but the [Gimp's] interface is not only (IMO) counter-intuitive, it goes against the established norm, in a way that could be very easily fixed.
The GIMP UI isn't the best in the world, but I certainly prefer it to Photoshop. Why do all applications have to look the same? Just because many people eat at Macdonalds, does everybody have to?
X11 has had automatic double buffering support for years. The X Render extension provides compositing, alpha support, full transparency (including for windows), antialiased drawing, antialiased text, affine transforms, and splines. X Render is part of XFree86 4.2.
Time will tell whether the Quartz or X11 approach is better. I suspect X Render will be more efficient and work with many more toolkits.
Why is having a different (and arguably better) graphics system 'part of the problem'?
The "part of the problem" that I see isn't the differences in the graphics models, it's the fact that Apple doesn't define a fixed protocol for communication between the display server and the applications--it is just whatever it is, and it can change at any point. That's fine for Apple's business, but it places it completely outside open standards or the open source world, and will probably limit its adoption.
but it provides capabilities and features that would be hard, if not impossible to do with X11 (true transparency, pdf based vector graphics, built-in color management all come to mind).
The X11 render extension offers true transparency, antialiased text and graphics, and an imaging model equivalent to PDF. While it's still evolving, most important capabilities are in XFree86 4.2 already. Color management has been in X11 for years, although it's not used much (and arguably not very useful despite its popularity).
Don't get me wrong, I love some of X11's capabilities, but I just don't see having Quartz rather than X11 being a bad thing.
PDF-like imaging models are nice, and display hardware is now up to it. But you don't need Postscript or PDF in order to get that kind of graphics. In less than a year, I predict most X11 toolkits will use a PDF-like imaging model but with much less overhead than Quartz or Cocoa. We'll just have to wait and see. If I'm right, Apple might want to consider retargetting their back-end.
Made the decision to keep their window manager closed, in order to keep the community from benefiting?
I think it's their full right to do this--they aren't obligated to give software away.
Of course, I also think it's foolish for Apple to invest much effort in this proprietary effort. Aqua looks nice and has set a new standard for appearance, but I predict that we will have smaller and faster X11-based systems with similar appearance and graphical capabilities within less than a year, and you'll be able to get that appearance using any of your favorite toolkits, without porting or any other hassles.
If you want the programming model of Apple's GUI, you already can get GNUStep. I think if Apple made their entire system open source, it would be about as successful--most developers apparently really don't want to develop for Objective-C and DisplayPostscript/DisplayPDF.
APSL has passed most open source advocates' definition of Open
It is arguably an open source license. Nevertheless, there are practical problems with it; RMS's commentary is based on things that anybody using APSL should take a close look at.
Anyone who's rational recognizes the great service Apple has done and continues to do for the open source community.
Oh, get a clue. Apple is a company, not a charity or a benevolent benefactor. They have realized that some open source efforts are good for their business. You don't owe them anything and they don't owe you or me anything. Apple is being rational, and so should anybody using their products.
It's free software and people try different things with it. Whether it catches on or not ultimately is decided by users.
I think it makes a lot of sense for RedHat to cherry-pick the best components from KDE and Gnome and integrate them at the desktop and packaging level. Such integration is long overdue. KDE seems a little better at presenting a Windows-like desktop, but Gnome has some key applications for which KDE does not have a mature equivalent. And third party additions like OpenOffice are yet different and also need to be accomodated
KDE developers can kick and scream all they want, this is good for users, and RedHat is doing the right thing.
Townsend of technology consultants EMA Inc. puts it, "the megahertz myth is a difficult one to overcome."
Perhaps that's because it may not be a myth. The fact that Apple has never submitted SPEC benchmarks is also disturbing. I think Apple is significantly behind in terms of performance and they know it. The question is whether they are going to fix it by moving to IBM or Pentium.
Oh, that will be so much fun. Ballmer will rally the Microsoft "community" by taking is Monkeyboy act on the road (you know, the one where he hops around the stage screaming to get Microsoft employees into the mood--quite funny if you haven't seen it). That will surely inspire the 98% of Microsoft users, you know, secretaries and home users,... well, I don't know. But it will surely inspire them.
(Of course, Ballmer has been "using" the MS community for a long time, he just zeroed in on their wallets.)
For film, you must go to larger formats in order to make pleasing large enlargements--otherwise, you get grain. For digital, that's not needed. You can make huge enlargements that look very nice from up close--they just won't have all the detail.
Gates & Balmer are billionairs because of the M$ stock price.
Well, the fact that they already have more money than half of humanity makes their desire to get even more even more pathological.
Re:as long as I have the option of not installing
on
More on KDE Groupware
·
· Score: 2
That's the beauty of KDE and GNome. They are totally configuarble.
