Your assumption is that others can enter the market. In the US, in most localities, both the physical phone and cable networks are monopolies, so you only have a single supplier for each. Until the service and the carrier are separated, this will continue to be a problem. Especially when the existing networks were built at taxpayer expense, and new systems would have to be built at cost.
The fair thing to do would be for localities/states/feds to divest the various companies of their physical networks, much as was done with electricity deregulation, which at least levels the playing field for everyone. After all, they were paid for with taxpayer dollars, so it only seems fair that the taxpayer owns them. That'd be us, btw.
To reject any application that can't explain in plain english and 2 sentences (120 words) or less why it is unique and deserving of a patent.
Why this criteria? Because if you have to draw comparisons with other items and state that this application improves incrementally over items 1-n, then it's not innovative and not deserving. Take the pet rock for instance (however trivial and droll):
It's a polished rock with googly eyes, marketed as a "pet". There is nothing like it in existance today.
I'm still not sure it should have a patent, but at least you can explain it in 2 sentences or less, including the all important "unlike anything else" clause. (whether that was true or not is a different issue)
As for funding the patent process:
Make patents holders pay a percentage take to the PTO, paid at least yearly, with a minimum fee of the application itself, increasing by some scale over the years. The older they get, the more expensive they get. Failure to pay on time means it becomes public domain.
I believe such an approach solves several issues, while still allowing invidividuals to profit from their work without undue hardships.
Nope, supercedes is not the word I wanted. It doesn't supercede, it directly contradicts certain aspects of clause 107, and with some byzantine clauses attempts to address certain issues that are part of clause 107, but were brought forward by interested parties.
Whether in the long run the DMCA supercedes the Copyright Act is for judges to decide. As far as I know, the Copyright Act still holds, except when a work has some form of DRM attached.
However, this makes me wonder: If I put a Sony Rootkit CD into my Mac or Linux box, am I violating the DMCA or exercising my 1978 Copyright Act Clause 107 Fair Use rights when I convert the CD to a harddrive losslessly compressed artifact? How about on a Windows machine with Autoplay turned off? How about on a Windows machine with a CD Ripper program that explicitly overrides OS control of the CDROM, thus leaving Autoplay "on", but effectively disabled for the controlled drive?
Questions like this merely highlight how flawed the DMCA is.
Since the inception of copyright you did not have the right to copy a copyrighted work and distribute it without permission. You know, I went about reading the copyright acts after this, and near as I can tell, nothing prohibits you from copying a copyrighted work (the exclusive right to copy appears to be explicitly bound to distribution due to the fair use clause in 107. If the copied work has no effect on the market, there is no infringement. It's that simple.
The DMCA appears to violate clause 107 in several ways, which it attempts to address via some byzantine clauses. Someone more knowledgable than I will probably proffer their counter opinion on this matter.
sounds like a subset of XML combined with XSLT. XSLT will allow you to create just about anything out of your XML. That's why people like XML these days. Now, I'll agree with you that some XML based implementations are not optimum, but you can create crap out of anything.
You write a single loop to dynamically generate 'x' lines of write output. Then you have folks download the saved file! That's how it's done in JS land!!!
Argh, the limits of textual communication. Einstein made a huge splash in the scientific community in 1905. My point was that the GGP was incorrect, physics was a major branch of science, as scientists were working studiously to discover the workings of the world/universe at that time. Whether the public had the slightest concept of this is irrelevant.
I'd say physics still isn't in the public consciousness as big science, although "nukulear" physics sure still seems to be, barely. After all, if it's not slapping Joe Sixpack in the face, it doesn't exist.
...pulps of the 30s and 40s....This was before physics was considered a major branch of science, So Einstein was just working on minor science? (1905 was his big year)
Non-scalable like Yahoo, the world's busiest web site, you mean? Yeah, thought so. Don't get me wrong, PHP has many problems[1]. The biggest of them is probably the ease with which an incompetent moron can create terrible code (a trait it shares with Perl).
