Im not talking about work. I'm talking about people taking a car to go buy milk one block away. Or for those who do use public transportation, use their car to get to the bus stop thats just across the block. The people who take their car to go to the gym thats 10 minutes away (come on...you're being lazy on your way to go exercise, the hell?)
Some people don't have a choice: but if the people who DO have a choice, made a different one, I feel a lot of fuel would be saved.
Being able to swap message format on the fly is really the way to go, since they all have advantages and issues, from speed, backward compatibility, debugging, security constraints, etc.
Thats one of the things I liked in messaging frameworks such as WCF (though its far from being the first to do it). Code to the API, the message protocol can be decided later, and changed/plugged in on the fly when something better comes.
I'd be interested in a study showing how much fuel we'd save if anytime people can get to somewhere on foot in 10 minutes, they'd walk instead of taking their car.
Just for kicks, I'd like to see those numbers.
Re:Redundent -- read selectively.
on
Head First C#
·
· Score: 1
Not quite. That "lingo" is the same that is part of the theoritical OOP model, and is also adopted in other more theoritical concepts such as UML representation (I guess that is still a tool, but yeah).
The difference is there because there is a large difference, both theoritical and in practice, between a fully abstract class and an interface, and it is important to keep that difference at the conceptual level. The "C++ way" exists because C++ is not a full implementation of the OOP model (neither is C# of course, but it is closer).
One very important time where that "theory affects reality to the point you really have to care" is when you start versionning your interfaces and classes. Long story short, you shouldn't, in general, expose an interface in a public API (well, the interface can be exposed, but I mean, a public method that will be used by a different consumer should not return an interface type, but only an implementation of it. So often, it is simply better to use fully abstract classes, since they are both interface an "implementation", so to speak, and will not suffer from the interface problem, especially if the abstract class realize/implement the interface, in which case you kill 2 birds with one stone).
The tool didn't take over the concept, since the concept predates the tools by decades in some cases:) (think of it in a similar way you would think about relational algebra vs a database).
Its a problem in the protocol. So the only systems that would not be vulnerable are those that did -not- follow the specs. Guess Windows is safe, since Microsoft never follows the specs:)
And really, all that is one thing: Web Development with the.NET/SQL Server stack. C#/HTML/CSS/Javascript/XML is basically ASP.NET, WPF is just a small piece of.NET 3.0, MSBuild is what runs when you hit F5 (and is incredibly simple compared to, let say, NAnt), SOAP/WS* is also part of ASP.NET, with a LITTLE added challenge of WSDL if you guys actually write those (and that fits with XML, since 90% of the learning curve of WSDL is knowing XSDs).
Thats a far cry from actually needing to know different echosystems or anything (though I'm not sure, was that actually your point?)
Re:How come nobody uses anonymous delegates?
on
Head First C#
·
· Score: 1
Good.NET developers will use delegates (not just anonymous ones) all over the place, as it is a core part of the framework, and is one of the best way to implement a large amount of the base design patterns (like Strategy and Observer) in.NET.
The thing is, there's a grand total of 7 good.NET developers in the world. If that.
.NET forms can be embedded as dll/activex in internet explorer if explicitely given appropriate permissions. Its kindda cute as a deployement platform for internal applications I suppose.
Personally, I prefer using ClickOnce or now Windows Presentation Foundation for that... easier, safer, more versatile, and better supported:)
Re:Redundent -- read selectively.
on
Head First C#
·
· Score: 1
The problem was in the context. If they're asking if there is multiple inheritance in C#, it is because they understand that it is an area where there can be confusion in the concepts of the language. Basically, it is a semantic question. And semantically, while a class, in the OOP and/or UML models, can be both "inherited" and "realized", an interface can only ever be "realized", and cannot be "inherited". Thus, saying that you can have multiple inheritance in C# is wrong.
Of course, in practice, as you did, you can just consider realization to be a public interface/behavior inheritance, but something like that holds no ground in a question about the semantic details of a language's implementation.
Re:Redundent -- read selectively.
on
Head First C#
·
· Score: 5, Interesting
I agree, but at the same time, you have to be careful. Different languages have different takes at things someone may think as "obvious". A while ago I went to an interview for a.NET dev job, and the people interviewing me were very obviously all former C++ programmers recycled in C#. One of the questions they asked me was: "How do you do multiple inheritance in C#".
