that SVG DOM support will be largely ignored by vendors for years to come, and even when it's available it will be wildly variant between implementations. It's a real bitch.
MJP
This is all quite true; XML works well for SVG but the DOM is just plain awful for event-driven development. CSS Events are pathetic, and the DOM is a memory pig.
I'm ambivalent at this point. You can't really afford to do without something like the XML DOM, but it leaves the programmer in the difficult position of having to do a lot of imaginative coding to balance functionality with standards-compliance.
I wish the Core DOM could be redesigned, but it's too late at this point.
MJP
Mises.org response
Reason Online response
Bottom line, the agenda held precious by Borsook and her colleagues is being torn apart by free thinking. They're terrified, and Borsook's awful writing confirms it.
MJPHell, gonna wear my shirt out to the bar tonight, always get into interesting discussions about it.
I must be doing it wrong. Every time I wear my DeCSS t-shirt I get a hundred people asking me how to pronounce DeCSS. I rarely get anything more interesting. What am I doing wrong?
MJP
This is terrific content for a discussion Website, and rather out-of-character for Slashdot (unfortunately). Congratulations are in order for jamie.
MJP
"Government has arrogated so extensive a role to itself that it's understandable that many people might imagine that nothing the government has a hand in could possibly have happened without it." -- Brian Doherty
This was exactly what I thought of when I saw the article posting. For the curious, "Tank Farm Dynamo" can be found in Brin's short story collection entitled "The River of Time".
MJP
I get the last word: Misspelling, dude.
MJP
Short version:
"Please, let's go back into hiding before we actually have to confront our fuckwitted government."
Looking at the span of history, I see the present as the first time free people have ever privately possessed the technology to *stay* free. I don't plan to walk away from that for any reason, lest my children call me a coward and a hypocrite.
MJP
If you are a hobbiest who gets gratification from writing elegant code, use Scheme. Ignore C++.
Hobby, hobbier, hobbiest?
Actually, no. The word is "hobbyist".
MJP
I'm glad to hear about your new book, and for the sake of posterity I'll read it when I can get my hands on it. A more complete single-volume effort would be great news.
I do understand the constraints of your publishing deadline, and in all fairness you did explain them in the class. In the context of the only C++ course I ever got at UTD, however, it was a pretty disappointing situation.
I am probably not the typical C++ student, and I understand that you can't please everybody. I offered my remarks only as a single datum for the feedback forum, not as an unbiased reviewer, myself.
Best wishes,
MJP
As a student of Mr. Heller at the University of Texas at Dallas, taking his C++ course in a CS track, I ended up hating this book.
First of all, this book should be given to people who don't know how to program anything beyond Visual Basic or Perl. It is fundamentally inappropriate for CS majors, especially since it covers -- as the reviewer accurately reports -- about half of the necessary material.
Mr. Heller would have done better introducing the C Programming Language, because he really fails to communicate more than the utter basics of object-oriented programming. It's not that he doesn't mean well, but the fact remains that C++ is a fascinatingly complex language, and anyone who is fooled into thinking that he/she can gain useful C++ skills from this book is in for a sorry time.
I realize that many people will need an extremely gradual introduction to the language, but at the very least Mr. Heller could have finished the job. I mean, really: there's no discussion of how to use polymorphism, no discussion of class modelling, no genuine engagement of inheritance issues, and -- except for the very basics -- almost no discussion of C++ class syntax whatsoever.
I know I'm being hard on the book and on Mr. Heller. It's not out of animosity toward Heller in the least; I just feel that those who are ready to tackle C++ should grab a good introduction -- "Practical C++ Programming" from O'Reilly comes to mind -- and just *learn* the language once and for all. Otherwise you'll end up learning it ten or more times over the next four years (the way I did) piecemeal, in a haphazard fashion that is likely to thoroughly confuse even the most brilliant student.
My apologies to Mr. Heller for harsh remarks; he's a fair and intelligent man, personally.
MJP
Anything that spreads Vacation Bible School files is a good thing, in my book.
MJP
(1) Piracy is stealing.
Wrong. Illegal copying is not "stealing". Go ahead, look it up in the dictionary. There are no illegal copying scenarios that I know of which involve deprivation of property.
(1) makes legitimate (2).
Given (1)'s inherent flaw, by your own admission this puts (2) on shaky ground.
