If only this device can deliver always on connection it would be a great gadget. Being an owner of the first version of Sidekick, however, I was disappointed on the connectivity. The problem in San Francisco Bay Area is Cingular's poor radio signal. The device frequently lose sync with the server. To resync it often takes one to two minutes. Everytime I click a hyperlink it may fetch in a reasonable amount of time or it may take many minutes. You are not going to enjoy web surfing given the unreliablity. I have adapted to the problem of this device. When I use it usually I have a book on one hand and the sidekick on the other hand.
And if you leave the device on overnight and you want to check the weather next morning. Chances are it has already lost sync. Wait another few minutes for it to get connected again. Sometimes you have to reboot the phone altogether in order to get connected.
A highly informative interview about the politics of municipal broadband. Solid opinion on issues that shapes the future of U.S. and aboard. An outline of an alternative to the pitiful service commercial companies are providing us today that called "broadband". This is a great read for anyone care about the future of our economy.
Why do we only get to moderate user comments? We should moderate article postings too and put great article above the mediocre ones.
Interesting comment about Minitel's retro character based UI. Sound rather similar to the range of *nix's terminal based applications. Despite the GUI there is always the large number of aficionados to those 30 years old technology!
Search google for many products available. One product we looked at is from TimesTen. It was very expensive!
The primary reason people use it is because of performance. They are situation where data update and access time is very important but the data is small enough to fit in memory and persistence is less important. One example is for keeping real time system status.
Today, without the language support, I ask every programmer to minimally add a comment to document the data type used in the collection, like
HashMap myTable;// username(String) => UserRecord
Still many people fail to do just that. I have to trace the typecast and the usage of the collection to discover what are they putting in the collection.:(
The proposal seems to keep Java collection's light -weight design. It simply provides some compile time checking and implicit typecasting. It avoids the complexity of C++'s template. I remember I have spent a lot time debugging the compilation. An error message would spend multiple lines with long and unreadable class names decompose of a template instances...
You question is addressed in Chapter 1
on
802.11 Security
·
· Score: 2, Informative
The phrase "wireless security" is considered by some to be an oxymoron. How can a system with no physical security hope to facilitate secure data transport? Well, with careful planning and configuration, a wireless network can protect itself from many types of attacks and become almost as secure as its wired counterpart. 802.11 can be deployed with various security mechanisms to provide robust, mobile, and hardened network infrastructure.
In sounds like common sense, if you want performance, port it to C, C++, etc. In practice, you may as well port it to assembly language to gain some performance.
Python has many high level data types such as tuple, list and dictionary as well as operators to make them useful. These datatype prove to be useful for a lot of programming tasks. This is why programmers find it so productive with Python. A simple pythonic operation such as getting an indexed item from a dictionary would be translated into many many lines of C code. Perhaps hundreds of lines if you want an efficient implementation. It just wouldn't be the same program if you rewrite it in C.
Nothing against rewriting code in different language. I just want to point out that one is lot more high level then the other. For me Python is a good choice for tasks where programmer's time is more important than CPU time.
XML is obviously not as compact as binary data format. The question is what is the most important attribute for a data format? Readability and maintainability or efficiency and compactness. The popularity of XML means today people value readability over compactness.
Is it silly to said XML is compressible when it is verbose to start with? I think this really separate the design into two layers. XML provides data presentation and interoperability. Compression can be applied on top when necessary. Since an XML and an equivalent binary file contain the same information, they should be of similar size in compressed form.
Don't forget bandwidth and processor speed will only get better. Today's efficiency concern would be irrelevant in the future. Access to legacy data would be a much bigger issue.
The memcmp() bug in the interview seems simple enough. May I ask a novice question? What is unchecked buffer size problem? How can you take over a system because of an unchecked buffer? Is it a C/C++ specific problem? Thanks for enlighting me.
After spending a good part of the day pulling my hair trying to help my wife read some files from a floppy disk, I figure that the fun must be in a pervert sense. (what is the device name to be used for 'mount'??? Let me google it... damn, do I need to give her root password so that she can 'su'...)
I gave her a Windows XP. She use it to download pictures from digicam and to organize them. I get some peace of mind and have my sanity preserved. Perhaps this can be called fun.
Linux is great for many purposes. But a viable and fun OS for the rest of us it is not.
Hard to diagnosis. Very hard to visualize the XSLT processing. And then there is the XMLized scripting language you have to learn. I wonder how many people uses XSLT vs printf() for generating XML.
The API that you describe exists. It is XPath. The next post of a C# example illustrate a possible use of XPath API.
XPath is especially great for getting a single value. It elminate that need to walk a DOM tree of or use callback. However it does not help on a more general case of stream processing a XML file.
