... anyone looked at what would be required to aquire either of these keys using reverse engineering and/or cracking the encryption??
The public key can be read straight from the file, as detailed in the initial report, however I've been unable to locate a copy, Cryptome have removed it, suspicious, almost certainly because of the NSA leaning on them using ITAR.
How about an opensource distributed processing project?
The RSA Key cracking challenge (http://news.zdnet.co.uk/story/0,,t269-s2091852,00.html), suggests
cracking this (or the original MS key) is probably a little beyond feasibillity for a collaborative effort.
However this chance would be considerably enhanced by locating a [number] signed messages (ActiveX controls).
I'm not aware of any ever being located in the wild, however, since the likely targets of this attack are also unlikely to be aware of it this probably unlikely to ever emerge.
Consider the situation M$ would face if it happened...! Noone would feel safe using their products and open source software would probably explode.
Nothing new there:)
Oh yeah.. and the NSA would be very annoyed.
Oh, I don't doubt they already are very annoyed. They agree to allow M$ to export a reasonably secure crypto API on the grounds that the key is essentially escrowed, then M$ accidentally leak the keys pressence, allowing it to be over written. You can bet they are very, Very pissed with M$.
shut up those idiots who think nasa is in the business of special effects and sound stages.
Some people are still absolutely certain the earth is flat (http://www.flat-earth.org/).
So if some fools cannot be convinced after nearly 400 years of science evidence to the contary, what hope is there that those moron's(http://www.apfn.org/apfn/moon.htm) will ever accept it even when they can make the observations, themselves (http://www.discovery.com/stories/science/entrepre neurs/tourist.html).
It a pity Natural Selection seems to reward stupidity (The Bell Curve) and not punish it (http://www.darwinawards.com/).
Without the source how can a government be sure that the software cannot be used to spy on them
They cannot, indeed there is plently of evidence the Microsoft have already installed a backdoor in the CryptoAPI, as part of the US Government Echelon project.
NSA key to Windows: an open question http://www.cnn.com/TECH/computing/9909/03/window s. nsa.02/
Eavesdropping on the Planet http://www.thirdworldtraveler.com/Blum/Eavesdrop pi ngPlanet_RS.html
Microsoft collaborating with US spymasters http://www.theregister.co.uk/content/archive/659 8. html
A multipage contents is no replacement for a poor index; to many otherwise excellent books suffer from poor indexs, it's the second thing I check in the book store after reading the title.
I want a clear distinction between tutorial & reference books, and theory and practice. I should not need to open the cover to make this distinction. However when a concrete topic (specific Language/Topic/Product) is covered use standardised methodologies (ERD/UML etc) to illustrate & document the concepts in general. Avoid the authors own invented methodology (unless that is the topic of course).
Always include a CD/DVD, don't skimp on the cost, include the books text not just the source code.
Produce a cost effective Compendium's on topics, it's quite common for Publishers to produce 10+ Books on specific topic. If the series is particularly good, I might buy 3-4 of them however I'm unlikely to get the full set as seperate items, however if a bundle of all where available, I certainly those linked to my core skills. These should also be available on on a single CD/DVD. Keep these upto date and cost effective and I'm probably upgrading regularly.
Cable Systems have the wrong network topology to become the long term solution to the requirements of a broadband society.
They are rings and as such will always suffer from the contention being too close to the customer, leeches will always have a very negative impact.
Star based solutions such as xDSL offer much between solution. The bandwidth becomes more dedicated and contention is moved up stream, where the capacity can be managed in a much more effective way. Over time the 'last mile' is reduced so the xDSL become a bigger pipe, until ultimatly we have a star made from fibre rather than a fibre ring. Everbody wins, consumer, supplier, society.
This is tangentially similar to an idea I've been playing with for some time. My main problem is that as as S/W enginner I have little idea where to start, I was thinking PIC, but I'm really not sure. The idea would facilitate real large scale automation. Perhaps this could be the first 'open-hardware design':-) However the main reason I'm raising this an attempt to invalidate any future patent.
It's a TCP/IP/UDP trigger/switch. However the concept relies on being tiny, simple and cheap to produce, pence rather like the 555 timers we played with in school electronics. It needs to be a mass produced chip with a [very] low price point.
The chip(s) must operate in two modes. Switch and Trigger.
The trigger operates by producing a [multicast]packet contanining a unique GUID, when a specific input line is triggered dragged high(low).
