The added benefits of tabs (compared to having just desktops):
Unlike desktops, tabs have visible names dynamically adjusted to the content they display.
Tabs supply another level of hierarchy for the grouping that is visually distinctive.
Tabs are automatically configured to group similar tasks (be it the same application or not) while desktops are manually configured on a per-window basis.
Its great that KDE avoided code duplication, but it is not only that I was talking about.
The feature should not be implemented at the application-level at all, but at the window manager level, as one window manager already does (partially), noted by the other comments.
This would allow no work to be required for tabs to exist, and exist merely as a configuration option by the user. This would also allow non-kde apps to benefit from tabs.
Opera added tabs. That was neat because it can let the user group similar tasks (web browsing pages) together hierarchially under the task of web browsing (and unlike grouping in Microsoft's/KDE's taskbar, remain one click away when in the browser).
Mozilla added tabs, that was also neat.
Konqueror added tabs, this was not neat! KDE's people, unlike Opera's or Mozilla's are in the exact right position to have a bit more of a vision, and encorporate tabs into KDE's general facilities, and not just a specific program (web browser). Instead, KDE's people choose to incorporate tabs separately in Konqeruror, Konsole, and other programs, such that non-KDE applications cannot benefit from it. Now it seems as though Microsoft is just as short-sighted and added tabs to Internet Explorer instead of adding tabs to the core window-switching facilities (by drawing a tab under title bars of a new concept of "window-group" that contains multiple windows of same applications or such).
What I believe should have been done, is something more along the lines of what was done with Mouse Gestures in KDE. Mouse Gestures in KDE are handled by a general facility (KHotKeys) such that not only Konqueror can benefit from it, but any KDE/non-KDE application.
Using passwords which are correct English sentences isn't much better.
Correct English sentences have about 1.2 bits per character. That means that for 10 words of 5 characters each, you have 50 characters which are 60 entropic bits (~7.5 entropic bytes).
That is as strong as a 10-character password, or so, but much much longer.
Not sure this is the solution. I think that whatever is easy to remember, is easy to remember because it has low entropy and is easy to attack.
The solution might be to use non-human memory? USB disk-on-keys containing crypto keys?
.Net isn't a GUI development solution. There's a lot of different resources out there for understanding what it is, but you can start here with the basics.
Thank you, King Obvious. I was obviously talking about the GUI libraries and RAD tools available for.NET.
but from what I can see, it's Python-based. That's fine if you know Python (and I do), but not everyone does
Qt is a C++ library usable from many languages. PyQt is its Python bindings. Not everyone does, but everyone should, because it is the right language for GUI and most of the other non-performance-critical applications.
VS.Net offers 3 or 4 different development languages that easily work together.
Besides for Python.NET, the rest are statically typed and otherwise very difficult to program in (compared to Python, ofcourse) and thus a very poor choice for high-level code such as GUI code.
YOU like PyQt/PyKDE, but that doesn't make it "far better" for everyone's purposes. It just means that you like it.
No, PyQt/PyKDE are better objectively in terms of API consistency, simplicity, power and expresiveness.
When I like something I say I like it. PyQt/PyKDE are simply better than all Microsoft GUI solutions out there.
And your point is basically immaterial to the main thread: there's a ton of support and information for and about developing Windows GUI applications, that is, the full application chain from front-end to back, and there's a paucity of such information for Linux
That is because a ton of information is needed to program Windows GUI applications as the Windows API's are so horribly broken and difficult to use.
As for Linux lacking documentation (more accurately PyQt/PyKDE as those are portable to non-Linux systems unlike Microsoft GUI tools), I did try a Google search and this is the first result.
Note that this documentation, unlike the MSDN has a consistent format, is usable as a reference and as a tutorial and is comprehensive and complete uniformly across all aspects of Qt. Can you claim this about Microsoft's documentation?
Actually, in my code, code coverage tools always show at least 90% coverage by my unit tests...
I use neat tricks and tools to simulate failures though, that I invented myself:-) Maybe others have less powerful ways to simulate the external environment...
His article simply tries to state the code coverage does not guarantee 0 bugs, which is ofcourse true, but completely off the point. Also, the "little return on investment" is simply untrue in my experience, because the investment is not as huge as they claim - unless lacking the proper tools.
