Only about 120 classes of the 1200 in the Dotnet platform are standardized as part of the C Sharp language, so standardization offers little protection if your application uses a GUI (Windows Forms, Web Forms) or a database (ADO Dotnet). Not only are these libraries not standardized, they are likely to be protected by patents.
Sun does not have the same room for manoeuvre as MS since the JCP has other powerful participants. In practice, there have been few ownership/legal issues in developing Open Source versions of the JVM - see the Kawa web site for a list of these. Their complaints revolve around issues such as access to the test suites - ultimately Sun just owns the Java name, not all the implementations.
I feel I should point out for any potential newbies that SuSE has the YAST2 (SAX2-based) tool that is basically just like the Windows display mode change tool, right down to the 'can you see the screen when I do this?' bit and the geometry settings available on the more sophisticated Windows drivers. There's no need to hack the X config files if you go this route. Some other distros must have equivalents, though I'm not aware of any.
IMHO SuSE 8 might be evolutionary in development terms but it is verging on the revolutionary in use - so much of the graft is now managed coherently in YAST2. As a comparison, it was much quicker to set up my SuSE partition than my W2K one because (a) SuSE detected my LAN card and W2K didn't and (b) every W2K update and most drivers required a reboot - over 20 in all - whereas SuSE's online updates were handled in one go.
FWIW, the issue of whether writing to volatile storage counts as a committed transaction has been kicking around for a long time.
I remember in the mid-80s, Stratus and Tandem would duel over TPC benchmarks, and while Stratus did respectably on conventional disk-based writes, they did try to get the TPC council to allow writes to their resilient (duplicated), battery-backed memory to count too. I don't think they succeeded then, and IMHO some rather cruddy PC memory system should not be allowed to count now.
Having been to a number of MS 'Executive Briefings' my impression is that by far the most requested item by large customers has been proper Java support. Right now it is costing companies a huge amount of effort to integrate Excel and Outlook apps with Java-based transactional systems, and going right back to 1998 the story from MS has never been "How can we help solve your problem?", only "How can we dominate this space and exclude competition?"
Ironically, we had MS people on site for over a year to gather 'requirements' and help 'influence strategy'. There's no real question that this was by and large ignored - a small insight into what perhaps has been one of the most dramatic examples of contempt for customers ever exhibited by a major corporation.
Beatles were more memorable
on
Live Via Satellite
·
· Score: 3, Informative
The first worldwide TV program was 5 years later (June 25, 1967) - the Beatles in their Magical Mystery Tour mode doing "All You Need is Love." Covered 24 countries, 5 continents via Echo II (?) a satellite which had no transmitter, just a reflector.
I'm sure some worthy celebs would like to commemorate this event - how about it Sir Paul/Mick?
(Unfortunately, though alive I think I was probably tuned to Listen with Mother instead:( )
Sounds to me like you're trying to be a little bit pregnant.
If I can run an arbitrary program then I can subvert any existing security policy - interpose layers to snoop key exchanges, spoof integrity checks etc. A locked-down platform is all or nothing - you're describing something like Java or ActiveX, but Palladium needs to be more than that if it is to work.
Yep, but you can give up on the idea of taking one of these boxes into work like you can today.
Linux as it is now won't be able to give guarantees of integrity equivalent to Dotnet + Palladium. The tables will turn quite quickly and corporate security policies will have good reason to prevent you running Linux. Ironic, I know.
Not really. At least here in the UK, an invention mustn't be obvious or well-known to practitioners. I'm sure that certification vendors such as Schulumberger in France or secure system developers such as the RSRE (Royal Signals and Radar Establishment) here are well aware of such a fundamental principle. My own former colleagues in ICL did a lot of work on trusted OSes - I'm sure that work is relevant and was published.
Stability isn't the same thing as security. I have exactly the same problem running a binary on Linux as I do on Windows - integrity isn't guaranteed (uncontrolled pointers...), rights can only be given at a very coarse level (run as root, write anything in this directory...) and so forth.