I'll still maintain that PHP is not inherently scalable for applications without major hack type gyrations. The persist to DB method discussed here is something that works for low-volume high latency type applications. While PHP may be great at serving semi-stale or static content, high volume HA systems are not in its forte at this time. Even if you could work around those issues, why force a square peg in a round hole when standard relatively easy to use solutions exist? (Look at Weblogic or Resin to see high volume high performance HA capabilities, and you'll understand one reason why Java holds the sway it does in large enterprises)
PHP certainly does not have a monopoly on the ease with which an incompetent moron can create terrible code. Java, C++, C# all have their hats in that ring as well with the slavish devotion to bad patterns witnessed in all of those. I'll never figure out why someone wants to have 400 cookie cutter classes with massive overlap in code (read as cut N paste) for something that could easily be coded in a 100 lines or less in a single class, all for the sake of following a pattern with "type-safety", which is completely bogus as objects are serialized or marshalled across the wire via binary or XML streams. It merely creates 400 points for 400 errors, versus a single point with a single error. I know which system I'd rather work on.
Actually, if you check the reviews of the Mac Pro, you'll see that the FSB 1333 is barely feeding the 4 cores at the moment, depending upon application of course.
AMD's solution scales much better, and combined with NUMA blows Intel out of the water. That's why AMD is rapidly overtaking Intel in the server market, even with supposedly "inferior" CPUs, and every PC manufacturer of note is now selling AMD CPUs.
You'll see the same effect in the client space as applications and OSes become more multi-threaded. Right now, only certain people will see benefits in AMDs CPUs, and those aren't cheap. But, when AMDs quads come out in 2007, combined with a 4X4 board, watch out - that's going to be one smoking machine. The mere fact that the 4X4 architecture rivals Intels best QX processor with yesterday's chips is a good indicator of what's to come. Yes, I'm aware that it's a new chip, but it's merely a faster version of a CPU built on 3 year old technology.
BTW, those new AMD quads actually have more new tech in them than was previously released. It's possible that AMD may get a significant boost out of that alone, and combined with a 65nm process...
How are you going to ask the client, when said DHCP client is one of those nifty routers we all own?
I don't think anyone on/., or even most in the world, directly connect their machines to a network connection anymore. All the broadband connections all go through some sort of router these days, provided by the ISPs themselves.
EJBs, and many other things as well. EJBs violate KISS on many levels, not the least of which is avoidance of boilerplate code. <-- read as "maintenance nightmare"
A lot of the "helpful kits" purport to make your life easier by vendor lockin to a dev kit with tons of boilerplate code that would have to be painfully changed on migrations or even upgrades <-- read this as "non-portability: maintenance nightmare part 2"
Lastly, a single appserver on modern relatively inexpensive hardware can easily serve 5K concurrent people, unless you're doing something computationally intensive for each connection. After all, most web-sites are basically serving up data and pumping it out, most of which is retrieval of said data from a third source (n-tier web design) and merely massaging it for the presentation layer to format.
The new system is actually pretty neat. 1 click expansion of an entire thread below the click point, ability to hide replies at any point in the thread, etc.
I agree about the lack of a feedback mechanism. I guess you just have to email the "overlords" directly. I did. In its initial version, the expansion of a thread was only the posting you clicked, all replies underneath it remained in their current state. I like the new full expansion much better.
I also like the new mod system - select and done. Granted, you need to be sure you want to mod the posting before setting a mod level - there's no second chances with this one.
As for the floating box on the left, I just ignore it, it's not in my way 99.9% of the time, due to how I surf/.
well said. PHP is great for "toy" web sites (by toy, I mean non-scalable SMB at most).
There's lots of things Java offers that are in addition to the language, which are what really won the enterprise space. C/C++/C#/.NET/PHP/Ruby/ROR do not have it. Basically, it's this - you can write relatively complex code quickly, easily, and have it scale while being HA (High Availability) and highly reliable. That doesn't mean dweeb code monkeys can produce such code, however. The quickly and easily terms are highly subjective and only apply to those of appropriate skill, otherwise such things are still "hard", "difficult", or "non-intuitive".