(Did you notice the linked article author is still in the dark Hungarian notation days?)
I think the point is not that Unit Tests are a good substitute for static type checks.
I think the point is that static type checks are not a good substitute for Unit Tests.
I trust a line that has never run, but passed static type tests pretty much the same as I trust a line that has never run and passed no tests - that is, I don't trust either.
Type inference sure is nice... But its main feature is performance. If you dont care about performance, use dynamic typing. Most programmers don't really code for performance..
The real distinction is between the Open Source movement and the Free Software movement. Both movements make software that is open-source and is free software.
The difference is in the motivations:
Free Software is motivated by the moral need to create a Free Way to use computers -- to free software users from their "masters".
Open Source software is motivated by the practical advantages of the Open Source development process.
The Free Software movement is more idealist: "Don't use it if its not free, whether or not there are practical advantages". The Open Source movement is more pragmatic, even at the cost of some Freedom: "Use whatever is better technically for your purpose, even if its not free".
As I pick up my jaw after reading his more than reasonable approach to porting Free software to proprietary platforms, I wonder why he allows this freedom, (to use your desired platform) but argues against using a best of breed code management system for the same reason.
Same reason??
He argues porting Free Software is to Non-Free systems is good because it exposes users to Free Software increasing their freedom and the chance that they move to a Free System.
He argues using Non-Free Software to manage the Linux Kernel is bad because it sets a bad example for Free Software developers and because it takes away the freedom of the users of the Source Control software.
Either you haven't read the reasons at all, or you read them and still thought they were the same, in which case you are a complete retard and better stop reading RMS since you won't comprehend what he's saying anyway.
But a CS degree does not really help you be a better programmer in the real world, for the most part:-)
Well, that's not 100% accurate, ofcourse a degree helps you program better, but so will a physics degree, a maths degree, and other degrees. CS and real-world programming are very indirectly and loosely connected.
The main points that separate good programmers from the bad:
Experience, experience, lots of experience.
Talent/Intelligence (These matter less for most (but not all) programming work). However, its that difficult task here and there that even an experienced programmer will fail, unless he also posseses the talent.
Ofcourse experience will help a lot less if you are gaining it without talent and far from a talented environment.
But you were probably not asking about how much a degree will help you be a better programmer, but how much it will help you get a better programming job. I'm sorry, but I've not seen enough job interviews to know that one:-)
"Object Oriented" means a lot of different things to a lot of different people.
If you cite examples of your designs or pieces of code of yours, I'm sure I can find some things that could look nicer or be better implemented via Python's OO mechanisms, and I am not just talking about the C++ kind of OO (encapsulation/inheritence/polymorphism) but also about Python's __getattr__ and other powerful features inherent in its OO design).
I tend to find that extremely competent programmers, with a lot of experience, tend to make nearly bug-free software...
Unfortunatly (or fortunatly for some of us:), the vast majority of programmers out there simply suck, bad. Judging by most faults I've seen, and despite what so many people say: MS programmers suck.
Stop saying people are "stealing" when they copy movies and other stuff against the law, when they are obviously committing an act of murder against the rights of the owners, and murder of creativity.
Copyright infringement is murder, and should be penalized accordingly!
The only insight I can read from your message is that maybe exceptions ought to be renamed, so that fools don't create silly rules about when they are to be used.
Exceptions are a powerful way/tool to return/propagate values on the call stack, and creating arbitrary limitations on them because they are named "exceptions" is silly.
Using exceptions for non-"exceptional" cases is often a lot more convinient/powerful than having complicated return values.
Slashdot Mirror
Actually that clause in the constitution of the United States allows the Congress to secure works for limited times, not the people.
The congress can choose to do this via copyright/patent laws, or it can choose not to because the constitution does not demand it.
Unlike desktops, tabs have visible names dynamically adjusted to the content they display.
Tabs supply another level of hierarchy for the grouping that is visually distinctive.
Tabs are automatically configured to group similar tasks (be it the same application or not) while desktops are manually configured on a per-window basis.
Its great that KDE avoided code duplication, but it is not only that I was talking about.
The feature should not be implemented at the application-level at all, but at the window manager level, as one window manager already does (partially), noted by the other comments.
This would allow no work to be required for tabs to exist, and exist merely as a configuration option by the user. This would also allow non-kde apps to benefit from tabs.