The switch operates the opposite way by dragging a line high(low) when it receives specific GUID, within a [multicast] packet.
I'm thinking PIC(s) Would it be possible to implement a TCP/IP stack in PIC logic ?
I'm thinking multicast packets with TTL:1, to keep everything withing the subnet.
I'm thinking the IP equivalent of mecrcury/magnetic switches, relays etc for burglar/fire alarms, door switches, light switches, thermostatic switches, infact massive automation. It would then be possible to control just about any device via pretty much any IP enabled/connected computing device.
Consider some applications.
Switching Night Lights.
Burglar/Fire Alarm switches.
Light switches
Thermostats
I waking up in the morning to my digital clock radio and reading the time from my digital clock, setting the shower temperature on digital thermostat. Catching the news on my digital tv and checking the time on my digital watch. I speed to work, listerning to my digital radio ignoring the digital my cars speedometer and reading the digital speed warning signs on the road-side. I read the digital display on the lift to get to my floor where I use a digital pass to enter my office, where I read the digital display on the digital coffee machine, before checking my appointment on my digital PDA to use a digital computer and listerning to music om my digital music player, and taking phone calls on my digital mobile.
And all before lunch.
I like to thing I'm pretty normal person in the digital age.
What fraction of those R&D costs occur because of the constantly requirement to reinvent a 'new and better' wheel, because somebody else already holds the IP/patent ?
The fact majority, that is why we have hundreds of pain-killers and no cure for, say, AIDS.
I'm going to assume your post is based on ignorance and is not a troll. Time will reveal the wisdom of this decision.
Firstly the primary assertion of my first post was that C(hash) and Java are the essentially the same thing. Your post despite the errors, actually illustrates more differences.
C# is a ECMA standard.
So What? ECMA is not 'open'; and ECMAScript is a standard, but hardly anybody uses it, most Web Developers use propriety non-standard implementions of JavaScript/JScript/Flash etc.
Java is a single supplier system.
Not true! This is a list of ~26 supplies of Java Application Servers. This is about 25 more than the number of alternative implementation of C#.
http://dmoz.org/Computers/Programming/Languages/ Ja va/Server-Side/Application_Servers/
This is a list of ~258 Suppliers of Java Development tools.
http://dmoz.org/Computers/Programming/Languages/ Ja va/Development_Tools/
I suspect this one was just a typo on your part, transposing C(hash) [sic] and Java.
No it was not a typo. Java IS platform agnostic, implementations are available on dozens of platforms from dozens of suppliers including all the biggest IT suppliers, IBM, HP, Netscape(IPlanet), Oracle, Fujitsu it's available open source, and last but not least Sun.
Java has primitive types
Java's primative should have been wrapped in an object, however wrapping anything in an Object is well within the capabilities of even the most junior programmer; and that deficiency is hardly a criticism in the league of a broken encapsulation mechanism. Encapsulation is Essential to OO, that all variables must be objects is not.
Your whining about encapsulation shows a profound ignorance of what you speak of.
Then best insult is to prove I am wrong not just assert it.
C#'s private fields and methods are fully protected, you're thinking of properties....
So you will be able to me the error message generated, when you try to access a private member of anObject with the following code:
Care to back that up with a fact based argument rather than a blanket assertion ?
The bottom link goes to a page that describes the differences between "fields" and "properties" in C#.
Perhaps I should have beem more specific about what the links represented.
The first link goes to a C# Advocacy site that admits the pertinent facts (though it attempts to dress them up as an ease of use advantage). Anybody with even remotest familiar with the concepts of encapsulation or OO, SHOULD understand the ramifications of this (summary below) and can draw their own conclusions.
This mechanism allows members to be accessed directly outside the Object in the following fashion.
anObject.aMember = anOther.aMember;
Since this can be done even when aMember is private, this is a blatant violation of encapsulation.
The top link simply defines the OO term "encapsulation".
Yes; since few slashdot readers are actually OO Software Engineers, this was an attempt inform those unfamiliar with the expression, of what it actually means.
In neither instance is there any evidence that C# is not OO.
The first explains what encapsulation is the second is how C# violates it, my post is the conclusion; that since C# violates encapsulation it is not an OO language. The links are most certainly the only evidence you need to draw that conclusion.
In OOP; a field, a property, a member, or what ever you care to call it are all the same thing. The terminology is different languages for different languages.
They are ALL a variable encapusulated within an object.