I would think the real problem is that VBA on Office is an excellent vector for hackers. Perhaps removing it from all versions of Office would be a good thing. It would probably help businesses out in the long run as spreadsheet macros are horribly inefficient means for executing business logic. Add to that the auditing issues surrounding laws like Sarbanes Oxley, and macros become unusable anyways.
If you recall, the "saturated" search engine market was saturated with subpar search engines. Everyone everywhere was complaining about the lack of relevant data showing up in searches as the web grew. Google was the first, and still arguably the best, search engine that tackled data retrieval in a different way. They used that to sell ads, which they've done very very very well at.
So it's still luck - many have tried since, and until they can better google by a large percentage, I think Google's probably safe. Mind share will keep them that way.
Hey!!! You must have worked at my last job - I got to see the results of that graduate (ok - he was a college graduate, but just as clueless....)
It's on its 3rd rewrite, because architect-boy "designed" the follow-up based on something about this thing called SOA he read in a magazine, and then added insult to injury by adding in Spring because it "solved" the cyclical redundancy issues in his design. I still haven't gotten over my distaste for Spring because of that.
First off, I'm not arguing your points about how using standards and configuration help your app. Grats that you got something to work the way you want. That doesn't change the fact that HTML, from the perspective of a data description language, is hopelessly broken. Using a subset to achieve your desires does not invalidate that truth. It's very similar to saying I use "English with a couple of symbols" as my data description language.
As to adding span tags, aren't spans merely divs that aren't "broken" (ie - still presentation). Meta data is a different beast, and is XML for the most part.
I'd argue straight XML is more valuable to search engines (HTML is a subset of XML, after all;). Think about how much cruft there is in a standard commercial web page with ads. I could point you to several very popular and known sites that would make you cringe. The funny thing is that the actual dynamic data is relatively miniscule, like 0.0001% of the page. So, with XML/XSLT, you could have a 1-time download of the template, and the remaining data would be 1/1000th of the current page. One site I know of has an alternate interface powered through XML which although verbose, is still 1000th of the data traffic that would otherwise be served. Note: I'm assuming that the XML data sent is only the actual data/meta data required for content, the XSLT handles all aspects of the presentation layer using that content.
As for Javascript, it doesn't even need to change. XML/XLST transforms to a nifty neat HTML DOM in your browser that you can walk with JS. Better yet, with a proper XML design, you're now setup to do AJAX where ever you feel like, and it will perform. Oh, and if you want to hide the XML underpinnings, perform the XSLT transform at your server, although that defeats the efficiency aspect.
I stopped counting. 3 tiers? I wish. There's at least 5 layers just in the JSP/servlet/XML -> HTML portion of an app. Add in the multiple layers of appservers, DBs, message buses that go to who knows where, etc, and "n-tier" becomes more than just a description implying complexity, it's an accurate description.
Oh, and don't forget, somehow it's the application programmer's responsibility to debug the DB, the queues, the network, the cross browser issues, and not to forget why CSS isn't making that image move 1 pixel to the right when javascript changes the underlying DOM.
What kills me is that a lot of businesses think they can hire someone for $40 hour to do all that, do it well, and make it scalable. Oh, and do it in multiple languages and platforms. Some job listings are worth the read for a good laugh.
having worked with tables and divs and CSS for a while as a programmer, not a web-designer, here's my take on why programmers almost always use tables: making formatted tabular layouts with divs and CSS is non-intuitive and a major pain in the ass for just one browser, much less a slew of them.
Why are div and CSS so damn difficult to use to make a row/column layout? Wait, could it be that that's the forte of tables? I want to divide my page into rectangular areas. The answer? Make divs as easy to use as tables for presentation.