Windows is about to fix this with Dotnet. Palladium will just be icing on the cake for the DRM crowd. Meanwhile, precisely nothing equivalent is happening on what we refer to as the Linux platform, only in assorted addons (Java, Dotnet, Parrot etc.) which are semi-integrated at best.
I implemented a system that worked that way on a crappy little Verifone credit card terminal 12 years ago. In fact, some smartcard firms must have done similiar things, if only to check the integrity of their own code.
You're missing the point. Palladium isn't merely a solution to keeping your secrets safe and your system integrity intact, it's a solution to keeping other people's secrets safe from you when in "your" environment. That's why is needs to be tamper-proof hardware - the first problem can be solved by using Java or a similar software-only platform.
Maybe, but it can be made a lot more secure than the X-Box. Think of something like a satellite CAM card - hardware encryption that you need an electron microscope to break. Encrypt the apps, seal the platform and you have a fairly tight setup, particularly if part of the app is remote.
Interesting to note that none of the six bulletted features, all relating to users' control and users' trust relationships, require anything new or different from current PC platforms. Therefore the only reason for Intel & co. to be involved must relate to other parties trust and control.
To give a concrete example, a virtual machine like Python or Java can offer complete control over what an application can do with your identity and information and guarantee the integrity of your PC. No hardware support is required at all.
However, for other parties to trust your identity and control the use of their information requires a locked-down platform. Again, a VM-like system is a solution, but the VM's integrity must be guaranteed for them to trust it, hence the need for a tamper-proof, hardware-based solution.
Now, here's the interesting bit. Both open source and closed systems appear to be converging on the use of VMs, but for different reasons. In the open source world, Java, Parrot and Mono/DotGNU are seen as simply practical solutions to portability problems, with security and other factors some way further down the list. For closed systems, security (meaning keeping the information closed) will soon be the priority, far surpassing the need to maintain cross-platform (i86, PPC, ARM etc.) builds cheaply.
Open source advocates should not respond by continuing to develop more monolithic and fundamentally insecure C binaries - this will just leave Linux exposed to criticisms from future security-related interests, such as corporate IT management. Instead, we should embrace systems that can guarantee security - the difference being that it is security on the user's terms, not the vendor's. In fact, a high-level VM (like Java's) is the ideal platform for open source because (thanks to decompilers and the semantic equivalence of bytecode and Java source) it is impossible to ship code that isn't open.
There's a lot of positive spin for Open Source to be gained from this development, but the first thing to recognize is the critical importance of VMs (preferably a single "anointed" VM) to the viability of Linux platform.
the next version of Pocket PC - that will surely have a.NET runtime in it. That will make all the problems go away, won't it? Just compile the apps to IL, target, and distribute.
Let's hope your skepticism is justified. Because if it isn't, Linux as a platform will be in very serious trouble.
Linux has no answer to cross-platform code, the one exception being Gnome with Mono. If that remains the only effort, and continues to attract hype and developer support, one day soon we'll wake up and find that the single viable open source platform to write to is under the technical direction of Microsoft.
FUD. Of the roughly 1200 classes in Dotnet today, less than 200 are part of the ECMA standards (CLR, C Sharp). Most of the Dotnet functionality is contained in APIs such as Windows Forms and ADO.NET and these are not standardized, and are covered by a number of patents. Ximian may be attempting to clone them but any use of these components will entail risk. This is not the case with the Java (J2SE, J2EE) platform, which is well established on Linux and other OSes.
In case it's not transparently obvious to everyone, I should point out that the above is complete FUD.
Anyone contemplating developing Java on Linux applications should read the following to set their mind at rest:
1. The list of already existing and completely open VM and class library ports at Kaffe.org. Sun can do nothing to restrict any of these clean-room developments, nor has it shown any sign of wishing to.
2. Sun's Community Source license. This explains the terms on which Sun's own code is shared. Note that this license has proven open enough for the Blackdown Linux port to be developed. Interestingly, the Blackdown group have listed what they would ideally want from Sun to keep their implementation fully compatible, and a "better definition" of the Java platform doesn't appear to be needed.