C# has already geared itself up for a dominant position in tomorrow's enterprise development environment
This reads like Marketing hype worthy of the FUD-Master General himself.
... its ECMA standardisation
ECMA standardisation is a red herring when it comes to openness. ECMA is a closed organisation. As an individual expert Software Enginner I cannot join and influence language development at ECMA. However I can (and have) joined the JDC and bring my ideas to bear on the development of Java at (http://developer.java.sun.com/). This is open to anybody, including Microsoft.
Furthermore you only have to consider the death of the ECMA standardisation of JavaScript which has been an abysmal failure to seem that, ECMA is not a guarantee of a successful standardisation.
and Microsoft's atypical encouragement of competing implementations
You've obviously not been keeping up on current affairs. Microsoft have been found guilty of anti-competative behaviour in the US and are about to get another nailed in the EU too.
.NET, C# and the CLR is going to vapourise Sun's marketshare in server applications and enterprise programming userbase due to sheer openness.
Hardly likely; Microsoft have minimal existing presence (or mindshare) in the heavy weight enterprise sector. As long as they stick to the attitude of stick security, robustness and quality, they never will.
once Microsoft's asserted their dominance in the field,.NET won't remain an open standard for long.
The only truth, in your entire post, Embrace and Extend.
Perhaps we need another Moderation option (-1 Astrotufer).
the technical distinctions between Java and C# are of little concern
You have fallen for the M$ FUD; IMHO these distinction(s) are pretty fundamental.
1) Java is platform agnostic, C(hash) is WOSA only; couple this with fact that WinTel platform is dying at the hands of a raft of alternative platforms(http://news.bbc.co.uk/hi/english/busines s/newsid_1767000/1767695.stm)
2) Java is an industry wide product, C(hash) is from only one supplier.
2) Java is OO, C(hash) is not. [It's encapsulation mechanism is broken, all members are essentially public].
This post perfectly illustrates the danger of making out of context quotes and not understanding the issue(s) involved. It seeks to suggest that Cooper loses because he cannot live with development [in the real world].
I would suggest this is poor advocacy at best and deceitful at worst.
Coopers comments below:
Cooper: I think XP has some really deep, deep tacit assumptions going on, and I think the deepest tacit assumption is that we have a significant organisational problem, but we can't fix the organisation....
This perfectly illustrate this issue, Coopers solution is to fix the root cause [poor project planning/management] and not the problem [imprecise requirements/impossible deadlines]. Kents solution (XP) is to paper over the cracks and try to live with these requirements, and quit before your 'fail'. It seems XP boils down to the old adage of fixing the symptom and not the cause. This approach only produces an adequate solution at best and never results in excellence. MHO is that excellence should be the target of any Software Engineer who regards himself as a professional.
[rant]
Quite how this deserves +5 informative is mystery to me, it appears to me that the moderators did not even check this against the linked article, and moderated based on their own beliefs on XP rather than the merit of the post.
[/rant]
I think the moderation of all the posts critical of XP to 'flamebait' is a perfect illustration that it is all style over substance, and stand on its own merit.
Have you even looked at extremeprogramming.org [extremeprogramming.org]?
Yes, browsed site, partly read the book (#1). I have used some common techniques such as paired codering, I've never 'used' XP as such.
(#1) Pretty much the reason I gave up on it was the egotism 'look what I've invented, aint I clever attitude' when none of it is really new, and the constant name dropping. Well I've also worked 'Blue Chip' and produced and design my own processes. However unlike Donavan and Kent, I'm well aware I did not invent the fundamental techniques, something that seems completely lost on them. The expression 'standing on the shoulders of giants' come to mind and they are not the giants they seem to think they are.
XP is divided into four major areas of activity: Planning, Designing, Coding, Testing
I strongly disagree, it pays lip service to Planning, Analysis and Design phases and focus on Iteration, iteration and more iteration of the coding and testing cycle. Indeed that is one of its key flaws, it specifically excludes planning ahead, excludes prototyping (spikes are not prototypes because the code base is included in the final project, they are an early/initial iteration).
You are correct that XP is not that new.
Well that is my main contention conceeded.:)
It bears many similarities to the Rational Unified Process, but is in no way related to RAD.
However I don't see how you can claim it is related to RUP but not RAD. When the most obvious similarity in all three is the shift from a traditional 'waterfall' process model to an iterative process model.