But, here's the real root of the problem: HTML is not a data description language, it's a presentation language, and it's been misused since its inception. If you want true separation, use XML for the data, and XSLT to transform it into HTML. If HTML were to truly be a data representation, you'd only have and
tags, and the rest would be CSS. Think about it a while. Everyone of those other tags are some variation related to presentation. (Yes, it's oversimplification, but it's not that big a stretch)
Slashdot Mirror
Your assumption is that others can enter the market. In the US, in most localities, both the physical phone and cable networks are monopolies, so you only have a single supplier for each. Until the service and the carrier are separated, this will continue to be a problem. Especially when the existing networks were built at taxpayer expense, and new systems would have to be built at cost.
The fair thing to do would be for localities/states/feds to divest the various companies of their physical networks, much as was done with electricity deregulation, which at least levels the playing field for everyone. After all, they were paid for with taxpayer dollars, so it only seems fair that the taxpayer owns them. That'd be us, btw.
To reject any application that can't explain in plain english and 2 sentences (120 words) or less why it is unique and deserving of a patent.
Why this criteria? Because if you have to draw comparisons with other items and state that this application improves incrementally over items 1-n, then it's not innovative and not deserving. Take the pet rock for instance (however trivial and droll):
It's a polished rock with googly eyes, marketed as a "pet". There is nothing like it in existance today.
I'm still not sure it should have a patent, but at least you can explain it in 2 sentences or less, including the all important "unlike anything else" clause. (whether that was true or not is a different issue)
As for funding the patent process:
Make patents holders pay a percentage take to the PTO, paid at least yearly, with a minimum fee of the application itself, increasing by some scale over the years. The older they get, the more expensive they get. Failure to pay on time means it becomes public domain.
I believe such an approach solves several issues, while still allowing invidividuals to profit from their work without undue hardships.
Mail (Mac) and Thunderbird have both done this for years, as do several other programs. No reason to run any specific OS at all.
Nope, supercedes is not the word I wanted. It doesn't supercede, it directly contradicts certain aspects of clause 107, and with some byzantine clauses attempts to address certain issues that are part of clause 107, but were brought forward by interested parties.
Whether in the long run the DMCA supercedes the Copyright Act is for judges to decide. As far as I know, the Copyright Act still holds, except when a work has some form of DRM attached.
However, this makes me wonder: If I put a Sony Rootkit CD into my Mac or Linux box, am I violating the DMCA or exercising my 1978 Copyright Act Clause 107 Fair Use rights when I convert the CD to a harddrive losslessly compressed artifact? How about on a Windows machine with Autoplay turned off? How about on a Windows machine with a CD Ripper program that explicitly overrides OS control of the CDROM, thus leaving Autoplay "on", but effectively disabled for the controlled drive?
Questions like this merely highlight how flawed the DMCA is.
The DMCA appears to violate clause 107 in several ways, which it attempts to address via some byzantine clauses. Someone more knowledgable than I will probably proffer their counter opinion on this matter.
sounds like a subset of XML combined with XSLT. XSLT will allow you to create just about anything out of your XML. That's why people like XML these days. Now, I'll agree with you that some XML based implementations are not optimum, but you can create crap out of anything.
No no no!!!!
You write a single loop to dynamically generate 'x' lines of write output. Then you have folks download the saved file! That's how it's done in JS land!!!
note that the readers of pulp SF are only an insignificantly small subset of the general public.
Argh, the limits of textual communication. Einstein made a huge splash in the scientific community in 1905. My point was that the GGP was incorrect, physics was a major branch of science, as scientists were working studiously to discover the workings of the world/universe at that time. Whether the public had the slightest concept of this is irrelevant.
I'd say physics still isn't in the public consciousness as big science, although "nukulear" physics sure still seems to be, barely. After all, if it's not slapping Joe Sixpack in the face, it doesn't exist.
Light travels about 310 times C in a cesium vapor.