No. Even Motif was much more configurable than KDE or Gnome. KDE and Gnome are basically Windows UI clones with a few more options in their "Preferences" panels. That kind of thing may have a place in this world, but don't delude yourself that it's innovative or configurable.
Re:as long as I have the option of not installing
on
More on KDE Groupware
·
· Score: 2
It might be the end of egotistical elitest Linux hippie types, as any and everyone would be able to use it rather than just the 1337 few
Yes, the same "egotistical elitist Linux hippie types" that created UNIX and Linux in the first place. I am one of them, and I am not ashamed of it.
I took refuge a little less than a decade ago in Linux from lousy Windows user interfaces and overpriced Sun workstations, and if that kind of junk follows me to the Linux platform, I have to move again.
if Linux were to become the same as Windows, it would be a crushing blow to Microsoft, not Linux. BillG can't sell his product against the same thing for free.
And why the hell would I care? If Linux becomes just like Windows, why would I care whether it comes from Bill Gates or whether it's free? What makes Linux interesting to me is that it is still different, in particular in its UI.
But, the KDE and Gnome interfaces to Linux are becoming more Windows like everyday.
As I said, as long as they don't become a core part of the Linux infrastructure, it's fine with me--it's a free country. The moment I have to run them in order to, say, configure the kernel (the new Qt-based kernel configurator is a bad step in that direction), I'm off the platform, as are many other "egotistical hippie types". Fortunately, there are still plenty of obscure, non-mainstream operating systems to go around.
as long as I have the option of not installing it
on
More on KDE Groupware
·
· Score: 1, Flamebait
KDE and Gnome seem very busy cloning every aspect of Microsoft software, good or bad. I guess that will give lots of people a familiar destkop "for free", and a good consequence of that is that protocols and file formats become more and more open.
Let's just not clone the aspects of Microsoft that are most annoying: the idea that there is one solution that works for everybody, and tying together clients and servers through proprietary protocols. If ever system configuration, mail, calendaring, or other imporant Linux functions started relying on Gnome or KDE, it would be the beginning of the end for Linux--because it would then really start being just like Windows.
For most people, selling the Ferrari and buying the Ford is an upgrade. Just try going shopping or taking several kids to grandma's place in a Ferrari. And try driving that Ferrari as fast as it can go on congested highways.
You see, theoretical top speed is not a primary concern for most car buyers--or programming language users. Safety, capacity, and comfort are much more important.
"We can't beat them on price, but we have to add value."
Microsoft has 90% of the desktop market, but enough just isn't enough for them. Their hunger to assimilate every last person on the planet is insatiable. If your tastes or working styles disagree from theirs, there is just no room for you. Microsoft's hunger for market domination is pathological. I suspect that they really do know deep down that their software is just an incoherent collection of marketing-driven features inplemented in a haphazard manner, and it scares them to think that the public at large realize that; that's why everybody with a brain needs to be assimilated before they can create resistance.
What Microsoft just doesn't get is that different people have different preferences. People use Linux not because it's cheaper in some absolute value metric, but because they like it. To Microsoft, "value" means more features, more buttons, and more conformity in terms of appearance. To many Linux users, "value" means fewer features, fewer buttons, more configurability, and standards compliance at the API level. Microsoft can't add that value to Windows; to achieve it, they'd have to subtract stuff from Windows, a lot of stuff, and they can't do it.
Sorry, Ballmer, but unless Microsoft gets the government to mandate Windows, you'll have to be satisfied with 90% market shares. And they may even go down as Linux (for better or for worse) steadily and unstoppably adds your kind of value--as an option for those who want it.
The have two choices for moving beyond C++: Java and C#. I don't see one as being a lot more attractive for HP than the other. And Java has some baggage: their past spat with Sun, and much more competition for consulting dollars.
(Of course, I know what I would want them to do: invest the $50m in gcj and Mono...)
Where did I say that MF was "no better"? I said most people use it for better tonality, not better resolution.
Film flexibility and viewing screns are, of course, also advantages of digital over 35mm. A high-end digital camera is more like a decent MF camera than 35mm.
(Any wedding photographer that uses MF for better resolution is going to have the bride hit him over the head with the bouquet. She wants her beautiful skin tone to show, not her crow's feet. Just you try it.)
will get from a 35 mm frame a little bit over 10 000 000 pixels (no interpolation, no tricks). This are "real", "color" pixels
Quite to the contrary. Film needs to create tonality from discrete grains of silver (converted into blobs of dye in color). That gives you anything but a "real" pixel. Film is more like a half-toned image: you only get one bit per pixel, but you can cleverly arrange those to give the appearance of smooth shades. But you can't have high resolution and smooth shades at the same time with film.
Good digital cameras have 12 bits per pixel and channel at full resolution. That means that you get nice gradations and tonality even at their highest resolutions. To get anything close to that with film, you have to sacrifice a lot of the resolution film has. That's why a 5 Mpixel camera is probably already significantly better for most kinds of pictures than a 35mm film.