Attempts by Mono developers and other interested parties to discredit Java-on-Linux should be treated with great caution. Not only do they try to represent Sun and IBM as imposing more restrictions on open developments than Microsoft (presumably their status as actual open source contributors counts for nothing against the neutral and fair-dealing Redmond empire, well known for its 'embrace' of standards) but they deliberately misrepresent what parts of Dotnet are actually open and standardized. Any open app that uses a Dotnet GUI (Windows Forms or Web Forms) or database API (ADO.NET) is not using standard mechanisms - these APIs, and in fact over 1000 of the 1200 classes in Dotnet today, are entirely proprietary and are protected by patents.
Oh, and I didn't use the word "all" - et al translates as 'and others', a somewhat different meaning.
I make a living coding Java on Linux. Frankly, your statement is silly.
How can the point be silly if you've decided to code in Java? Or are your technical choices made at random?
Second, having to deal with with what Java version is on the end user's machine leads to most of us distributing the JRE with our code.
That's bizarre - most companies do precisely the opposite. For example, BEA provide a WebLogic install specific to Windows that does include a VM, because Windows is a relatively well-defined base, but for Linux, Solaris, HP-UX etc. they provide one without. This is because they realize that choosing a VM on these platforms is generally the job of a sys admin and not something that their product alone should dictate.
Given those (and other) issues, I think I would seriously consider C or C++ with perhaps wxwindows as the GUI toolkit for future end user apps.
I suspect that there's some hollow laughter from other/. developers at this point. Anybody proposing C++ with GUI libraries (not to mention STL, database and threading libraries) as a more portable platform than Java is seriously out of touch with reality.
Java's standard library is nice, but the performance, memory and distribution hassles are (or can be) a bitch.
Performance is good enough to run high-volume transactions behind innumerable web sites. It's also good enough for intensive GUI work (e.g. JBuilder). Memory demands don't seem to be a problem, after all, people are developing apps that can run on phones. 'Distribution hassles' presumably refers to the same self-inflicted 'problem' above.
Anyhow, there are plenty of better candidates for 'redundant language' to pick on (XML and its myriad of ways to computational completeness, Python, Javascript...)
And the closed Java implementations on Linux and BSD are a lot more promising (as in production-ready) than Mono et al.
In fact, with alternative hardware platforms such as PPC and ARM diluting the hold of x86-based Linux, anyone now writing end-user Linux apps in something other than Java should pause to consider how accessible and maintainable their products will be.
I agree Java will certainly face a challenge from the Dotnet platform. F Sharp is a very clear warning signal to Sun, and it won't just be academics who want to move on from C++ and Simula 67.
The quotes were just accessible examples of why this book is inappropriate as an introduction to conceptual computing fundamentals. There are other papers highlighting problems with the GoF book in its own terms, if that's what you're interested in. LISP and Dylan aren't perfect models either, and I certainly wouldn't base a course entirely on them.
I'm sure you are aware of established books on computing fundamentals, such as How to Design Programs, so it must be fairly obvious that the GoF is not remotely comparable to these.
Just as a matter of interest, did you honestly believe that this comment had some relevance to the substance of the story?
If no, what problem do you have that forces you to share with us the first thing that comes into your head?
If yes, you might like to join the privileged elite (10%) of your country and travel outside it occasionally to improve your sociability in international forums.
Nonsense. This book has nothing to do with IT fundamentals, OO or otherwise. As manypeople have pointed out, at best it can be regarded as a roll-your-own appendage to cover the flaws of current mainstream languages. This situation does not apply in academia - they can use any language they like.
Lucky you. Some poor recommendations so far IMHO, with the honourable exception of SICP.
The key thing is to keep a sense of proportion - anyone referring to UML or the GoF design patterns book has failed to understand what the fundamentals of IT are, and is certainly overestimating the relevance of their own preferred language or paradigm.