My answer was the obvious: "There isn't any, you can only inherit from one class". They started arguing with me about how it WAS possible, and actually common, to do multiple inheritance in C#. After a bit of discussion, I realised that they were still thinking in C++ term, where an interface is nothing more than a fully abstract class, so even if you only use "interfaces", you ARE doing multiple inheritance (while in C#, there are -core- differences between a fully abstract class and an interface). They had skipped that part of the book thinking they didn't need it:)
Super users being able to access any account can still be done through session or other server side mechanism:) The product we worked on at my previous job worked like that, and it went quite well too:)
Usually a big company will run through a caching proxy... so you'll see one user, hit all of your cachable content exactly one time. So unless your site is more of a web application than a site, you shouldn't see very high traffic coming from corporations.
Well, according to data from 2007, tokyo's population density was of 4700 person per square km, vs 2050 for new york. Los Angele, the closest USA city on the list stands at about 2700. Nowhere close. Plus, if you go by population density of the country (since federal governments have a hand in pushing these things along), the US have an insignificant population density, while in Japan you'd have to cater to less total surface. Big difference.
In many asian countries, multi-douzen megabit connections for the equivalent of 30$ a month is the norm (and thats based on several years old info, I didn't check recently, so its probably even more).
Japan's population density is a LOT higher than the US, so having fiber in every house is quite simple to do.
Software licenses are such an insignificant part of a business' spending... it doesn't make a dent in most' budget. There's the occasional one that can seriously mess you up if you're silly (Magicsoftware's eDeveloper comes to mind... awful licensing term, cost 3 arms and 4 legs if you base a software development project on it, especially if you're an ISV, but there's a reason its not very popular:) ), but as a general rule, the couple thousand dollars for the windows server web farm is quite insignificant when you sell an ERP priced in the 6-7 digits, for example.
Because it is usually simply not worth spending more time than I already have answering to someone who claim ignorance, yet is quite ignorant himself/herself
Half of your post made sense. The other half was filled with that non-sense.
The fact that its made to take notes... its organized in a folder hierarchy, you can put anything anywhere, and it is made to clip and paste stuff (for example, if I copy paste a random HTML table from a web site into a word processor, it will be ugly as hell, while in OneNote it will be nicely formatted and can be put anywhere).
It is used to make arrangements of clips and clippings, scribble with a tablet PC or a wacom or something, hand writing recognition (and its -really- good...). You can even record voice messages, and it will index them using voice recognitions and make them searchable (videos too).
Taking notes in a word processor vs something like OneNote, is like doing word processing in Excel vs a word processor. It doesn't compare.
The license is for non-commercial use. So any computer you do not use for business (read: probably most, though I realise not all, home computers), not just for educational computers:)
You don't see OneNote so much because it is not part of Office 2003, which is what most businesses you saw use. It -is- part of Office 2007, which is where I first saw it. Plus, the 2003 version blows (and is a stand alone software, so obviously NO ONE is going to bug their boss for a license of OneNote stand alone... I tried once and I got turned around quite fast by the guy who managed the licenses).
I'm a consultant too. I've seen it used heavily in all companies that migrated to Office 2007 (which is quite a lot for me, but isn't a representative sample: I'm mainly an architect for more bleeding edge systems, so obviously I'll end up working almost exclusively for the companies that already migrated to the new stuff)
A) Already supports multiple browsers in the released version (IE, Firefox and Safari) B) -already- is out for Mac (both PowerPC and Intel based) C) Moonlight (the Linux version) isn't out because they are only targeting Silverlight 2.0. That being said, they tend to be -ahead- of Microsoft in its implementation. The only reason it isn't done yet is because you cannot complete software based on incomplete reference (that is, even MS doesn't know 100% what Silverlight 2.0 will be yet, so Moonlight currently match whatever MS said, and MS is working with its devs to boot). It will be out very shortly after the Windows version comes out.
That said, Flash has too big an install base. Silverlight will tank on the public internet. The reason it is cool though, is like Flex -> for internal web based applications. Windows Presentation Foundation is better at it (but is Windows only), but when you need to support multiple platforms internally (more and more common), Silverlight will be sweet for that. You can reuse your.NET client code for it instead of recoding everything from scratch in a different environment.
