Where I live, we will refer to a burger shaped piece of potato that is deep fried as a "potato scallop" whereas our New Zealand neighbors (and some people in the south of Australia) will refer to one of them as a "potato cake".
In an NZ fish and chip shop which is usually the only place you get such things, that would be called a "potato fritter" if it is a whole slice of potato. A NZer would guess that "scallop" means either a slice of something (usually potato) or the shellfish depending on context. "Potato cake" is uncommon, but to a Kiwi probably conjures up something like a fish cake but with mashed or grated potato instead of fish (most commercial fish cakes are mostly just potato anyway).
Typically a fried disc of anything in a fish and chip shop is called a fritter - although a fish cake is the main exception to that.
Calling a filled roll "a sandwich" is creeping in slightly though, but it hasn't spread to burgers yet as far as I can tell. Subway TV ads direct from the US are probably to blame for that, even though Subway locally doesn't really use the word sandwich much at all.
Phew, such a long post about nothing of any importance:)
Only on slashdot could a response to a story about patenting name to gender associations for avatars have a response featuring ridicule of people's poor naming choices and a commentary on said IP law be moderated "Off topic".
I suspect (correct me if I'm assuming too much) the real reason for the original "Off topic" mod was that you posted this comment to a story about Linux hardware support rather than the story about IBMs patent.
If you want me to go on arguing, you'll have to pay for another five minutes.
If Kerberos was GPL it would not have stopped Microsoft from reimplementing it clean-room and THEN embracing and extending it, or perhaps even worse, using some other authentication system which you would have absolutely NO starting point to derive the differences from (at least MS Kerberos is a well known weirdness and you could patch the standard Kerberos server to support it.
Not only that, the Kerberos RFCs allowed for vendor extensions like the PAC. The PAC didn't break anything MIT Kerberos already did - it was just added an extra bit of info for Windows specific stuff. The only crappy part was MS not documenting how it worked until much later on, temporarily preventing other systems making use of it.
Anyone that paints this as a GPL is better than BSD/MIT issue is either ignorant or trolling. The Kerberos protocols are open standards defined in RFCs, and MS would've largely implemented their codebase from scratch even with the MIT licenses. As you say the only thing a GPLed Kerberos would've achieved is making MS much more likely to have rolled their own completely proprietary solution instead and Samba 4 might've never been possible due to the increased complexity involved.
I've never seen a "enterprise class" server-side Java app that didn't scale except with lots and lots of servers.
Isn't that kinda the definition of "scale"? Maybe by "scale" you really meant "perform well"? In which case I'd agree with you.
Scaling well doesn't necessarily mean something performs well, rather it means performance keeps increasing as you add servers rather than leveling off.
Could someone explain where the cost comes from? Why does using more of my bandwidth cost comcast extra money?
It isn't your bandwidth, or more explicitly it isn't the amount of bandwidth dedicated to you. It's just the upper limit on the rate at which you can use a portion of the ISPs dedicated bandwidth.
An ISP has a finite capacity (uplinks, routers etc etc). Increasing that capacity costs serious money - ie way way more than what you pay the ISP for it. The business model of an ISP is to pay serious money for a chunk of dedicated capacity and spread that capacity and cost over their customer base on the assumption that the customers will each only use a small percentage of the capacity they are capable of.
With flat rate plans, what that small percentage ends up being directly affects the entire business model and even small increases in it could make the difference between being a profitable ISP and an unprofitable one. The low usage customers are directly subsidizing the high usage ones.
If I buy a firewire cable, it costs the same whether I transfer 80 MB or 80GB using it, and it doesn't wear out with use.
And if you lease a dedicated line from a telco to connect you to a backbone provider you'll have a situation like your firewire cable. You might want to try comparing the costs of that to a DSL or cable connection.
But an ISP connection is more like a shared bus (eg USB) with lots of other devices on it - you need to share the total bandwidth with all the other devices and the bus can either have lots of devices using a little bit of bandwidth or only a few devices using lots of bandwidth. If the devices start wanting more bandwidth, extra buses will need to be added to the system (at a cost) to keep each device working at the same speed it used to by having less devices on each bus.
To qualify your point, from the birth of home 3D acceleration until about 1999, OpenGL was the de facto standard for 3D games that supported hardware acceleration, with few notable exceptions.
Really? You're not thinking of Glide (the 3Dfx API) are you? I seem to remember there were far more Glide games than OpenGL ones.
From memory, even GL Quake depended on Glide via a OpenGL -> Glide wrapper for 3D hardware acceleration.