You are right that the resolution numbers of digital cameras are cheating a bit. However, it's not as bad as you may think. You really don't need high resolution for color--the eye can't see it anyway, and the compression algorithms and printers will just throw it out anyway even if you capture it.
Yes, those are the usual claims, but they really are misleading. With resolution targets, if you use the right lenses and film, you may get something that by some naive calculation corresponds to 8 Megapixels.
However, the imaging characteristics of film and digital are just different; resolution isn't everything and you can't compare them that way. For practical purposes, a 4 or 5 Megapixel gives most amateurs and professionals similar functionality to what they get with 35mm. Under many conditions, a 5 Mpixel digital camera produces nicer images than a 35mm film camera with any film, and under some other conditions, it's worse.
In fact, scaling up such estimates to medium format is particularly silly. People generally don't use MF or LF for higher resolution, but for characteristics like tonality and DOF. Even if you manage to get the right film and the right lens on your MF system, the longer shutter speeds and smaller DOF will likely counteract any theoretical gain in resolution under most conditions; and films for MF on the whole aren't as good for 35mm either. A Rolleiflex (maybe even loaded with Tri-X) or Pentax67 can't compete with modern 35mm SLR systems and lenses, but that's not the point.
So, don't wait for some big, high-resolution digital camera to do digital imaging. 12 Megapixels is nice, and some people may need, but most amateurs and professionals probably don't. The real news about the two latest SLRs is that they have no focal length multiplier, and that means that we finally get real wide-angles. That's been the real limitation of digital cameras so far.
The big news about these new cameras is that they finally don't have a focal length factor--with older digital SLRs, your nice, expensive 20mm lens turned into 28mm or 32mm. To me, that matters much more than the extra resolution--there are few or no decent wide angles on current high-end digital cameras.
Do 802.11a or 802.11g fix the lousy security of 802.11b? I mean, no wireless network will ever be secure as a wired one because making it secure requires key management, but at least I should be able to expect that if I do my key management correctly, other people can't break in.
IBM has hundreds of employees that might speak out on this. I'm sure many of the folks working in the AIX division hate Linux as much as they ever did and wouldn't be shy to tell you all the ways in which AIX is better, if anybody still bothered to talk to them, that is.
IBM does have a point, insofar that many enterprises have feature checklists that DB2 satisfies and MySQL doesn't, regardless of whether those enterprises actually need those features. And both DB2 and Oracle do have some features that a few customers actually need but MySQL doesn't provide.
Slashdot Mirror
Funny, that's what I recommend you do. History is never just black and white.
The GIMP UI isn't the best in the world, but I certainly prefer it to Photoshop. Why do all applications have to look the same? Just because many people eat at Macdonalds, does everybody have to?
Time will tell whether the Quartz or X11 approach is better. I suspect X Render will be more efficient and work with many more toolkits.
The "part of the problem" that I see isn't the differences in the graphics models, it's the fact that Apple doesn't define a fixed protocol for communication between the display server and the applications--it is just whatever it is, and it can change at any point. That's fine for Apple's business, but it places it completely outside open standards or the open source world, and will probably limit its adoption.
but it provides capabilities and features that would be hard, if not impossible to do with X11 (true transparency, pdf based vector graphics, built-in color management all come to mind).
The X11 render extension offers true transparency, antialiased text and graphics, and an imaging model equivalent to PDF. While it's still evolving, most important capabilities are in XFree86 4.2 already. Color management has been in X11 for years, although it's not used much (and arguably not very useful despite its popularity).
Don't get me wrong, I love some of X11's capabilities, but I just don't see having Quartz rather than X11 being a bad thing.
PDF-like imaging models are nice, and display hardware is now up to it. But you don't need Postscript or PDF in order to get that kind of graphics. In less than a year, I predict most X11 toolkits will use a PDF-like imaging model but with much less overhead than Quartz or Cocoa. We'll just have to wait and see. If I'm right, Apple might want to consider retargetting their back-end.
It isn't, and that's part of the problem. if it were a protocol, it would be documented and could be implemented on other platforms.
X11, in contrast, is a well-documented protocol, which is why there are dozens of different server and client library implementations for it.
I think it's their full right to do this--they aren't obligated to give software away.
Of course, I also think it's foolish for Apple to invest much effort in this proprietary effort. Aqua looks nice and has set a new standard for appearance, but I predict that we will have smaller and faster X11-based systems with similar appearance and graphical capabilities within less than a year, and you'll be able to get that appearance using any of your favorite toolkits, without porting or any other hassles.
If you want the programming model of Apple's GUI, you already can get GNUStep. I think if Apple made their entire system open source, it would be about as successful--most developers apparently really don't want to develop for Objective-C and DisplayPostscript/DisplayPDF.