One of the key aspects of RAD was two people around each PC, XP has certainly take this to extremes, they seem to have increased this to 2 Coders and a Domain expert now. A key difference with RUP, is it includes Architecture, Analysis and Design phases which are absent from XP.
XP stresses the importance of an underlying architecture and extensible design, before coding is started.
This is just plain wrong, these claims cannot be justified, indeed this one area where XP is self contraditory, it 'claims' it but also prohibits the process features that make it possible. It specifically excludes any architecture phases and 'rules out' forward planning, and
There are other contraditions, the idea that you don't waste time 'documenting' yet demand's strong adherence to code standards.
does not require any knowledge, on the part of the programmer, of the business processes...
Just how this is reconciled with this:
stresses customer satisfaction.
I have no idea.
Would you care to expand on what you mean by "REAL Software Engineering texts"?
I mean texts about Software Engineering as a discipline/subject not a Software Engineering texts about a specific methodology. So :
Software Engineering, Somerville.
Software Engineering Practice, Pressman.
Mythical Man Month.
Code Complete.
IMHO, All essential reading for any professional developer. Indeed I always seem to find the biggest advocates of XP have never read any of these. They are all hackers, not Software Engineers. The whole XP thing is a hackers charter, with little to do with Software Engineering, it may be suitable for small scale bespoke systems for single users, but not large scale, high value, multi-user projects.
Most people seem to believe that scientifically provable methodologies are the only option for software engineering...
Well I don't.
I have a range of both macro (RAD, JSD/JSP, Prince, OOAD, SAD, SSADM, OMT, UML, RUP, ERD, FARP) and micro (CRC,Peer-Review/Codewalking, etc)tools/techniques at my disposal, I use them when relevant and use some constantly (ERD, UML, Peer-Review) some only slightly (i.e. FARP, Prince).
If you are advocating XP over other methodologies, how many other Systems Development Methodogies have you used ? Just to provide a a point of reference !
What is it about the these XP advocates, are they really that poorly read that they genuinuely think they invented these ideas ?
XP is NOT the awesome paradigm shift that it is made out to be by its advocates; it is NOT even that new. It's just a repackaging of the failed RAD fad. Now XPer's can make the same mistakes all over again, produce poorly specified, unmaintainable mess. XP is an exercise in self promotion and marketing hype.
Those that have not come across it (or the ideas) before should read more REAL Software Engineering texts.
This system is already installed in 10,000 of the 30K homes in the City. Supports High Speed internet, Interactive Digital Television, Video on Demand, Council/Health and Educational Services.
There are a lot more Jobs in IT than just programmer, and nearly all of them require expertise in the problem domain. I would suggest that he should capitalise on this knowledge. He should approach the Companies that would tried to sell him software or supply his industry. I think he would struggle to compete with fresh grad's with current skill as an entry level coder, but would probably bring a range of skills as a Analysis, Designer, Tester, Pre/Post-Sales Support.
For this to work, the network must deliver a unique Mb/s data stream from the headend to each consumer. How many cable plants can do that today?
This is a very good point, most cable systems cannot do this because there network topology is a ring. ADSL however uses a star topology and this is probably the Killer Application for it!
Last time they tried this, the major roadblock was that no one could figure out how to build a server fast enough to stream multiple, unique video streams.
The solution is custom hardware, Kingston use hardware from www.nCube.com. This is about the best (only) kit you can get in the world for this application. Each Server support's about ~500Mbps of streams, divide this by the bit rate, say 2.5-4.5Mbps, gives 120+ quality streams per server.
BTW Two years ago these cost >150,000 uk pounds for each node(server), this year they cost 40K each. Next year who knows ?
Slashdot Mirror
The public key can be read straight from the file, as detailed in the initial report, however I've been unable to locate a copy, Cryptome have removed it, suspicious, almost certainly because of the NSA leaning on them using ITAR.
How about an opensource distributed processing project?
The RSA Key cracking challenge (http://news.zdnet.co.uk/story/0,,t269-s2091852,0
cracking this (or the original MS key) is probably a little beyond feasibillity for a collaborative effort.
However this chance would be considerably enhanced by locating a [number] signed messages (ActiveX controls).
I'm not aware of any ever being located in the wild, however, since the likely targets of this attack are also unlikely to be aware of it this probably unlikely to ever emerge.
Consider the situation M$ would face if it happened...! Noone would feel safe using their products and open source software would probably explode.
Nothing new there:)
Oh yeah.. and the NSA would be very annoyed.