Opera added tabs. That was neat because it can let the user group similar tasks (web browsing pages) together hierarchially under the task of web browsing (and unlike grouping in Microsoft's/KDE's taskbar, remain one click away when in the browser).
Mozilla added tabs, that was also neat.
Konqueror added tabs, this was not neat! KDE's people, unlike Opera's or Mozilla's are in the exact right position to have a bit more of a vision, and encorporate tabs into KDE's general facilities, and not just a specific program (web browser).
Instead, KDE's people choose to incorporate tabs separately in Konqeruror, Konsole, and other programs, such that non-KDE applications cannot benefit from it.
Now it seems as though Microsoft is just as short-sighted and added tabs to Internet Explorer instead of adding tabs to the core window-switching facilities (by drawing a tab under title bars of a new concept of "window-group" that contains multiple windows of same applications or such).
What I believe should have been done, is something more along the lines of what was done with Mouse Gestures in KDE. Mouse Gestures in KDE are handled by a general facility (KHotKeys) such that not only Konqueror can benefit from it, but any KDE/non-KDE application.
This is what should be done with tabs!
Using passwords which are correct English sentences isn't much better.
Correct English sentences have about 1.2 bits per character. That means that for 10 words of 5 characters each, you have 50 characters which are 60 entropic bits (~7.5 entropic bytes).
That is as strong as a 10-character password, or so, but much much longer.
Not sure this is the solution.
I think that whatever is easy to remember, is easy to remember because it has low entropy and is easy to attack.
The solution might be to use non-human memory? USB disk-on-keys containing crypto keys?
This is Slashdot, nobody expects real discussion here, lighten up :)
Slashdot was always a contest of who yells more towards which side, and I find it fun sometimes to participate in it..
Anyhow I was just responding in the same spirit of "Sadly Windows is far ahead of Linux in the GUI and development department" with the opposite.
If a real discussion is due, I can participate in that as well.
Ruby is a nice language too, but too bad all the sheeple are stuck with C# and other Microsoft-said-so languages.
.Net isn't a GUI development solution. There's a lot of different resources out there for understanding what it is, but you can start here with the basics.
.NET.
Thank you, King Obvious. I was obviously talking about the GUI libraries and RAD tools available for
but from what I can see, it's Python-based. That's fine if you know Python (and I do), but not everyone does
Qt is a C++ library usable from many languages. PyQt is its Python bindings. Not everyone does, but everyone should, because it is the right language for GUI and most of the other non-performance-critical applications.
VS.Net offers 3 or 4 different development languages that easily work together.
Besides for Python.NET, the rest are statically typed and otherwise very difficult to program in (compared to Python, ofcourse) and thus a very poor choice for high-level code such as GUI code.
YOU like PyQt/PyKDE, but that doesn't make it "far better" for everyone's purposes. It just means that you like it.
No, PyQt/PyKDE are better objectively in terms of API consistency, simplicity, power and expresiveness.
When I like something I say I like it. PyQt/PyKDE are simply better than all Microsoft GUI solutions out there.
And your point is basically immaterial to the main thread: there's a ton of support and information for and about developing Windows GUI applications, that is, the full application chain from front-end to back, and there's a paucity of such information for Linux
That is because a ton of information is needed to program Windows GUI applications as the Windows API's are so horribly broken and difficult to use.
As for Linux lacking documentation (more accurately PyQt/PyKDE as those are portable to non-Linux systems unlike Microsoft GUI tools), I did try a Google search and this is the first result.
Note that this documentation, unlike the MSDN has a consistent format, is usable as a reference and as a tutorial and is comprehensive and complete uniformly across all aspects of Qt. Can you claim this about Microsoft's documentation?
Qt, KDE and PyQt/PyKDE are far better GUI development solutions than anything Microsoft has ever put out, .NET included.
Contrary to common belief.. Random bits "installed" on the hard disk do not affect operation of the machine, unless explicitly accessed.
In other words, you don't have to run those programs!
Starcraft single player is worthless :-P
Actually, in my code, code coverage tools always show at least 90% coverage by my unit tests...
:-)
I use neat tricks and tools to simulate failures though, that I invented myself
Maybe others have less powerful ways to simulate the external environment...