...pulps of the 30s and 40s.I'll still maintain that PHP is not inherently scalable for applications without major hack type gyrations. The persist to DB method discussed here is something that works for low-volume high latency type applications. While PHP may be great at serving semi-stale or static content, high volume HA systems are not in its forte at this time. Even if you could work around those issues, why force a square peg in a round hole when standard relatively easy to use solutions exist? (Look at Weblogic or Resin to see high volume high performance HA capabilities, and you'll understand one reason why Java holds the sway it does in large enterprises)
PHP certainly does not have a monopoly on the ease with which an incompetent moron can create terrible code. Java, C++, C# all have their hats in that ring as well with the slavish devotion to bad patterns witnessed in all of those. I'll never figure out why someone wants to have 400 cookie cutter classes with massive overlap in code (read as cut N paste) for something that could easily be coded in a 100 lines or less in a single class, all for the sake of following a pattern with "type-safety", which is completely bogus as objects are serialized or marshalled across the wire via binary or XML streams. It merely creates 400 points for 400 errors, versus a single point with a single error. I know which system I'd rather work on.
Actually, if you check the reviews of the Mac Pro, you'll see that the FSB 1333 is barely feeding the 4 cores at the moment, depending upon application of course.
AMD's solution scales much better, and combined with NUMA blows Intel out of the water. That's why AMD is rapidly overtaking Intel in the server market, even with supposedly "inferior" CPUs, and every PC manufacturer of note is now selling AMD CPUs.
You'll see the same effect in the client space as applications and OSes become more multi-threaded. Right now, only certain people will see benefits in AMDs CPUs, and those aren't cheap. But, when AMDs quads come out in 2007, combined with a 4X4 board, watch out - that's going to be one smoking machine. The mere fact that the 4X4 architecture rivals Intels best QX processor with yesterday's chips is a good indicator of what's to come. Yes, I'm aware that it's a new chip, but it's merely a faster version of a CPU built on 3 year old technology.
BTW, those new AMD quads actually have more new tech in them than was previously released. It's possible that AMD may get a significant boost out of that alone, and combined with a 65nm process...
How are you going to ask the client, when said DHCP client is one of those nifty routers we all own?
/., or even most in the world, directly connect their machines to a network connection anymore. All the broadband connections all go through some sort of router these days, provided by the ISPs themselves.
I don't think anyone on
EJBs, and many other things as well. EJBs violate KISS on many levels, not the least of which is avoidance of boilerplate code. <-- read as "maintenance nightmare"
A lot of the "helpful kits" purport to make your life easier by vendor lockin to a dev kit with tons of boilerplate code that would have to be painfully changed on migrations or even upgrades <-- read this as "non-portability: maintenance nightmare part 2"
Lastly, a single appserver on modern relatively inexpensive hardware can easily serve 5K concurrent people, unless you're doing something computationally intensive for each connection. After all, most web-sites are basically serving up data and pumping it out, most of which is retrieval of said data from a third source (n-tier web design) and merely massaging it for the presentation layer to format.
The new system is actually pretty neat. 1 click expansion of an entire thread below the click point, ability to hide replies at any point in the thread, etc.
/.
I agree about the lack of a feedback mechanism. I guess you just have to email the "overlords" directly. I did. In its initial version, the expansion of a thread was only the posting you clicked, all replies underneath it remained in their current state. I like the new full expansion much better.
I also like the new mod system - select and done. Granted, you need to be sure you want to mod the posting before setting a mod level - there's no second chances with this one.
As for the floating box on the left, I just ignore it, it's not in my way 99.9% of the time, due to how I surf
well said. PHP is great for "toy" web sites (by toy, I mean non-scalable SMB at most).
There's lots of things Java offers that are in addition to the language, which are what really won the enterprise space. C/C++/C#/.NET/PHP/Ruby/ROR do not have it. Basically, it's this - you can write relatively complex code quickly, easily, and have it scale while being HA (High Availability) and highly reliable. That doesn't mean dweeb code monkeys can produce such code, however. The quickly and easily terms are highly subjective and only apply to those of appropriate skill, otherwise such things are still "hard", "difficult", or "non-intuitive".
I wish I'd had mod points, you deserve a few.
I would think the real problem is that VBA on Office is an excellent vector for hackers. Perhaps removing it from all versions of Office would be a good thing. It would probably help businesses out in the long run as spreadsheet macros are horribly inefficient means for executing business logic. Add to that the auditing issues surrounding laws like Sarbanes Oxley, and macros become unusable anyways.