Oh, I don't doubt they already are very annoyed. They agree to allow M$ to export a reasonably secure crypto API on the grounds that the key is essentially escrowed, then M$ accidentally leak the keys pressence, allowing it to be over written. You can bet they are very, Very pissed with M$.
shut up those idiots who think nasa is in the business of special effects and sound stages.
e neurs/tourist.html).
Some people are still absolutely certain the earth is flat (http://www.flat-earth.org/).
So if some fools cannot be convinced after nearly 400 years of science evidence to the contary, what hope is there that those moron's(http://www.apfn.org/apfn/moon.htm) will ever accept it even when they can make the observations, themselves (http://www.discovery.com/stories/science/entrepr
It a pity Natural Selection seems to reward stupidity (The Bell Curve) and not punish it (http://www.darwinawards.com/).
Without the source how can a government be sure that the software cannot be used to spy on them
w s. nsa.02/
p pi ngPlanet_RS.html
9 8. html
e &q =_NSAKEY+Key+Microsoft+CryptoAPI&btnG=Google+Searc h&meta=
They cannot, indeed there is plently of evidence the Microsoft have already installed a backdoor in the CryptoAPI, as part of the US Government Echelon project.
NSA key to Windows: an open question
http://www.cnn.com/TECH/computing/9909/03/windo
Eavesdropping on the Planet
http://www.thirdworldtraveler.com/Blum/Eavesdro
Microsoft collaborating with US spymasters
http://www.theregister.co.uk/content/archive/65
http://www.google.co.uk/search?hl=en&safe=activ
No1 - Good Indexes.
No2 - Better Indexes.
A multipage contents is no replacement for a poor index; to many otherwise excellent books suffer from poor indexs, it's the second thing I check in the book store after reading the title.
I want a clear distinction between tutorial & reference books, and theory and practice. I should not need to open the cover to make this distinction. However when a concrete topic (specific Language/Topic/Product) is covered use standardised methodologies (ERD/UML etc) to illustrate & document the concepts in general. Avoid the authors own invented methodology (unless that is the topic of course).
Always include a CD/DVD, don't skimp on the cost, include the books text not just the source code.
Produce a cost effective Compendium's on topics, it's quite common for Publishers to produce 10+ Books on specific topic. If the series is particularly good, I might buy 3-4 of them however I'm unlikely to get the full set as seperate items, however if a bundle of all where available, I certainly those linked to my core skills. These should also be available on on a single CD/DVD. Keep these upto date and cost effective and I'm probably upgrading regularly.
Cable Systems have the wrong network topology to become the long term solution to the requirements of a broadband society.
They are rings and as such will always suffer from the contention being too close to the customer, leeches will always have a very negative impact.
Star based solutions such as xDSL offer much between solution. The bandwidth becomes more dedicated and contention is moved up stream, where the capacity can be managed in a much more effective way. Over time the 'last mile' is reduced so the xDSL become a bigger pipe, until ultimatly we have a star made from fibre rather than a fibre ring. Everbody wins, consumer, supplier, society.
This is tangentially similar to an idea I've been playing with for some time. My main problem is that as as S/W enginner I have little idea where to start, I was thinking PIC, but I'm really not sure. The idea would facilitate real large scale automation. Perhaps this could be the first 'open-hardware design' :-) However the main reason I'm raising this an attempt to invalidate any future patent.
It's a TCP/IP/UDP trigger/switch. However the concept relies on being tiny, simple and cheap to produce, pence rather like the 555 timers we played with in school electronics. It needs to be a mass produced chip with a [very] low price point.
The chip(s) must operate in two modes. Switch and Trigger.
The trigger operates by producing a [multicast]packet contanining a unique GUID, when a specific input line is triggered dragged high(low).
The switch operates the opposite way by dragging a line high(low) when it receives specific GUID, within a [multicast] packet.
I'm thinking PIC(s) Would it be possible to implement a TCP/IP stack in PIC logic ?
I'm thinking multicast packets with TTL:1, to keep everything withing the subnet.
I'm thinking the IP equivalent of mecrcury/magnetic switches, relays etc for burglar/fire alarms, door switches, light switches, thermostatic switches, infact massive automation. It would then be possible to control just about any device via pretty much any IP enabled/connected computing device.
Consider some applications.
Switching Night Lights.
Burglar/Fire Alarm switches.
Light switches
Thermostats
See the potential ?
What do you think ?
Martin
I too enjoy a digital lifestyle.