Scheme/LISP, logic and some database theory is a good way to approach the fundamentals, as it was 15 years ago when I went through it. They won't thank you to begin with, but its what college is for! I'm not totally sold on SICP, students might think it's a bit pedantic, you might like to look at How to Design Programs as an alternative. I don't have a good reference for database and logic texts - I use Joe Celko's Data and Databases book, but this isn't suitable as an introduction.
Windows CE.NET allows developers to select the optimal hardware architecture for their projects by supporting a broad range of processor families, including x86, Xscale, ARM, MIPS and the SH series. Device bring-up is now faster than ever with Windows CE.NET providing numerous board support packages and features such as the Platform Wizard, which provides preconfigured design templates for embedded devices such as handhelds, smart phones, Web pads, retail point-of-sale devices and set-top boxes.
Your devastating critique continues:
What does this [deployment problems] have to do with anything? There are many, _many_ sites that are dedicated repositories of binaries for particular architectures.
There are? I think you'll find that most applications (as opposed to platform builds) are offered in a few builds that the authors know about, and finding any others is down to luck.
Let's try searching for 'linux tetris' in Google, shall we? OK, here's the first one. Oh dear! There's one build, and it's for x86 Mandrake. Care to point to one of your "many, many binary sites" that has it built for a PPC, for example?
And, lastly, our special bonus, the extra-insightful:
Having a VM is a waste of time... Any architecture can be virtualized
Riiiigghhtt. Well, I hope you've taken the opportunity to tell Microsoft, Sun and all those phone vendors that they're wasting their time pushing VMs. I bet they'll be kicking themselves soon! Perhaps you'd like to point to a phone or PDA that runs a foreign CPU instruction set, just as an example?
Slashdot Mirror
This doesn't really add up to much.
Only about 120 classes of the 1200 in the Dotnet platform are standardized as part of the C Sharp language, so standardization offers little protection if your application uses a GUI (Windows Forms, Web Forms) or a database (ADO Dotnet). Not only are these libraries not standardized, they are likely to be protected by patents.
Sun does not have the same room for manoeuvre as MS since the JCP has other powerful participants. In practice, there have been few ownership/legal issues in developing Open Source versions of the JVM - see the Kawa web site for a list of these. Their complaints revolve around issues such as access to the test suites - ultimately Sun just owns the Java name, not all the implementations.
I feel I should point out for any potential newbies that SuSE has the YAST2 (SAX2-based) tool that is basically just like the Windows display mode change tool, right down to the 'can you see the screen when I do this?' bit and the geometry settings available on the more sophisticated Windows drivers. There's no need to hack the X config files if you go this route. Some other distros must have equivalents, though I'm not aware of any.
IMHO SuSE 8 might be evolutionary in development terms but it is verging on the revolutionary in use - so much of the graft is now managed coherently in YAST2. As a comparison, it was much quicker to set up my SuSE partition than my W2K one because (a) SuSE detected my LAN card and W2K didn't and (b) every W2K update and most drivers required a reboot - over 20 in all - whereas SuSE's online updates were handled in one go.
FWIW, the issue of whether writing to volatile storage counts as a committed transaction has been kicking around for a long time.
I remember in the mid-80s, Stratus and Tandem would duel over TPC benchmarks, and while Stratus did respectably on conventional disk-based writes, they did try to get the TPC council to allow writes to their resilient (duplicated), battery-backed memory to count too. I don't think they succeeded then, and IMHO some rather cruddy PC memory system should not be allowed to count now.
Having been to a number of MS 'Executive Briefings' my impression is that by far the most requested item by large customers has been proper Java support. Right now it is costing companies a huge amount of effort to integrate Excel and Outlook apps with Java-based transactional systems, and going right back to 1998 the story from MS has never been "How can we help solve your problem?", only "How can we dominate this space and exclude competition?"
Ironically, we had MS people on site for over a year to gather 'requirements' and help 'influence strategy'. There's no real question that this was by and large ignored - a small insight into what perhaps has been one of the most dramatic examples of contempt for customers ever exhibited by a major corporation.
