Will this make Inprise less inclined to port it's development tools to Linux? (Interbase is already ported). C++ Builder or Delphi for Linux would have been nice.
This is not embrace and extend
on
Corel Linux FAQ
·
· Score: 2
If Corel Office only worked with the Corel Linux distro, that would be embrace and extend -- lock people into the distro just because they want the office suite. As long as the office suite runs on all distros, which is their stated plan, there's no danger of that.
I don't think the reviewer was against writing memory allocators, per see. I think he was objecting to the fact that the book author didn't integrate the memory allocator into new and delete. By overriding new and delete, as the reviewer suggested, you are in fact writing a new memory allocator (or at least customizing the standard one). But I have to admit, that part of the review confused me a bit.
Some of the things the reviewer objected to (page long inline functions, premature optimizations that didn't achieve anything, etc.) really did sound remarkably stupid. What's the point of optimizing a structure from 16 to 13 bytes (to save 3 bytes!!! whoopee!!!), only to have the compiler pad it to 16 anyway? You've broken the "rules" (or, more accurately, the standard guidelines) and achieved absolutely nothing. You're code is more complicated and harder to follow, but you've gained nothing in efficiency. To quote Donald Knuth, "Premature optimization is the root of all evil." Coding like this is what gives C++ a bad name. Yes, it might (through some stroke of luck) still work, but what's the point? Why not code properly, make it work even better, and make the code more maintainable?
I really, really liked this review. It was critical, and critics are supposed to be critical. Of course, some people my disagree with the reviewer, and that's valid too.
I suppose that Gnome hitting 1.0 just before RedHat 6.0 was released was a coincidence? And the fact that RedHat employs Gnome coders gives it no influence over things like release dates?
Regarding your "it was good that we released buggy software as a 1.0 release because that makes it better" argument, are you advocating this as a general policy? Do you fault the KDE people for waiting before the stuff is stable before releasing it? Do you think the 2.2 kernel was released too late, that it would have been better to release it earlier, when it was buggier? Do you admire Microsoft's strategy of releasing buggy stuff, because it makes it better? And do you beleive Microsoft when they say that release dates have nothing to do with marketing considerations? If not, why should we beleive that Gnome releasing buggy software too early had nothing with marketing? Do you think that this strategy helps Linux and Gnome's reputation?
I agree. The article was typical Village-Voice political correctness, but with a twist -- putting down victomization politics because the victims are white males. This shows the emptiness and hypocrisy of pc politics.
Of course, part of being an adult is getting past feeling sorry for yourself, and these kids will have to learn that. But they are just kids, so give them time! We do need to understand what went wrong in the killers heads. If we just say "well, its just the problems of privileged white males, so ignore it," more of this things will occurr. This does NOT excuse the killers, of course.
I agree. The best protection is strong encryption (which government's should allow).
Re:Silent cries from distant places
on
The KDE Future
·
· Score: 1
Linux on the cutting edge? Linux started by making a Unix clone, instead of some fancy whiz bang micro-kernel thingy. That's part of what made Linux so stable -- instead of trying to soemthing really rad, Linus copied something that was proven.
On the desktop, Linux and Unix have some catching up to do.
Re:Some Thoughts on KDE/GNOME
on
The KDE Future
·
· Score: 1
Amen
Re:GNOME/KDE what's the big difference?
on
The KDE Future
·
· Score: 1
>
That makes a BIG difference to newbies.
Re:Internet transparency, document embedding, ...
on
The KDE Future
·
· Score: 1
You can bloat in any language, including C (witness Gnome). Object oriented programming, in any language, including C, can add some bloat if you're not careful. Doing object-oriented programming in C is awkward and error-prone, and can add it's own bloat. The nice thing about C++, as opposed to other OO languages like Java, Python, and Smalltalk, is that it takes a pay when you use it approach to bloating -- only if you use a certain OO feature (like virtual functions) do you pay the performance penalty.
Re:Some Thoughts on KDE/GNOME
on
The KDE Future
·
· Score: 1
Fair enough. But some of the posters we're complaing about Gnome people (Miguel) NOT giving credit where credit is due (like the infamous "legacy" comment, or Miguel's "Gnome is technically superior" to KDE FUD-comment -- how does more bugs and less features == technically superior?)