I waking up in the morning to my digital clock radio and reading the time from my digital clock, setting the shower temperature on digital thermostat. Catching the news on my digital tv and checking the time on my digital watch. I speed to work, listerning to my digital radio ignoring the digital my cars speedometer and reading the digital speed warning signs on the road-side. I read the digital display on the lift to get to my floor where I use a digital pass to enter my office, where I read the digital display on the digital coffee machine, before checking my appointment on my digital PDA to use a digital computer and listerning to music om my digital music player, and taking phone calls on my digital mobile.
And all before lunch.
I like to thing I'm pretty normal person in the digital age.
Broadly speaking I agree. However:
What fraction of those R&D costs occur because of the constantly requirement to reinvent a 'new and better' wheel, because somebody else already holds the IP/patent ?
The fact majority, that is why we have hundreds of pain-killers and no cure for, say, AIDS.
Ok, I accept that i was mistaken over 'broken encapsulation'.
I'm going to assume your post is based on ignorance and is not a troll. Time will reveal the wisdom of this decision.
/ Ja va/Server-Side/Application_Servers/
/ Ja va/Development_Tools/
...
Firstly the primary assertion of my first post was that C(hash) and Java are the essentially the same thing. Your post despite the errors, actually illustrates more differences.
C# is a ECMA standard.
So What? ECMA is not 'open'; and ECMAScript is a standard, but hardly anybody uses it, most Web Developers use propriety non-standard implementions of JavaScript/JScript/Flash etc.
Java is a single supplier system.
Not true! This is a list of ~26 supplies of Java Application Servers. This is about 25 more than the number of alternative implementation of C#.
http://dmoz.org/Computers/Programming/Languages
This is a list of ~258 Suppliers of Java Development tools.
http://dmoz.org/Computers/Programming/Languages
I suspect this one was just a typo on your part, transposing C(hash) [sic] and Java.
No it was not a typo. Java IS platform agnostic, implementations are available on dozens of platforms from dozens of suppliers including all the biggest IT suppliers, IBM, HP, Netscape(IPlanet), Oracle, Fujitsu it's available open source, and last but not least Sun.
Java has primitive types
Java's primative should have been wrapped in an object, however wrapping anything in an Object is well within the capabilities of even the most junior programmer; and that deficiency is hardly a criticism in the league of a broken encapsulation mechanism. Encapsulation is Essential to OO, that all variables must be objects is not.
Your whining about encapsulation shows a profound ignorance of what you speak of.
Then best insult is to prove I am wrong not just assert it.
C#'s private fields and methods are fully protected, you're thinking of properties.
So you will be able to me the error message generated, when you try to access a private member of anObject with the following code:
localMember = anObject.privateMember ;
anObject.privateMember ++ ;
In the future, I suggest actually knowing what you're talking about before accusing people of FUDing.
It's just wrong.
;
Care to back that up with a fact based argument rather than a blanket assertion ?
The bottom link goes to a page that describes the differences between "fields" and "properties" in C#.
Perhaps I should have beem more specific about what the links represented.
The first link goes to a C# Advocacy site that admits the pertinent facts (though it attempts to dress them up as an ease of use advantage). Anybody with even remotest familiar with the concepts of encapsulation or OO, SHOULD understand the ramifications of this (summary below) and can draw their own conclusions.
This mechanism allows members to be accessed directly outside the Object in the following fashion.
anObject.aMember = anOther.aMember
Since this can be done even when aMember is private, this is a blatant violation of encapsulation.
The top link simply defines the OO term "encapsulation".
Yes; since few slashdot readers are actually OO Software Engineers, this was an attempt inform those unfamiliar with the expression, of what it actually means.
In neither instance is there any evidence that C# is not OO.
The first explains what encapsulation is the second is how C# violates it, my post is the conclusion; that since C# violates encapsulation it is not an OO language. The links are most certainly the only evidence you need to draw that conclusion.
C# has properties ...
In OOP; a field, a property, a member, or what ever you care to call it are all the same thing. The terminology is different languages for different languages.
They are ALL a variable encapusulated within an object.
It's encapsulation mechanism is broken, all members are essentially public
i on /0,,sid7_gci212060,00.html
e .h tm#2
That's news to me. Can you provide details?