The first worldwide TV program was 5 years later (June 25, 1967) - the Beatles in their Magical Mystery Tour mode doing "All You Need is Love." Covered 24 countries, 5 continents via Echo II (?) a satellite which had no transmitter, just a reflector.
:( )
I'm sure some worthy celebs would like to commemorate this event - how about it Sir Paul/Mick?
(Unfortunately, though alive I think I was probably tuned to Listen with Mother instead
Sounds to me like you're trying to be a little bit pregnant.
If I can run an arbitrary program then I can subvert any existing security policy - interpose layers to snoop key exchanges, spoof integrity checks etc. A locked-down platform is all or nothing - you're describing something like Java or ActiveX, but Palladium needs to be more than that if it is to work.
Yep, but you can give up on the idea of taking one of these boxes into work like you can today.
Linux as it is now won't be able to give guarantees of integrity equivalent to Dotnet + Palladium. The tables will turn quite quickly and corporate security policies will have good reason to prevent you running Linux. Ironic, I know.
Not really. At least here in the UK, an invention mustn't be obvious or well-known to practitioners. I'm sure that certification vendors such as Schulumberger in France or secure system developers such as the RSRE (Royal Signals and Radar Establishment) here are well aware of such a fundamental principle. My own former colleagues in ICL did a lot of work on trusted OSes - I'm sure that work is relevant and was published.
Stability isn't the same thing as security. I have exactly the same problem running a binary on Linux as I do on Windows - integrity isn't guaranteed (uncontrolled pointers...), rights can only be given at a very coarse level (run as root, write anything in this directory...) and so forth.
Windows is about to fix this with Dotnet. Palladium will just be icing on the cake for the DRM crowd. Meanwhile, precisely nothing equivalent is happening on what we refer to as the Linux platform, only in assorted addons (Java, Dotnet, Parrot etc.) which are semi-integrated at best.
Grief, did they get a patent on that?
I implemented a system that worked that way on a crappy little Verifone credit card terminal 12 years ago. In fact, some smartcard firms must have done similiar things, if only to check the integrity of their own code.
You're missing the point. Palladium isn't merely a solution to keeping your secrets safe and your system integrity intact, it's a solution to keeping other people's secrets safe from you when in "your" environment. That's why is needs to be tamper-proof hardware - the first problem can be solved by using Java or a similar software-only platform.
Maybe, but it can be made a lot more secure than the X-Box. Think of something like a satellite CAM card - hardware encryption that you need an electron microscope to break. Encrypt the apps, seal the platform and you have a fairly tight setup, particularly if part of the app is remote.
Interesting to note that none of the six bulletted features, all relating to users' control and users' trust relationships, require anything new or different from current PC platforms. Therefore the only reason for Intel & co. to be involved must relate to other parties trust and control.
To give a concrete example, a virtual machine like Python or Java can offer complete control over what an application can do with your identity and information and guarantee the integrity of your PC. No hardware support is required at all.
However, for other parties to trust your identity and control the use of their information requires a locked-down platform. Again, a VM-like system is a solution, but the VM's integrity must be guaranteed for them to trust it, hence the need for a tamper-proof, hardware-based solution.
Now, here's the interesting bit. Both open source and closed systems appear to be converging on the use of VMs, but for different reasons.
In the open source world, Java, Parrot and Mono/DotGNU are seen as simply practical solutions to portability problems, with security and other factors some way further down the list. For closed systems, security (meaning keeping the information closed) will soon be the priority, far surpassing the need to maintain cross-platform (i86, PPC, ARM etc.) builds cheaply.
Open source advocates should not respond by continuing to develop more monolithic and fundamentally insecure C binaries - this will just leave Linux exposed to criticisms from future security-related interests, such as corporate IT management. Instead, we should embrace systems that can guarantee security - the difference being that it is security on the user's terms, not the vendor's. In fact, a high-level VM (like Java's) is the ideal platform for open source because (thanks to decompilers and the semantic equivalence of bytecode and Java source) it is impossible to ship code that isn't open.