OK you've pointed out the difference between xVM and XenServer from Citrix. But what are the difference between xVM and the open source Xen project? Is it just the continuation of the old OpenSolaris port of Xen under a new name?
Because the differences you mentioned seem (to me at least) to mostly be roughly the same differences between XenServer and open source Xen anyway. What has Sun added to xVM that makes it better than the open source Xen?
Note: I'm just curious, I'm familiar with Xen but not xVM.
I think the posters point was that Google charging advertisers for ad views that are never shown is fraudulent. Their point wasn't that blocking ads is illegal.
Maybe it would all make sense if you read the article rather than the summary.
A very rough vague summary:
This guy invented a portable digital music player 30 yrs ago. Apple gets sued by some other company who reckons the iPod infringes on their patent, and for their defense Apple hires this guy as a consultant to show prior art etc. The original guy gets some money for his efforts and is pleased his original invention is now recognised. Everyone is happy - well except for the other company and their patent lawyers.
As another OpenBSD fan, there are still things Linux does better IMO.
Linux still performs better - especially on todays multicore systems.
As secure as the core of OpenBSD is, it is only the core systems security that is looked after by the OpenBSD team. Unless things have changed recently (corrections welcome), security updates for 3rd party apps you've installed are your responsibility with OpenBSD.
Rebuilding for security patches is tedious compared to letting the package manager just download binary updates.
Um, fires can pretty much bring down every building without some sort of suppression.
Yep, fire is structurally damaging to just about anything (steel especially).
The sole exception is concrete, which can leave a hollow shell.
Reinforced concrete is also somewhat vulnerable in a serious fire - especially when the reinforcing is relatively close to the surface. The heat can cause the concrete to spall off beams exposing the reinforcing from residual water in the concrete turning into steam. The expansion of the steel in the heat causes tensile stress on the surrounding concrete and concrete isn't very good under tension.
The ONLY thing anyone needs to know? Most of the internet relies on the security of DNS - does that mean that is the ONLY thing anyone has to know about anything else on the internet as well?
I have fiddled around with Windows/Linux integration for central authentication and found that the only alternative TODAY that works acceptable is to use the "Windows Services for Unix [microsoft.com]" (SFU) add-on for Windows Server. And you can download that from Microsoft.
Just an update - SFU is now built into Windows 2003 R2 and Windows 2008. And the AD schema extensions now use the standard RFC2307 attributes rather than the SFU specific ones.
You are jumping to conclusions about what I said and missing the point. I was just pointing out the similarity between that and the same rhetorical device used in another situation as a divide and conquer strategy.
A quote from the top of your link:
The terms, CSS and CSS-P (or full css) are two (2) very different things in this article. The term "CSS" is more of the regular CSS used for font styles and the like. The term, "CSS-P", is specifically referring to use of <div> tags to replace <table> tags. This article challenges CSS-P, not CSS.
That is the vocal table layout defender I referred to, and their use of an artifical distinction.
The rest of the page reads just like many other sites on all kinds of things created by crackpots with a chip on their shoulder. It's full of straw-man arguments and misunderstandings of what they are arguing against.
That you describe it as exploding some of the myths about tables vs CSS-P is telling.
I am fairly neutral about using tables for layout (although I draw the line at nested tables), they aren't as elegant but can be a pragmatic choice when you need to implement certain designs while still supporting IE.
I have no religious position on this topic (besides calling for a fatwa against IE - joking).
But CSS is just CSS. The CSS-P distinction was made up so that people who cling to tables for layout and get defensive about that don't have to look like idiots by saying CSS as a whole is wrong.
If CSS-P was an actual defined subset of CSS - where is the breakdown of which parts are CSS and which parts are CSS-P?
I love the way vocal table layout defenders invented an artificial "CSS" vs "CSS-P" distinction. It reminds me of the artificial macro/micro evolution distinction that evolution disbelievers insist upon.
In both cases the goal is to separate out the absolutely undeniable part of something so they can better fight against the rest.
If thats the O'Reilly book I'm thinking of (The Definitive Guide by Eric Meyer?), that's because it was a reference book not a beginners tutorial book.
Trying to learn anything from a reference book is hard and confusing (I can sympathise because I originally learnt CSS 1 & 2 from the W3C specs), but that doesn't diminish the important value of a good reference book. That book was for a long time (and may still be) the only useful CSS book for people that had already picked up the basics and needed the nitty gritty details.
Now the market is completely swamped by beginner level CSS books.
Soupiness aside, DIVs make more sense, although I hate it when people claim that DIVs are 'semantic'.
Table elements have semantic meaning, DIVs do not.