Also, I don't think you can call Gnome the underdog. Technically, yes, it has a long way to go before it matches KDE. But it has corporate funding from RedHat (which is partially funded by Intel, IBM, etc.). As such, Gnome does get a lot more mainstream press coverage. Until a few months ago, Gnome also got a lot more Slashdot coverage. So in that sense (resources, press coverage), Gnome has the advantage and KDE is the underdog.
Slashdot Mirror
Will this make Inprise less inclined to port it's development tools to Linux? (Interbase is already ported). C++ Builder or Delphi for Linux would have been nice.
If Corel Office only worked with the Corel Linux distro, that would be embrace and extend -- lock people into the distro just because they want the office suite. As long as the office suite runs on all distros, which is their stated plan, there's no danger of that.
I don't think the reviewer was against writing memory allocators, per see. I think he was objecting to the fact that the book author didn't integrate the memory allocator into new and delete. By overriding new and delete, as the reviewer suggested, you are in fact writing a new memory allocator (or at least customizing the standard one). But I have to admit, that part of the review confused me a bit.
Some of the things the reviewer objected to (page long inline functions, premature optimizations that didn't achieve anything, etc.) really did sound remarkably stupid. What's the point of optimizing a structure from 16 to 13 bytes (to save 3 bytes!!! whoopee!!!), only to have the compiler pad it to 16 anyway? You've broken the "rules" (or, more accurately, the standard guidelines) and achieved absolutely nothing. You're code is more complicated and harder to follow, but you've gained nothing in efficiency. To quote Donald Knuth, "Premature optimization is the root of all evil." Coding like this is what gives C++ a bad name. Yes, it might (through some stroke of luck) still work, but what's the point? Why not code properly, make it work even better, and make the code more maintainable?
I really, really liked this review. It was critical, and critics are supposed to be critical. Of course, some people my disagree with the reviewer, and that's valid too.
I suppose that Gnome hitting 1.0 just before RedHat 6.0 was released was a coincidence? And the fact that RedHat employs Gnome coders gives it no influence over things like release dates?
Regarding your "it was good that we released buggy software as a 1.0 release because that makes it better" argument, are you advocating this as a general policy? Do you fault the KDE people for waiting before the stuff is stable before releasing it? Do you think the 2.2 kernel was released too late, that it would have been better to release it earlier, when it was buggier? Do you admire Microsoft's strategy of releasing buggy stuff, because it makes it better? And do you beleive Microsoft when they say that release dates have nothing to do with marketing considerations? If not, why should we beleive that Gnome releasing buggy software too early had nothing with marketing? Do you think that this strategy helps Linux and Gnome's reputation?
I agree. The article was typical Village-Voice political correctness, but with a twist -- putting down victomization politics because the victims are white males. This shows the emptiness and hypocrisy of pc politics.
Of course, part of being an adult is getting past feeling sorry for yourself, and these kids will have to learn that. But they are just kids, so give them time! We do need to understand what went wrong in the killers heads. If we just say "well, its just the problems of privileged white males, so ignore it," more of this things will occurr. This does NOT excuse the killers, of course.
I agree. The best protection is strong encryption (which government's should allow).
Linux on the cutting edge? Linux started by making a Unix clone, instead of some fancy whiz bang micro-kernel thingy. That's part of what made Linux so stable -- instead of trying to soemthing really rad, Linus copied something that was proven.
On the desktop, Linux and Unix have some catching up to do.
Amen
>
That makes a BIG difference to newbies.
You can bloat in any language, including C (witness Gnome). Object oriented programming, in any language, including C, can add some bloat if you're not careful. Doing object-oriented programming in C is awkward and error-prone, and can add it's own bloat. The nice thing about C++, as opposed to other OO languages like Java, Python, and Smalltalk, is that it takes a pay when you use it approach to bloating -- only if you use a certain OO feature (like virtual functions) do you pay the performance penalty.
Fair enough. But some of the posters we're complaing about Gnome people (Miguel) NOT giving credit where credit is due (like the infamous "legacy" comment, or Miguel's "Gnome is technically superior" to KDE FUD-comment -- how does more bugs and less features == technically superior?)
Also, I don't think you can call Gnome the underdog. Technically, yes, it has a long way to go before it matches KDE. But it has corporate funding from RedHat (which is partially funded by Intel, IBM, etc.). As such, Gnome does get a lot more mainstream press coverage. Until a few months ago, Gnome also got a lot more Slashdot coverage. So in that sense (resources, press coverage), Gnome has the advantage and KDE is the underdog.
Steve Molitor