There's a lot of positive spin for Open Source to be gained from this development, but the first thing to recognize is the critical importance of VMs (preferably a single "anointed" VM) to the viability of Linux platform.
the next version of Pocket PC - that will surely have a .NET runtime in it. That will make all the problems go away, won't it? Just compile the apps to IL, target, and distribute.
Let's hope your skepticism is justified. Because if it isn't, Linux as a platform will be in very serious trouble.
Linux has no answer to cross-platform code, the one exception being Gnome with Mono. If that remains the only effort, and continues to attract hype and developer support, one day soon we'll wake up and find that the single viable open source platform to write to is under the technical direction of Microsoft.
However did this happen?
.NET is already being ported to other platforms.
FUD. Of the roughly 1200 classes in Dotnet today, less than 200 are part of the ECMA standards (CLR, C Sharp). Most of the Dotnet functionality is contained in APIs such as Windows Forms and ADO.NET and these are not standardized, and are covered by a number of patents. Ximian may be attempting to clone them but any use of these components will entail risk. This is not the case with the Java (J2SE, J2EE) platform, which is well established on Linux and other OSes.
Hmmm, I was thinking more along the lines of 1x2 GHz AMD, 1GB, RAID 0 Barracuda drives. Any good?
Anyone know if Premiere makes use of SSE2, or other Intel-only tweaks?
TVM
In case it's not transparently obvious to everyone, I should point out that the above is complete FUD.
Anyone contemplating developing Java on Linux applications should read the following to set their mind at rest:
1. The list of already existing and completely open VM and class library ports at Kaffe.org. Sun can do nothing to restrict any of these clean-room developments, nor has it shown any sign of wishing to.
2. Sun's Community Source license. This explains the terms on which Sun's own code is shared. Note that this license has proven open enough for the Blackdown Linux port to be developed. Interestingly, the Blackdown group have listed what they would ideally want from Sun to keep their implementation fully compatible, and a "better definition" of the Java platform doesn't appear to be needed.
Attempts by Mono developers and other interested parties to discredit Java-on-Linux should be treated with great caution. Not only do they try to represent Sun and IBM as imposing more restrictions on open developments than Microsoft (presumably their status as actual open source contributors counts for nothing against the neutral and fair-dealing Redmond empire, well known for its 'embrace' of standards) but they deliberately misrepresent what parts of Dotnet are actually open and standardized. Any open app that uses a Dotnet GUI (Windows Forms or Web Forms) or database API (ADO.NET) is not using standard mechanisms - these APIs, and in fact over 1000 of the 1200 classes in Dotnet today, are entirely proprietary and are protected by patents.
Oh, and I didn't use the word "all" - et al translates as 'and others', a somewhat different meaning.
I make a living coding Java on Linux. Frankly, your statement is silly.
/. developers at this point. Anybody proposing C++ with GUI libraries (not to mention STL, database and threading libraries) as a more portable platform than Java is seriously out of touch with reality.
How can the point be silly if you've decided to code in Java? Or are your technical choices made at random?
Second, having to deal with with what Java version is on the end user's machine leads to most of us distributing the JRE with our code.
That's bizarre - most companies do precisely the opposite. For example, BEA provide a WebLogic install specific to Windows that does include a VM, because Windows is a relatively well-defined base, but for Linux, Solaris, HP-UX etc. they provide one without. This is because they realize that choosing a VM on these platforms is generally the job of a sys admin and not something that their product alone should dictate.
Given those (and other) issues, I think I would seriously consider C or C++ with perhaps wxwindows as the GUI toolkit for future end user apps.
I suspect that there's some hollow laughter from other
Java's standard library is nice, but the performance, memory and distribution hassles are (or can be) a bitch.
Performance is good enough to run high-volume transactions behind innumerable web sites. It's also good enough for intensive GUI work (e.g. JBuilder). Memory demands don't seem to be a problem, after all, people are developing apps that can run on phones. 'Distribution hassles' presumably refers to the same self-inflicted 'problem' above.