The people you were quoting were probably just muppets repeating something they didn't really understand, but it is possible to make some logical sense out of those kinds of statements.
so playing around with it a bit...
A div is generally defined as a block element with no inherent semantic meaning.
So if a div used for layout has a semantic value of zero, and a table element used for layout is semantically wrong (ie has a negative semantic value), then you could say that the div is 'more semantic' than the table element;)
Likewise if some other elements do have row and column like relationships with each other, the table elements semantic value is now positive and the table becomes 'more semantic' than divs or spans used in the same situation.
Slashdot Mirror
In an NZ fish and chip shop which is usually the only place you get such things, that would be called a "potato fritter" if it is a whole slice of potato. A NZer would guess that "scallop" means either a slice of something (usually potato) or the shellfish depending on context. "Potato cake" is uncommon, but to a Kiwi probably conjures up something like a fish cake but with mashed or grated potato instead of fish (most commercial fish cakes are mostly just potato anyway).
Typically a fried disc of anything in a fish and chip shop is called a fritter - although a fish cake is the main exception to that.
Calling a filled roll "a sandwich" is creeping in slightly though, but it hasn't spread to burgers yet as far as I can tell. Subway TV ads direct from the US are probably to blame for that, even though Subway locally doesn't really use the word sandwich much at all.
Phew, such a long post about nothing of any importance :)
Probably both.
Draughts? No wonder daylight saving has a checkered history!
Or maybe some of that foamy tape stuff in his door/window frames would help fix the problem?
I suspect (correct me if I'm assuming too much) the real reason for the original "Off topic" mod was that you posted this comment to a story about Linux hardware support rather than the story about IBMs patent.
If you want me to go on arguing, you'll have to pay for another five minutes.
You look lost - don't be afraid to ask for directions. I think the patent story is two blocks that way...
Not only that, the Kerberos RFCs allowed for vendor extensions like the PAC. The PAC didn't break anything MIT Kerberos already did - it was just added an extra bit of info for Windows specific stuff. The only crappy part was MS not documenting how it worked until much later on, temporarily preventing other systems making use of it.
Anyone that paints this as a GPL is better than BSD/MIT issue is either ignorant or trolling. The Kerberos protocols are open standards defined in RFCs, and MS would've largely implemented their codebase from scratch even with the MIT licenses. As you say the only thing a GPLed Kerberos would've achieved is making MS much more likely to have rolled their own completely proprietary solution instead and Samba 4 might've never been possible due to the increased complexity involved.
Isn't that kinda the definition of "scale"? Maybe by "scale" you really meant "perform well"? In which case I'd agree with you.
Scaling well doesn't necessarily mean something performs well, rather it means performance keeps increasing as you add servers rather than leveling off.
Ummm is it really that difficult to understand?
It isn't your bandwidth, or more explicitly it isn't the amount of bandwidth dedicated to you. It's just the upper limit on the rate at which you can use a portion of the ISPs dedicated bandwidth.
An ISP has a finite capacity (uplinks, routers etc etc). Increasing that capacity costs serious money - ie way way more than what you pay the ISP for it. The business model of an ISP is to pay serious money for a chunk of dedicated capacity and spread that capacity and cost over their customer base on the assumption that the customers will each only use a small percentage of the capacity they are capable of.
With flat rate plans, what that small percentage ends up being directly affects the entire business model and even small increases in it could make the difference between being a profitable ISP and an unprofitable one. The low usage customers are directly subsidizing the high usage ones.
And if you lease a dedicated line from a telco to connect you to a backbone provider you'll have a situation like your firewire cable. You might want to try comparing the costs of that to a DSL or cable connection.
But an ISP connection is more like a shared bus (eg USB) with lots of other devices on it - you need to share the total bandwidth with all the other devices and the bus can either have lots of devices using a little bit of bandwidth or only a few devices using lots of bandwidth. If the devices start wanting more bandwidth, extra buses will need to be added to the system (at a cost) to keep each device working at the same speed it used to by having less devices on each bus.
Really? You're not thinking of Glide (the 3Dfx API) are you? I seem to remember there were far more Glide games than OpenGL ones.
From memory, even GL Quake depended on Glide via a OpenGL -> Glide wrapper for 3D hardware acceleration.
OK you've pointed out the difference between xVM and XenServer from Citrix. But what are the difference between xVM and the open source Xen project? Is it just the continuation of the old OpenSolaris port of Xen under a new name?
Because the differences you mentioned seem (to me at least) to mostly be roughly the same differences between XenServer and open source Xen anyway. What has Sun added to xVM that makes it better than the open source Xen?