Back in the Win95 days I used Borland Delphi to build UI. I like them lot better than MS tools. One of the advantage is it support layout management.
self absorbed in his craftsman analogy
on
Software Craftsmanship
·
· Score: 2, Interesting
I certainly agree that a small team of talented developers are more productive than horde of average programmer. What I'm looking for in his book is some hard research on why this works better, some real world cases and some partical advice on how to make this work better. I find little of these in this book. Instead the author is just self absorbed in his craftsman analogy. The main theme: a bunch of apprentices would work with a craftsman to create quality product that the craftsman would personally sign off. That's a way too simplistic idea to improve software development by putting people in the apprentice-journeymen-craftsman role.
Big regret to have spent money and time on this book.
Unfortunately when faced the choice between performance and battery most vendor bias toward performance. I am a programmer spending a lot of time on commute train. 3+ hour is what I demand. I spend more time on text editor and web browser than any other thing I don't really need high speed in that circumstance. Unfortunately with CPU speed turned low (Intel SpeedStep?) I still get only around 2 hours, much less than my PIII.
With the interest shift to building web based applications in the last few years, client side application is not as important as it use to be. Even MFC is somewhat dead.
I hope the court ruling on bundling up-to-date version of JVM would bring more life to client side Java. I like to see it fulfill the cross-platform promise. I'm looking forward for a great, secure email client to replace MS outlook.
"Actually, we rewrote JUnit's basic functionality."
Aha! You build a new framework because you are not happy with JUnit's "functionality". It true that JUnit's "functionality", mostly in the runner, is not super strong. I don't think they mean to provide anything more than a few basic runners. It is mean to be extended. I wrote my own runner that setup my environment, parse my own suite file and output HTML reports. I ask myself am I still using JUnit when I have built 50% of "functionality" myself? In the end the value of JUnit do not lie in their code, which is thankfully simple. It is the concept of building unit testing modules and do it as part of the development process that is really innovative. So whether one use JUnit's or SuiteRunner is not that important. It would be win for software engineering in any case!
Wai Yip Tung
Jamie's writing larger reflect my Linux experience
on
JWZ Reviews Video on Linux
·
· Score: 3, Insightful
That Linux is flooded with low quality apps.
1. Many apps can't get basic thing done easily (or at all). E.g. Nautilus (a file manager) could crash when copying files.
2. Many GUI are badly designed. They have complex interaction and fail to guide new users to do basic things. Think how many steps are needed to use xcdroast to duplicate a CD. How intimidating it would be to a new user. And how many opportunities for user to do wrong things.
3. Many GUI apps look amateurish. And when some window doesn't fit in a 1024x768 screen this just drive you mad.
4. When an app exhaust my patience I go for a different app. Only to find it have its own set of problem or sucks even more. And then I still can't even get my basic work done.
The enthusiastic crowd of Linux would insist the app works would great if only you do this configure and/or use a different version and/or recompile from source code. We need to get real and have a objective evaluation on the state of art. If we oversold on Linux and it doesn't meet the quality standard for average user it would only damage Linux's image.
Note that after 2 months of frustrating experience and I still in a quixotic attempt to get Linux to work for me. It is only because I'm serious in finding an alternative to MS. If any windoze app give me this kind of crap it will be uninstalled and will never be seen again. (But I think my next machine is going to be a Mac).
Java detect some dead code and treat it as compiler error (Java Language Specification 14.20 - Unreachable Statements).
In the paper, some cases classified as redundant assignment are also dead code cases. Fig 3 is a dead code case. Fig 2 is would be a dead code case instead if it is a more straight forward link list traversal case. E.g.
for (entry=priv->lec_arp_tables[i]; entry != NULL; entry = entry->next) { /* next = entry->next; */ .... return 0; }
the expression entry = entry->next would become dead code.
Java actually make some dead code a compilation error (The Java Language Specification 14.20).
The author claim redundant assignment signal most bug. Interestly I think some of them can be classified as dead code as well. Fig.3 is obviously contain dead code. For Fig.2, if the example were a more stright forward link list traversal instead of deleting, it would become a dead code case. E.g.
for (entry = xxx;
entry != NULL
entry = entry->next) {/* next = entry->next the redundant assigment is not needed for simple traversal */...
return 0;
}
The entry=entry->next would be a dead code. No redundant assignment in this case.
Wai Yip Tung
I have the pleasure to meet Mr O'Reilly in one of his speaking event. It was an insightful and throught provoking event. What set O'Reilly and other publisher apart is that they are truely interested in technology and even take a stance on issues (such as Amazon's 1-click patent).
Many times I visit a book store I go straight to O'Reilly rack. Can't say every book has consistent highest quality. But at least they are thinner that other competitor's book and waste less paper on dumb topics. Keep on! We will continue to explore the world of technology together!
The PC Case is just ugly. I'm considering to buy an iMac, partly because of it look, seriously. I'm a hardcore programmer. But Mac's look is just irresistible. I think the Unix core make it a partical machine for coding (besides web browsing, etc).