Of interest, that was its original intent, too. It used to be called WPF/E (Windows Presentation Foundation/Everywhere), which was really just to leverage WPF in a cross browser/cross platform environment, and WPF was never meant for the public web. Its just, it CAN be used on the public web, so it doesn't cost MS anything beyond marketing (but no developer hours, or at least very little) to try and make a tiny dent in Flash' market. But don't be mistaken, Silverlight doesn't compete with Flash, it competes with -Flex-.
Giant txt files aren't a big deal. Just get something to return the first couple of records/lines so you can figure out the format, then push it through an ETL tool, and you're done.
In one of my last jobs, in about an hour I had made an SSIS package to parse and dump in a database the data from 10 gigs text files (which ran every morning), and the parsing took only a couple of minutes on a developer machine in debug mode. On an enterprise server with a lot of raw power, 12 terrabyte dumped in an OLAP system wouldn't be a big deal, and then you could analyse it at will.
Hmm? WPF's XBAP was IE only (now works on Firefox, though still Windows only), and is made mostly for actual applications/software without requiring an explicit installer. Thats the closest thing I can think of that you may be refering to, since it shares a common architecture with Silverlight 2.0.
Silverlight however, is NOT exclusive to IE, the official version also works on Macs, and there's a Linux version too, that will all be available when Silverlight 2.0 hits the market (Silverlight 1 is Windows and Mac only, but it barely counts, it is such a pathetic technology. 2 is where its at).
So I'm not sure why you're talking about IE's market share when talking about silverlight.
The version of Office that is more than enough for most people (Excel/Word/Powerpoint and OneNote... is there an open source equivalent of OneNote that is roughly as good, btw?) is far from a "lot". Its 140-150$ for 3 licenses. (it does have limitations though, so often you'll want open office, but its still not "a lot").
Keep in mind this subscription thing is also that: 70$/year for 3 licenses + support.
If you convert the energy in electricity, you won't heat yourself up. Though if you DO use it to heat up your house, well... you've never lived up north eh? When its getting awkwardly close to -40 degrees outside on a sunny day, I definately don't want to keep my house from warming up. Heating can make my power bill goes up quite a bit.
Slashdot Mirror
Im not talking about work. I'm talking about people taking a car to go buy milk one block away. Or for those who do use public transportation, use their car to get to the bus stop thats just across the block. The people who take their car to go to the gym thats 10 minutes away (come on...you're being lazy on your way to go exercise, the hell?)
Some people don't have a choice: but if the people who DO have a choice, made a different one, I feel a lot of fuel would be saved.
Being able to swap message format on the fly is really the way to go, since they all have advantages and issues, from speed, backward compatibility, debugging, security constraints, etc.
Thats one of the things I liked in messaging frameworks such as WCF (though its far from being the first to do it). Code to the API, the message protocol can be decided later, and changed/plugged in on the fly when something better comes.
I'd be interested in a study showing how much fuel we'd save if anytime people can get to somewhere on foot in 10 minutes, they'd walk instead of taking their car.
Just for kicks, I'd like to see those numbers.
Not quite. That "lingo" is the same that is part of the theoritical OOP model, and is also adopted in other more theoritical concepts such as UML representation (I guess that is still a tool, but yeah).
The difference is there because there is a large difference, both theoritical and in practice, between a fully abstract class and an interface, and it is important to keep that difference at the conceptual level. The "C++ way" exists because C++ is not a full implementation of the OOP model (neither is C# of course, but it is closer).
One very important time where that "theory affects reality to the point you really have to care" is when you start versionning your interfaces and classes. Long story short, you shouldn't, in general, expose an interface in a public API (well, the interface can be exposed, but I mean, a public method that will be used by a different consumer should not return an interface type, but only an implementation of it. So often, it is simply better to use fully abstract classes, since they are both interface an "implementation", so to speak, and will not suffer from the interface problem, especially if the abstract class realize/implement the interface, in which case you kill 2 birds with one stone).
The tool didn't take over the concept, since the concept predates the tools by decades in some cases :) (think of it in a similar way you would think about relational algebra vs a database).