Note: I'm just curious, I'm familiar with Xen but not xVM.
Read it again.
I think the posters point was that Google charging advertisers for ad views that are never shown is fraudulent. Their point wasn't that blocking ads is illegal.
Maybe it would all make sense if you read the article rather than the summary.
A very rough vague summary:
This guy invented a portable digital music player 30 yrs ago. Apple gets sued by some other company who reckons the iPod infringes on their patent, and for their defense Apple hires this guy as a consultant to show prior art etc. The original guy gets some money for his efforts and is pleased his original invention is now recognised. Everyone is happy - well except for the other company and their patent lawyers.
Yeah I got it, but... _Monty_ ? shouldn't that be __Monty__(self) ?
As another OpenBSD fan, there are still things Linux does better IMO.
Linux still performs better - especially on todays multicore systems.
As secure as the core of OpenBSD is, it is only the core systems security that is looked after by the OpenBSD team. Unless things have changed recently (corrections welcome), security updates for 3rd party apps you've installed are your responsibility with OpenBSD.
Rebuilding for security patches is tedious compared to letting the package manager just download binary updates.
Doh - that meant to read something more like:
"The heat can cause residual water in the concrete turning into steam causing the concrete to spall off beams exposing the reinforcing."
Yep, fire is structurally damaging to just about anything (steel especially).
Reinforced concrete is also somewhat vulnerable in a serious fire - especially when the reinforcing is relatively close to the surface. The heat can cause the concrete to spall off beams exposing the reinforcing from residual water in the concrete turning into steam. The expansion of the steel in the heat causes tensile stress on the surrounding concrete and concrete isn't very good under tension.
What if it gets cooler then warmer?
The ONLY thing anyone needs to know? Most of the internet relies on the security of DNS - does that mean that is the ONLY thing anyone has to know about anything else on the internet as well?
Thank you.
The level of ignorance about what OpenID is or isn't is fairly staggering even amongst technical people.
I think the OpenID people have an uphill battle trying to educate the masses. I hope they can succeed, but I have my doubts.
Just an update - SFU is now built into Windows 2003 R2 and Windows 2008. And the AD schema extensions now use the standard RFC2307 attributes rather than the SFU specific ones.
You are jumping to conclusions about what I said and missing the point. I was just pointing out the similarity between that and the same rhetorical device used in another situation as a divide and conquer strategy.
A quote from the top of your link:
That is the vocal table layout defender I referred to, and their use of an artifical distinction.
The rest of the page reads just like many other sites on all kinds of things created by crackpots with a chip on their shoulder. It's full of straw-man arguments and misunderstandings of what they are arguing against.
That you describe it as exploding some of the myths about tables vs CSS-P is telling.
I am fairly neutral about using tables for layout (although I draw the line at nested tables), they aren't as elegant but can be a pragmatic choice when you need to implement certain designs while still supporting IE.
I have no religious position on this topic (besides calling for a fatwa against IE - joking).
But CSS is just CSS. The CSS-P distinction was made up so that people who cling to tables for layout and get defensive about that don't have to look like idiots by saying CSS as a whole is wrong.
If CSS-P was an actual defined subset of CSS - where is the breakdown of which parts are CSS and which parts are CSS-P?
I love the way vocal table layout defenders invented an artificial "CSS" vs "CSS-P" distinction. It reminds me of the artificial macro/micro evolution distinction that evolution disbelievers insist upon.
In both cases the goal is to separate out the absolutely undeniable part of something so they can better fight against the rest.
If thats the O'Reilly book I'm thinking of (The Definitive Guide by Eric Meyer?), that's because it was a reference book not a beginners tutorial book.
Trying to learn anything from a reference book is hard and confusing (I can sympathise because I originally learnt CSS 1 & 2 from the W3C specs), but that doesn't diminish the important value of a good reference book. That book was for a long time (and may still be) the only useful CSS book for people that had already picked up the basics and needed the nitty gritty details.
Now the market is completely swamped by beginner level CSS books.
The people you were quoting were probably just muppets repeating something they didn't really understand, but it is possible to make some logical sense out of those kinds of statements.
so playing around with it a bit...
A div is generally defined as a block element with no inherent semantic meaning.
So if a div used for layout has a semantic value of zero, and a table element used for layout is semantically wrong (ie has a negative semantic value), then you could say that the div is 'more semantic' than the table element ;)
Likewise if some other elements do have row and column like relationships with each other, the table elements semantic value is now positive and the table becomes 'more semantic' than divs or spans used in the same situation.
OK, it was a bit of a stretch :)
Your friends and family wouldn't want it that way though.