Slashdot Mirror
If only this device can deliver always on connection it would be a great gadget. Being an owner of the first version of Sidekick, however, I was disappointed on the connectivity. The problem in San Francisco Bay Area is Cingular's poor radio signal. The device frequently lose sync with the server. To resync it often takes one to two minutes. Everytime I click a hyperlink it may fetch in a reasonable amount of time or it may take many minutes. You are not going to enjoy web surfing given the unreliablity. I have adapted to the problem of this device. When I use it usually I have a book on one hand and the sidekick on the other hand.
And if you leave the device on overnight and you want to check the weather next morning. Chances are it has already lost sync. Wait another few minutes for it to get connected again. Sometimes you have to reboot the phone altogether in order to get connected.
Wai Yip
This posting deserves 5 points!
A highly informative interview about the politics of municipal broadband. Solid opinion on issues that shapes the future of U.S. and aboard. An outline of an alternative to the pitiful service commercial companies are providing us today that called "broadband". This is a great read for anyone care about the future of our economy.
Why do we only get to moderate user comments? We should moderate article postings too and put great article above the mediocre ones.
Wai Yip Tung
Interesting comment about Minitel's retro character based UI. Sound rather similar to the range of *nix's terminal based applications. Despite the GUI there is always the large number of aficionados to those 30 years old technology!
Search google for many products available. One product we looked at is from TimesTen. It was very expensive!
The primary reason people use it is because of performance. They are situation where data update and access time is very important but the data is small enough to fit in memory and persistence is less important. One example is for keeping real time system status.
Generics should have come much sooner.
// username(String) => UserRecord
:(
Today, without the language support, I ask every programmer to minimally add a comment to document the data type used in the collection, like
HashMap myTable;
Still many people fail to do just that. I have to trace the typecast and the usage of the collection to discover what are they putting in the collection.
The proposal seems to keep Java collection's light -weight design. It simply provides some compile time checking and implicit typecasting. It avoids the complexity of C++'s template. I remember I have spent a lot time debugging the compilation. An error message would spend multiple lines with long and unreadable class names decompose of a template instances...
The phrase "wireless security" is considered by some to be an oxymoron. How can a system with no physical security hope to facilitate secure data transport? Well, with careful planning and configuration, a wireless network can protect itself from many types of attacks and become almost as secure as its wired counterpart. 802.11 can be deployed with various security mechanisms to provide robust, mobile, and hardened network infrastructure.
In sounds like common sense, if you want performance, port it to C, C++, etc. In practice, you may as well port it to assembly language to gain some performance.
Python has many high level data types such as tuple, list and dictionary as well as operators to make them useful. These datatype prove to be useful for a lot of programming tasks. This is why programmers find it so productive with Python. A simple pythonic operation such as getting an indexed item from a dictionary would be translated into many many lines of C code. Perhaps hundreds of lines if you want an efficient implementation. It just wouldn't be the same program if you rewrite it in C.
Nothing against rewriting code in different language. I just want to point out that one is lot more high level then the other. For me Python is a good choice for tasks where programmer's time is more important than CPU time.
XML is obviously not as compact as binary data format. The question is what is the most important attribute for a data format? Readability and maintainability or efficiency and compactness. The popularity of XML means today people value readability over compactness.
Is it silly to said XML is compressible when it is verbose to start with? I think this really separate the design into two layers. XML provides data presentation and interoperability. Compression can be applied on top when necessary. Since an XML and an equivalent binary file contain the same information, they should be of similar size in compressed form.
Don't forget bandwidth and processor speed will only get better. Today's efficiency concern would be irrelevant in the future. Access to legacy data would be a much bigger issue.
The memcmp() bug in the interview seems simple enough. May I ask a novice question? What is unchecked buffer size problem? How can you take over a system because of an unchecked buffer? Is it a C/C++ specific problem? Thanks for enlighting me.
After spending a good part of the day pulling my hair trying to help my wife read some files from a floppy disk, I figure that the fun must be in a pervert sense. (what is the device name to be used for 'mount'??? Let me google it... damn, do I need to give her root password so that she can 'su'...)
I gave her a Windows XP. She use it to download pictures from digicam and to organize them. I get some peace of mind and have my sanity preserved. Perhaps this can be called fun.
Linux is great for many purposes. But a viable and fun OS for the rest of us it is not.
Hard to diagnosis. Very hard to visualize the XSLT processing. And then there is the XMLized scripting language you have to learn. I wonder how many people uses XSLT vs printf() for generating XML.
The API that you describe exists. It is XPath. The next post of a C# example illustrate a possible use of XPath API.
XPath is especially great for getting a single value. It elminate that need to walk a DOM tree of or use callback. However it does not help on a more general case of stream processing a XML file.