Yes, A private member may be accessed directly, just as if it is a public member, this breaks encapsulation, for example:
localMember = anObject.privateMember ;
anObject.privateCounter ++;
http://searchnetworking.techtarget.com/sDefinit
http://genamics.com/developer/csharp_comparativ
C# has already geared itself up for a dominant position in tomorrow's enterprise development environment
... its ECMA standardisation
d _1 635000/1635317.stm
d _1 697000/1697766.stm
.NET, C# and the CLR is going to vapourise Sun's marketshare in server applications and enterprise programming userbase due to sheer openness.
.NET won't remain an open standard for long.
This reads like Marketing hype worthy of the FUD-Master General himself.
ECMA standardisation is a red herring when it comes to openness. ECMA is a closed organisation. As an individual expert Software Enginner I cannot join and influence language development at ECMA. However I can (and have) joined the JDC and bring my ideas to bear on the development of Java at (http://developer.java.sun.com/). This is open to anybody, including Microsoft.
Furthermore you only have to consider the death of the ECMA standardisation of JavaScript which has been an abysmal failure to seem that, ECMA is not a guarantee of a successful standardisation.
and Microsoft's atypical encouragement of competing implementations
You've obviously not been keeping up on current affairs. Microsoft have been found guilty of anti-competative behaviour in the US and are about to get another nailed in the EU too.
http://news.bbc.co.uk/hi/english/business/newsi
http://news.bbc.co.uk/hi/english/business/newsi
Hardly likely; Microsoft have minimal existing presence (or mindshare) in the heavy weight enterprise sector. As long as they stick to the attitude of stick security, robustness and quality, they never will.
once Microsoft's asserted their dominance in the field,
The only truth, in your entire post, Embrace and Extend.
Perhaps we need another Moderation option (-1 Astrotufer).
the technical distinctions between Java and C# are of little concern
s s/newsid_1767000/1767695.stm)
You have fallen for the M$ FUD; IMHO these distinction(s) are pretty fundamental.
1) Java is platform agnostic, C(hash) is WOSA only; couple this with fact that WinTel platform is dying at the hands of a raft of alternative platforms(http://news.bbc.co.uk/hi/english/busine
2) Java is an industry wide product, C(hash) is from only one supplier.
2) Java is OO, C(hash) is not. [It's encapsulation mechanism is broken, all members are essentially public].
This post perfectly illustrates the danger of making out of context quotes and not understanding the issue(s) involved. It seeks to suggest that Cooper loses because he cannot live with development [in the real world].
...
I would suggest this is poor advocacy at best and deceitful at worst.
Coopers comments below:
Cooper: I think XP has some really deep, deep tacit assumptions going on, and I think the deepest tacit assumption is that we have a significant organisational problem, but we can't fix the organisation.
This perfectly illustrate this issue, Coopers solution is to fix the root cause [poor project planning/management] and not the problem [imprecise requirements/impossible deadlines]. Kents solution (XP) is to paper over the cracks and try to live with these requirements, and quit before your 'fail'. It seems XP boils down to the old adage of fixing the symptom and not the cause. This approach only produces an adequate solution at best and never results in excellence. MHO is that excellence should be the target of any Software Engineer who regards himself as a professional.
[rant]
Quite how this deserves +5 informative is mystery to me, it appears to me that the moderators did not even check this against the linked article, and moderated based on their own beliefs on XP rather than the merit of the post.
[/rant]
I think the moderation of all the posts critical of XP to 'flamebait' is a perfect illustration that it is all style over substance, and stand on its own merit.
Have you even looked at extremeprogramming.org [extremeprogramming.org]?
:)
...
...
Yes, browsed site, partly read the book (#1). I have used some common techniques such as paired codering, I've never 'used' XP as such.
(#1) Pretty much the reason I gave up on it was the egotism 'look what I've invented, aint I clever attitude' when none of it is really new, and the constant name dropping. Well I've also worked 'Blue Chip' and produced and design my own processes. However unlike Donavan and Kent, I'm well aware I did not invent the fundamental techniques, something that seems completely lost on them. The expression 'standing on the shoulders of giants' come to mind and they are not the giants they seem to think they are.
XP is divided into four major areas of activity: Planning, Designing, Coding, Testing
I strongly disagree, it pays lip service to Planning, Analysis and Design phases and focus on Iteration, iteration and more iteration of the coding and testing cycle. Indeed that is one of its key flaws, it specifically excludes planning ahead, excludes prototyping (spikes are not prototypes because the code base is included in the final project, they are an early/initial iteration).
You are correct that XP is not that new.