Its a problem in the protocol. So the only systems that would not be vulnerable are those that did -not- follow the specs. Guess Windows is safe, since Microsoft never follows the specs :)
And really, all that is one thing: Web Development with the .NET/SQL Server stack. C#/HTML/CSS/Javascript/XML is basically ASP.NET, WPF is just a small piece of .NET 3.0, MSBuild is what runs when you hit F5 (and is incredibly simple compared to, let say, NAnt), SOAP/WS* is also part of ASP.NET, with a LITTLE added challenge of WSDL if you guys actually write those (and that fits with XML, since 90% of the learning curve of WSDL is knowing XSDs).
Thats a far cry from actually needing to know different echosystems or anything (though I'm not sure, was that actually your point?)
Good .NET developers will use delegates (not just anonymous ones) all over the place, as it is a core part of the framework, and is one of the best way to implement a large amount of the base design patterns (like Strategy and Observer) in .NET.
The thing is, there's a grand total of 7 good .NET developers in the world. If that.
.NET forms can be embedded as dll/activex in internet explorer if explicitely given appropriate permissions. Its kindda cute as a deployement platform for internal applications I suppose.
Personally, I prefer using ClickOnce or now Windows Presentation Foundation for that... easier, safer, more versatile, and better supported :)
The problem was in the context. If they're asking if there is multiple inheritance in C#, it is because they understand that it is an area where there can be confusion in the concepts of the language. Basically, it is a semantic question. And semantically, while a class, in the OOP and/or UML models, can be both "inherited" and "realized", an interface can only ever be "realized", and cannot be "inherited". Thus, saying that you can have multiple inheritance in C# is wrong.
Of course, in practice, as you did, you can just consider realization to be a public interface/behavior inheritance, but something like that holds no ground in a question about the semantic details of a language's implementation.
I agree, but at the same time, you have to be careful. Different languages have different takes at things someone may think as "obvious". A while ago I went to an interview for a .NET dev job, and the people interviewing me were very obviously all former C++ programmers recycled in C#. One of the questions they asked me was: "How do you do multiple inheritance in C#".
My answer was the obvious: "There isn't any, you can only inherit from one class". They started arguing with me about how it WAS possible, and actually common, to do multiple inheritance in C#. After a bit of discussion, I realised that they were still thinking in C++ term, where an interface is nothing more than a fully abstract class, so even if you only use "interfaces", you ARE doing multiple inheritance (while in C#, there are -core- differences between a fully abstract class and an interface). They had skipped that part of the book thinking they didn't need it :)
Super users being able to access any account can still be done through session or other server side mechanism :) The product we worked on at my previous job worked like that, and it went quite well too :)
Usually a big company will run through a caching proxy... so you'll see one user, hit all of your cachable content exactly one time. So unless your site is more of a web application than a site, you shouldn't see very high traffic coming from corporations.
Well, because the US doesn't even have enough money left to attack Canada, obviously :)
Well, according to data from 2007, tokyo's population density was of 4700 person per square km, vs 2050 for new york. Los Angele, the closest USA city on the list stands at about 2700. Nowhere close. Plus, if you go by population density of the country (since federal governments have a hand in pushing these things along), the US have an insignificant population density, while in Japan you'd have to cater to less total surface. Big difference.
In many asian countries, multi-douzen megabit connections for the equivalent of 30$ a month is the norm (and thats based on several years old info, I didn't check recently, so its probably even more).
Japan's population density is a LOT higher than the US, so having fiber in every house is quite simple to do.
Software licenses are such an insignificant part of a business' spending... it doesn't make a dent in most' budget. There's the occasional one that can seriously mess you up if you're silly (Magicsoftware's eDeveloper comes to mind... awful licensing term, cost 3 arms and 4 legs if you base a software development project on it, especially if you're an ISV, but there's a reason its not very popular :) ), but as a general rule, the couple thousand dollars for the windows server web farm is quite insignificant when you sell an ERP priced in the 6-7 digits, for example.
Because it is usually simply not worth spending more time than I already have answering to someone who claim ignorance, yet is quite ignorant himself/herself
Half of your post made sense. The other half was filled with that non-sense.
No worries, you proved your point on that quite well all on your own :)
The fact that its made to take notes... its organized in a folder hierarchy, you can put anything anywhere, and it is made to clip and paste stuff (for example, if I copy paste a random HTML table from a web site into a word processor, it will be ugly as hell, while in OneNote it will be nicely formatted and can be put anywhere).