Yeah, he probably meant to say LISP :-)
Anyhow, there are plenty of better candidates for 'redundant language' to pick on (XML and its myriad of ways to computational completeness, Python, Javascript...)
And the closed Java implementations on Linux and BSD are a lot more promising (as in production-ready) than Mono et al.
In fact, with alternative hardware platforms such as PPC and ARM diluting the hold of x86-based Linux, anyone now writing end-user Linux apps in something other than Java should pause to consider how accessible and maintainable their products will be.
I agree Java will certainly face a challenge from the Dotnet platform. F Sharp is a very clear warning signal to Sun, and it won't just be academics who want to move on from C++ and Simula 67.
The quotes were just accessible examples of why this book is inappropriate as an introduction to conceptual computing fundamentals. There are other papers highlighting problems with the GoF book in its own terms, if that's what you're interested in. LISP and Dylan aren't perfect models either, and I certainly wouldn't base a course entirely on them.
I'm sure you are aware of established books on computing fundamentals, such as How to Design Programs, so it must be fairly obvious that the GoF is not remotely comparable to these.
...another one.
Just as a matter of interest, did you honestly believe that this comment had some relevance to the substance of the story?
If no, what problem do you have that forces you to share with us the first thing that comes into your head?
If yes, you might like to join the privileged elite (10%) of your country and travel outside it occasionally to improve your sociability in international forums.
Nonsense. This book has nothing to do with IT fundamentals, OO or otherwise. As many people have pointed out, at best it can be regarded as a roll-your-own appendage to cover the flaws of current mainstream languages. This situation does not apply in academia - they can use any language they like.
Lucky you. Some poor recommendations so far IMHO, with the honourable exception of SICP.
The key thing is to keep a sense of proportion - anyone referring to UML or the GoF design patterns book has failed to understand what the fundamentals of IT are, and is certainly overestimating the relevance of their own preferred language or paradigm.
Scheme/LISP, logic and some database theory is a good way to approach the fundamentals, as it was 15 years ago when I went through it. They won't thank you to begin with, but its what college is for! I'm not totally sold on SICP, students might think it's a bit pedantic, you might like to look at How to Design Programs as an alternative. I don't have a good reference for database and logic texts - I use Joe Celko's Data and Databases book, but this isn't suitable as an introduction.
Hmmm, you don't work in this field, do you?
...obscure architectures such as ARM do not require payment, just extra legwork. Have you seen .NET on ARM lately? I don't think so.
.NET allows developers to select the optimal hardware architecture for their projects by supporting a broad range of processor families, including x86, Xscale, ARM, MIPS and the SH series. Device bring-up is now faster than ever with Windows CE .NET providing numerous board support packages and features such as the Platform Wizard, which provides preconfigured design templates for embedded devices such as handhelds, smart phones, Web pads, retail point-of-sale devices and set-top boxes.
ARM will be the single most widely deployed PDA and phone CPU. Period.
.NET on ARM? Try here on the Microsoft site (of all places). To quote:
Windows CE
Your devastating critique continues:
What does this [deployment problems] have to do with anything? There are many, _many_ sites that are dedicated repositories of binaries for particular architectures.
There are? I think you'll find that most applications (as opposed to platform builds) are offered in a few builds that the authors know about, and finding any others is down to luck.
Let's try searching for 'linux tetris' in Google, shall we? OK, here's the first one. Oh dear! There's one build, and it's for x86 Mandrake. Care to point to one of your "many, many binary sites" that has it built for a PPC, for example?
And, lastly, our special bonus, the extra-insightful:
Having a VM is a waste of time... Any architecture can be virtualized
Riiiigghhtt. Well, I hope you've taken the opportunity to tell Microsoft, Sun and all those phone vendors that they're wasting their time pushing VMs. I bet they'll be kicking themselves soon! Perhaps you'd like to point to a phone or PDA that runs a foreign CPU instruction set, just as an example?