If you recall, the "saturated" search engine market was saturated with subpar search engines. Everyone everywhere was complaining about the lack of relevant data showing up in searches as the web grew. Google was the first, and still arguably the best, search engine that tackled data retrieval in a different way. They used that to sell ads, which they've done very very very well at.
So it's still luck - many have tried since, and until they can better google by a large percentage, I think Google's probably safe. Mind share will keep them that way.
Hey!!! You must have worked at my last job - I got to see the results of that graduate (ok - he was a college graduate, but just as clueless....)
It's on its 3rd rewrite, because architect-boy "designed" the follow-up based on something about this thing called SOA he read in a magazine, and then added insult to injury by adding in Spring because it "solved" the cyclical redundancy issues in his design. I still haven't gotten over my distaste for Spring because of that.
First off, I'm not arguing your points about how using standards and configuration help your app. Grats that you got something to work the way you want. That doesn't change the fact that HTML, from the perspective of a data description language, is hopelessly broken. Using a subset to achieve your desires does not invalidate that truth. It's very similar to saying I use "English with a couple of symbols" as my data description language.
As to adding span tags, aren't spans merely divs that aren't "broken" (ie - still presentation). Meta data is a different beast, and is XML for the most part.
I'd argue straight XML is more valuable to search engines (HTML is a subset of XML, after all;). Think about how much cruft there is in a standard commercial web page with ads. I could point you to several very popular and known sites that would make you cringe. The funny thing is that the actual dynamic data is relatively miniscule, like 0.0001% of the page. So, with XML/XSLT, you could have a 1-time download of the template, and the remaining data would be 1/1000th of the current page. One site I know of has an alternate interface powered through XML which although verbose, is still 1000th of the data traffic that would otherwise be served. Note: I'm assuming that the XML data sent is only the actual data/meta data required for content, the XSLT handles all aspects of the presentation layer using that content.
As for Javascript, it doesn't even need to change. XML/XLST transforms to a nifty neat HTML DOM in your browser that you can walk with JS. Better yet, with a proper XML design, you're now setup to do AJAX where ever you feel like, and it will perform. Oh, and if you want to hide the XML underpinnings, perform the XSLT transform at your server, although that defeats the efficiency aspect.
I stopped counting. 3 tiers? I wish. There's at least 5 layers just in the JSP/servlet/XML -> HTML portion of an app. Add in the multiple layers of appservers, DBs, message buses that go to who knows where, etc, and "n-tier" becomes more than just a description implying complexity, it's an accurate description.
Oh, and don't forget, somehow it's the application programmer's responsibility to debug the DB, the queues, the network, the cross browser issues, and not to forget why CSS isn't making that image move 1 pixel to the right when javascript changes the underlying DOM.
What kills me is that a lot of businesses think they can hire someone for $40 hour to do all that, do it well, and make it scalable. Oh, and do it in multiple languages and platforms. Some job listings are worth the read for a good laugh.
Let me say that my anchor and div tags were removed... whoopsie.:)
Those would be the only 2 tags in a data representation, and div tags could even be considered a stretch.
having worked with tables and divs and CSS for a while as a programmer, not a web-designer, here's my take on why programmers almost always use tables: making formatted tabular layouts with divs and CSS is non-intuitive and a major pain in the ass for just one browser, much less a slew of them.
Why are div and CSS so damn difficult to use to make a row/column layout? Wait, could it be that that's the forte of tables? I want to divide my page into rectangular areas. The answer? Make divs as easy to use as tables for presentation.
But, here's the real root of the problem: HTML is not a data description language, it's a presentation language, and it's been misused since its inception. If you want true separation, use XML for the data, and XSLT to transform it into HTML. If HTML were to truly be a data representation, you'd only have and tags, and the rest would be CSS. Think about it a while. Everyone of those other tags are some variation related to presentation. (Yes, it's oversimplification, but it's not that big a stretch)