It is used to make arrangements of clips and clippings, scribble with a tablet PC or a wacom or something, hand writing recognition (and its -really- good...). You can even record voice messages, and it will index them using voice recognitions and make them searchable (videos too).
Taking notes in a word processor vs something like OneNote, is like doing word processing in Excel vs a word processor. It doesn't compare.
The license is for non-commercial use. So any computer you do not use for business (read: probably most, though I realise not all, home computers), not just for educational computers :)
You don't see OneNote so much because it is not part of Office 2003, which is what most businesses you saw use. It -is- part of Office 2007, which is where I first saw it. Plus, the 2003 version blows (and is a stand alone software, so obviously NO ONE is going to bug their boss for a license of OneNote stand alone... I tried once and I got turned around quite fast by the guy who managed the licenses).
I'm a consultant too. I've seen it used heavily in all companies that migrated to Office 2007 (which is quite a lot for me, but isn't a representative sample: I'm mainly an architect for more bleeding edge systems, so obviously I'll end up working almost exclusively for the companies that already migrated to the new stuff)
Silverlight:
A) Already supports multiple browsers in the released version (IE, Firefox and Safari)
B) -already- is out for Mac (both PowerPC and Intel based)
C) Moonlight (the Linux version) isn't out because they are only targeting Silverlight 2.0. That being said, they tend to be -ahead- of Microsoft in its implementation. The only reason it isn't done yet is because you cannot complete software based on incomplete reference (that is, even MS doesn't know 100% what Silverlight 2.0 will be yet, so Moonlight currently match whatever MS said, and MS is working with its devs to boot). It will be out very shortly after the Windows version comes out.
That said, Flash has too big an install base. Silverlight will tank on the public internet. The reason it is cool though, is like Flex -> for internal web based applications. Windows Presentation Foundation is better at it (but is Windows only), but when you need to support multiple platforms internally (more and more common), Silverlight will be sweet for that. You can reuse your .NET client code for it instead of recoding everything from scratch in a different environment.
Of interest, that was its original intent, too. It used to be called WPF/E (Windows Presentation Foundation /Everywhere), which was really just to leverage WPF in a cross browser/cross platform environment, and WPF was never meant for the public web. Its just, it CAN be used on the public web, so it doesn't cost MS anything beyond marketing (but no developer hours, or at least very little) to try and make a tiny dent in Flash' market. But don't be mistaken, Silverlight doesn't compete with Flash, it competes with -Flex-.
Giant txt files aren't a big deal. Just get something to return the first couple of records/lines so you can figure out the format, then push it through an ETL tool, and you're done.
In one of my last jobs, in about an hour I had made an SSIS package to parse and dump in a database the data from 10 gigs text files (which ran every morning), and the parsing took only a couple of minutes on a developer machine in debug mode. On an enterprise server with a lot of raw power, 12 terrabyte dumped in an OLAP system wouldn't be a big deal, and then you could analyse it at will.
Hmm? WPF's XBAP was IE only (now works on Firefox, though still Windows only), and is made mostly for actual applications/software without requiring an explicit installer. Thats the closest thing I can think of that you may be refering to, since it shares a common architecture with Silverlight 2.0.
Silverlight however, is NOT exclusive to IE, the official version also works on Macs, and there's a Linux version too, that will all be available when Silverlight 2.0 hits the market (Silverlight 1 is Windows and Mac only, but it barely counts, it is such a pathetic technology. 2 is where its at).
So I'm not sure why you're talking about IE's market share when talking about silverlight.
The version of Office that is more than enough for most people (Excel/Word/Powerpoint and OneNote... is there an open source equivalent of OneNote that is roughly as good, btw?) is far from a "lot". Its 140-150$ for 3 licenses. (it does have limitations though, so often you'll want open office, but its still not "a lot").
Keep in mind this subscription thing is also that: 70$/year for 3 licenses + support.
If you convert the energy in electricity, you won't heat yourself up. Though if you DO use it to heat up your house, well... you've never lived up north eh? When its getting awkwardly close to -40 degrees outside on a sunny day, I definately don't want to keep my house from warming up. Heating can make my power bill goes up quite a bit.