Back in the Win95 days I used Borland Delphi to build UI. I like them lot better than MS tools. One of the advantage is it support layout management.
I certainly agree that a small team of talented developers are more productive than horde of average programmer. What I'm looking for in his book is some hard research on why this works better, some real world cases and some partical advice on how to make this work better. I find little of these in this book. Instead the author is just self absorbed in his craftsman analogy. The main theme: a bunch of apprentices would work with a craftsman to create quality product that the craftsman would personally sign off. That's a way too simplistic idea to improve software development by putting people in the apprentice-journeymen-craftsman role.
Big regret to have spent money and time on this book.
Mac is probably a more viable choice. Although I don't think the $2.4mil can justify the switch.
Although here is slashdot, I would still say Linux is not a solution for desktop computing.
Unfortunately when faced the choice between performance and battery most vendor bias toward performance. I am a programmer spending a lot of time on commute train. 3+ hour is what I demand. I spend more time on text editor and web browser than any other thing I don't really need high speed in that circumstance. Unfortunately with CPU speed turned low (Intel SpeedStep?) I still get only around 2 hours, much less than my PIII.
With the interest shift to building web based applications in the last few years, client side application is not as important as it use to be. Even MFC is somewhat dead.
I hope the court ruling on bundling up-to-date version of JVM would bring more life to client side Java. I like to see it fulfill the cross-platform promise. I'm looking forward for a great, secure email client to replace MS outlook.
"Actually, we rewrote JUnit's basic functionality."
Aha! You build a new framework because you are not happy with JUnit's "functionality". It true that JUnit's "functionality", mostly in the runner, is not super strong. I don't think they mean to provide anything more than a few basic runners. It is mean to be extended. I wrote my own runner that setup my environment, parse my own suite file and output HTML reports. I ask myself am I still using JUnit when I have built 50% of "functionality" myself? In the end the value of JUnit do not lie in their code, which is thankfully simple. It is the concept of building unit testing modules and do it as part of the development process that is really innovative. So whether one use JUnit's or SuiteRunner is not that important. It would be win for software engineering in any case!
Wai Yip Tung
That Linux is flooded with low quality apps.
1. Many apps can't get basic thing done easily (or at all). E.g. Nautilus (a file manager) could crash when copying files.
2. Many GUI are badly designed. They have complex interaction and fail to guide new users to do basic things. Think how many steps are needed to use xcdroast to duplicate a CD. How intimidating it would be to a new user. And how many opportunities for user to do wrong things.
3. Many GUI apps look amateurish. And when some window doesn't fit in a 1024x768 screen this just drive you mad.
4. When an app exhaust my patience I go for a different app. Only to find it have its own set of problem or sucks even more. And then I still can't even get my basic work done.
The enthusiastic crowd of Linux would insist the app works would great if only you do this configure and/or use a different version and/or recompile from source code. We need to get real and have a objective evaluation on the state of art. If we oversold on Linux and it doesn't meet the quality standard for average user it would only damage Linux's image.
Note that after 2 months of frustrating experience and I still in a quixotic attempt to get Linux to work for me. It is only because I'm serious in finding an alternative to MS. If any windoze app give me this kind of crap it will be uninstalled and will never be seen again. (But I think my next machine is going to be a Mac).
Wai Yip Tung
Java detect some dead code and treat it as compiler error (Java Language Specification 14.20 - Unreachable Statements).
In the paper, some cases classified as redundant assignment are also dead code cases. Fig 3 is a dead code case. Fig 2 is would be a dead code case instead if it is a more straight forward link list traversal case. E.g.
the expression entry = entry->next would become dead code.
Wai Yip Tung
Java actually make some dead code a compilation error (The Java Language Specification 14.20). The author claim redundant assignment signal most bug. Interestly I think some of them can be classified as dead code as well. Fig.3 is obviously contain dead code. For Fig.2, if the example were a more stright forward link list traversal instead of deleting, it would become a dead code case. E.g. for (entry = xxx; entry != NULL entry = entry->next) { /* next = entry->next the redundant assigment is not needed for simple traversal */ ...
return 0;
}
The entry=entry->next would be a dead code. No redundant assignment in this case.
Wai Yip Tung
I have the pleasure to meet Mr O'Reilly in one of his speaking event. It was an insightful and throught provoking event. What set O'Reilly and other publisher apart is that they are truely interested in technology and even take a stance on issues (such as Amazon's 1-click patent). Many times I visit a book store I go straight to O'Reilly rack. Can't say every book has consistent highest quality. But at least they are thinner that other competitor's book and waste less paper on dumb topics. Keep on! We will continue to explore the world of technology together!
The PC Case is just ugly. I'm considering to buy an iMac, partly because of it look, seriously. I'm a hardcore programmer. But Mac's look is just irresistible. I think the Unix core make it a partical machine for coding (besides web browsing, etc).