His article simply tries to state the code coverage does not guarantee 0 bugs, which is ofcourse true, but completely off the point.
Also, the "little return on investment" is simply untrue in my experience, because the investment is not as huge as they claim - unless lacking the proper tools.
(Did you notice the linked article author is still in the dark Hungarian notation days?)
I think the point is not that Unit Tests are a good substitute for static type checks.
I think the point is that static type checks are not a good substitute for Unit Tests.
I trust a line that has never run, but passed static type tests pretty much the same as I trust a line that has never run and passed no tests - that is, I don't trust either.
Type inference sure is nice... But its main feature is performance.
If you dont care about performance, use dynamic typing.
Most programmers don't really code for performance..
Just clears up the obstacles, politics and beauracracy that could slow down his employees and then gets the hell out of the way...
(Assuming good workers ofcourse)
The difference is in the motivations:
Free Software is motivated by the moral need to create a Free Way to use computers -- to free software users from their "masters".
Open Source software is motivated by the practical advantages of the Open Source development process.
The Free Software movement is more idealist: "Don't use it if its not free, whether or not there are practical advantages".
The Open Source movement is more pragmatic, even at the cost of some Freedom: "Use whatever is better technically for your purpose, even if its not free".
As I pick up my jaw after reading his more than reasonable approach to porting Free software to proprietary platforms, I wonder why he allows this freedom, (to use your desired platform) but argues against using a best of breed code management system for the same reason.
Same reason??
He argues porting Free Software is to Non-Free systems is good because it exposes users to Free Software increasing their freedom and the chance that they move to a Free System.
He argues using Non-Free Software to manage the Linux Kernel is bad because it sets a bad example for Free Software developers and because it takes away the freedom of the users of the Source Control software.
Either you haven't read the reasons at all, or you read them and still thought they were the same, in which case you are a complete retard and better stop reading RMS since you won't comprehend what he's saying anyway.
I did the first make-based boot system, IBM did it later :-)
Well, that's not 100% accurate, ofcourse a degree helps you program better, but so will a physics degree, a maths degree, and other degrees. CS and real-world programming are very indirectly and loosely connected.
The main points that separate good programmers from the bad:
Experience, experience, lots of experience.
Talent/Intelligence (These matter less for most (but not all) programming work). However, its that difficult task here and there that even an experienced programmer will fail, unless he also posseses the talent.
:-)
Ofcourse experience will help a lot less if you are gaining it without talent and far from a talented environment.
But you were probably not asking about how much a degree will help you be a better programmer, but how much it will help you get a better programming job. I'm sorry, but I've not seen enough job interviews to know that one
"Object Oriented" means a lot of different things to a lot of different people.
If you cite examples of your designs or pieces of code of yours, I'm sure I can find some things that could look nicer or be better implemented via Python's OO mechanisms, and I am not just talking about the C++ kind of OO (encapsulation/inheritence/polymorphism) but also about Python's __getattr__ and other powerful features inherent in its OO design).
I tend to find that extremely competent programmers, with a lot of experience, tend to make nearly bug-free software...
:), the vast majority of programmers out there simply suck, bad. Judging by most faults I've seen, and despite what so many people say: MS programmers suck.
Unfortunatly (or fortunatly for some of us
The google cache is a copyright breach?
If that is so, its because they make "unauthorized" copies of the web page by retransmitting it to anyone who wants it.
However, if you think about it - so does every goddamn router on the path the page goes through.
So maybe routers are the biggest breach of copyrights?
I decided that I shall invest in stocks of companies that will be successful.
That way, when they are successful, I get richer!
Ingenious, isn't it?
Stop saying people are "stealing" when they copy movies and other stuff against the law, when they are obviously committing an act of murder against the rights of the owners, and murder of creativity.
Copyright infringement is murder, and should be penalized accordingly!
Shapiro is the guy working on a research Operating System project (The EROS system).
EROS was originally implemented in C++, but then
it was reimplemented in C.
The only insight I can read from your message is that maybe exceptions ought to be renamed, so that fools don't create silly rules about when they are to be used.
Exceptions are a powerful way/tool to return/propagate values on the call stack, and creating arbitrary limitations on them because they are named "exceptions" is silly.
Using exceptions for non-"exceptional" cases is often a lot more convinient/powerful than having complicated return values.