(2) can't be enforced for all.
This isn't about enforcement. This is about the moral status of data copying. If Metallica's copyright infringement was forgivable, you still fail to justify the Napster lawsuit. If it was not forgivable, then we might have a different conversation about this lawsuit.
(3) makes legit (4)
No, it doesn't, lackwit. Metallica copied The Misfits' album, apparently without fear of law-enforcement. I still can't see how the enforcement issue has anything to do with the scale of the transgression. They are independently variable.
and gives an easy attack point for (2), which is covered under law.
This is your conclusion? The justification for law is its ease of use within the context of specific circumstance?
This is simple. If you have any further questions I will write in simple terms.
A real logical argument, rather than this valueless exposition of nonsense, would be a worthwhile follow-up.
MJP
In other words, the recording industry made money in exactly the way everyone else in a free market makes money: by exploiting a scarcity. They took advantage of the scarcity of recorded media and built a massive business around it.
The scarcity is coming to an end, and so is the business model of the recording industry. The question for defenders of the Napster lawsuits is: what makes you think that we should give the recording industry special treatment so that it can sustain its business model?
The copyright laws are explicitly designed to create an artificial scarcity. The justification for this legislation is that it specially motivates artists to create music. I think that the burden of proof is upon those who demand special treatment; do we have a shortage of musicians?
MJP
Using Napster cannot be wrong because it's easy. That is an idiotic justification for law. If it really matters that the scale of bootlegging has changed, has it ever occurred to anyone that it might not be wrong in the first place?
MJP
Moderate this up. Everyone must read this, it is the very crux of the entire issue.
MJP
and when was the last time you found it nessasary to make use of a gun to maintain "the security of a free state?"
That would be the point, genius.
MJP
As a programmer, my natural inclination is to reject management forces. But I have come to the inescapable conclusion that software development is a process that must be managed, on some level, as an organized task.
The current systems and processes for software development are revolting, but that doesn't change the need for organization. Existing processes are modelled on old engineering practices and driven by corporate machines. Open Source development is changing the underlying models, and we can expect processes to follow suit as the effectiveness of OS organizational models gains recognition.
The important thing is to continue developing and practicing those new organizational models, and to continue pointing out those models which bear fruit. The Linux kernel and the Apache Web server stand as monuments to efficiency and quality, and they are destined to be among the forefathers of whatever processes we use in the next decade.
I just hope that we don't reject corporate development models by rejecting the entire notion of organizational intelligence. Next-generation management models would be a giant step forward for software development; management-annihilation would be a step backward.
MJP
Dude, the reason you're getting laughed off the stage is for saying things like "Malda has a massive influence in Open Source" and "his followers blindly, uncritically accept his every word". Gross overstatement is a recipe for disaster in rhetoric. If you want to be an effective speaker, you need to pick up on that.
In this specific case, it's worse than usual. Malda almost never offers editorial opinions. Mostly he offers them on Geeks in Space, where they're offered 1) on demand, and 2) mildly, to say the least. Malda is not a demagogue, plain and simple. He's a personality, mostly for entertainment purposes.
Anyway, your analysis is appreciated, but you're really making a mountain out of a molehill. That's the reason so many people are dismissing you out of hand.
MJP
You sound like a little kid who thinks that they're smarter than their parents and should be able to do whatever they want.
If you wanted an ad hominem debate, why didn't you say so from the beginning? You sound like a desperate curmudgeon who wants to paint the world with a single color to save his preconceived notions.
All of which is beside the point of logical argument, of course. But you already knew that, didn't you?
In case you didn't notice, the law routinely disagrees with your position. And after reading your post in toto, I'm way more inclined to trust their intelligence than yours.
If I hear you right, this issue has nothing to do with what you think for yourself; you're more inclined to leave the thinking to someone else, the one whose "intelligence" you trust the most.
Silly me, arguing with a muppet.
No shit, Einstein. Now here's a clue for you: Distributing it freely isn't using it legally.
my.mp3.com isn't distributing MP3 files freely. They're distributing them to owners of the IP in question. What, the finer points of the issue escaped you? Which part of the my.mp3.com business model did you miss?
Yeah, viva la revolucion, dude. Death to the infidels, and stuff.
So this is what passes for debate with you. What a privilege to match wits...
MJP