Well that is my main contention conceeded.
It bears many similarities to the Rational Unified Process, but is in no way related to RAD.
However I don't see how you can claim it is related to RUP but not RAD. When the most obvious similarity in all three is the shift from a traditional 'waterfall' process model to an iterative process model.
One of the key aspects of RAD was two people around each PC, XP has certainly take this to extremes, they seem to have increased this to 2 Coders and a Domain expert now. A key difference with RUP, is it includes Architecture, Analysis and Design phases which are absent from XP.
XP stresses the importance of an underlying architecture and extensible design, before coding is started.
This is just plain wrong, these claims cannot be justified, indeed this one area where XP is self contraditory, it 'claims' it but also prohibits the process features that make it possible. It specifically excludes any architecture phases and 'rules out' forward planning, and
There are other contraditions, the idea that you don't waste time 'documenting' yet demand's strong adherence to code standards.
does not require any knowledge, on the part of the programmer, of the business processes
Just how this is reconciled with this:
stresses customer satisfaction.
I have no idea.
Would you care to expand on what you mean by "REAL Software Engineering texts"?
I mean texts about Software Engineering as a discipline/subject not a Software Engineering texts about a specific methodology. So :
Software Engineering, Somerville.
Software Engineering Practice, Pressman.
Mythical Man Month.
Code Complete.
IMHO, All essential reading for any professional developer. Indeed I always seem to find the biggest advocates of XP have never read any of these. They are all hackers, not Software Engineers. The whole XP thing is a hackers charter, with little to do with Software Engineering, it may be suitable for small scale bespoke systems for single users, but not large scale, high value, multi-user projects.
Most people seem to believe that scientifically provable methodologies are the only option for software engineering
Well I don't.
I have a range of both macro (RAD, JSD/JSP, Prince, OOAD, SAD, SSADM, OMT, UML, RUP, ERD, FARP) and micro (CRC,Peer-Review/Codewalking, etc)tools/techniques at my disposal, I use them when relevant and use some constantly (ERD, UML, Peer-Review) some only slightly (i.e. FARP, Prince).
If you are advocating XP over other methodologies, how many other Systems Development Methodogies have you used ? Just to provide a a point of reference !
What is it about the these XP advocates, are they really that poorly read that they genuinuely think they invented these ideas ?
XP is NOT the awesome paradigm shift that it is made out to be by its advocates; it is NOT even that new. It's just a repackaging of the failed RAD fad. Now XPer's can make the same mistakes all over again, produce poorly specified, unmaintainable mess. XP is an exercise in self promotion and marketing hype.
Those that have not come across it (or the ideas) before should read more REAL Software Engineering texts.
http://whatis.techtarget.com/definition/0,,sid9
IANAL, but if you ...win in court and end up bankrupt due to legal fees. Then the loser pays your legal fees.
I submitted this story about my home City, Kingston upon Hull in the UK, which announced similar project over a year ago.
http://news.bbc.co.uk/hi/english/sci/tech/newsi
This system is already installed in 10,000 of the 30K homes in the City. Supports High Speed internet, Interactive Digital Television, Video on Demand, Council/Health and Educational Services.
More US bias from Slashdots ?
There are a lot more Jobs in IT than just programmer, and nearly all of them require expertise in the problem domain. I would suggest that he should capitalise on this knowledge. He should approach the Companies that would tried to sell him software or supply his industry. I think he would struggle to compete with fresh grad's with current skill as an entry level coder, but would probably bring a range of skills as a Analysis, Designer, Tester, Pre/Post-Sales Support.
For this to work, the network must deliver a unique Mb/s data stream from the headend to each consumer. How many cable plants can do that today?
This is a very good point, most cable systems cannot do this because there network topology is a ring. ADSL however uses a star topology and this is probably the Killer Application for it!
I have tyhe best VOD solution in my livingroom...
True VOD does not even require you to visit the Video Store in the first place or worse take it back on a cold rainy night.
Last time they tried this, the major roadblock was that no one could figure out how to build a server fast enough to stream multiple, unique video streams.
The solution is custom hardware, Kingston use hardware from www.nCube.com. This is about the best (only) kit you can get in the world for this application. Each Server support's about ~500Mbps of streams, divide this by the bit rate, say 2.5-4.5Mbps, gives 120+ quality streams per server.
BTW Two years ago these cost >150,000 uk pounds for each node(server), this year they cost 40K each. Next year who knows ?