It is arguably an open source license. Nevertheless, there are practical problems with it; RMS's commentary is based on things that anybody using APSL should take a close look at.
Anyone who's rational recognizes the great service Apple has done and continues to do for the open source community.
Oh, get a clue. Apple is a company, not a charity or a benevolent benefactor. They have realized that some open source efforts are good for their business. You don't owe them anything and they don't owe you or me anything. Apple is being rational, and so should anybody using their products.
socialist radicals
You say that as if it's a bad thing.
I think it makes a lot of sense for RedHat to cherry-pick the best components from KDE and Gnome and integrate them at the desktop and packaging level. Such integration is long overdue. KDE seems a little better at presenting a Windows-like desktop, but Gnome has some key applications for which KDE does not have a mature equivalent. And third party additions like OpenOffice are yet different and also need to be accomodated
KDE developers can kick and scream all they want, this is good for users, and RedHat is doing the right thing.
Perhaps that's because it may not be a myth. The fact that Apple has never submitted SPEC benchmarks is also disturbing. I think Apple is significantly behind in terms of performance and they know it. The question is whether they are going to fix it by moving to IBM or Pentium.
(Of course, Ballmer has been "using" the MS community for a long time, he just zeroed in on their wallets.)
For film, you must go to larger formats in order to make pleasing large enlargements--otherwise, you get grain. For digital, that's not needed. You can make huge enlargements that look very nice from up close--they just won't have all the detail.
Well, the fact that they already have more money than half of humanity makes their desire to get even more even more pathological.
No. Even Motif was much more configurable than KDE or Gnome. KDE and Gnome are basically Windows UI clones with a few more options in their "Preferences" panels. That kind of thing may have a place in this world, but don't delude yourself that it's innovative or configurable.
Yes, the same "egotistical elitist Linux hippie types" that created UNIX and Linux in the first place. I am one of them, and I am not ashamed of it.
I took refuge a little less than a decade ago in Linux from lousy Windows user interfaces and overpriced Sun workstations, and if that kind of junk follows me to the Linux platform, I have to move again.
if Linux were to become the same as Windows, it would be a crushing blow to Microsoft, not Linux. BillG can't sell his product against the same thing for free.
And why the hell would I care? If Linux becomes just like Windows, why would I care whether it comes from Bill Gates or whether it's free? What makes Linux interesting to me is that it is still different, in particular in its UI.
But, the KDE and Gnome interfaces to Linux are becoming more Windows like everyday.
As I said, as long as they don't become a core part of the Linux infrastructure, it's fine with me--it's a free country. The moment I have to run them in order to, say, configure the kernel (the new Qt-based kernel configurator is a bad step in that direction), I'm off the platform, as are many other "egotistical hippie types". Fortunately, there are still plenty of obscure, non-mainstream operating systems to go around.
Let's just not clone the aspects of Microsoft that are most annoying: the idea that there is one solution that works for everybody, and tying together clients and servers through proprietary protocols. If ever system configuration, mail, calendaring, or other imporant Linux functions started relying on Gnome or KDE, it would be the beginning of the end for Linux--because it would then really start being just like Windows.
You see, theoretical top speed is not a primary concern for most car buyers--or programming language users. Safety, capacity, and comfort are much more important.
Microsoft has 90% of the desktop market, but enough just isn't enough for them. Their hunger to assimilate every last person on the planet is insatiable. If your tastes or working styles disagree from theirs, there is just no room for you. Microsoft's hunger for market domination is pathological. I suspect that they really do know deep down that their software is just an incoherent collection of marketing-driven features inplemented in a haphazard manner, and it scares them to think that the public at large realize that; that's why everybody with a brain needs to be assimilated before they can create resistance.
What Microsoft just doesn't get is that different people have different preferences. People use Linux not because it's cheaper in some absolute value metric, but because they like it. To Microsoft, "value" means more features, more buttons, and more conformity in terms of appearance. To many Linux users, "value" means fewer features, fewer buttons, more configurability, and standards compliance at the API level. Microsoft can't add that value to Windows; to achieve it, they'd have to subtract stuff from Windows, a lot of stuff, and they can't do it.
Sorry, Ballmer, but unless Microsoft gets the government to mandate Windows, you'll have to be satisfied with 90% market shares. And they may even go down as Linux (for better or for worse) steadily and unstoppably adds your kind of value--as an option for those who want it.
(Of course, I know what I would want them to do: invest the $50m in gcj and Mono...)
Film flexibility and viewing screns are, of course, also advantages of digital over 35mm. A high-end digital camera is more like a decent MF camera than 35mm.
(Any wedding photographer that uses MF for better resolution is going to have the bride hit him over the head with the bouquet. She wants her beautiful skin tone to show, not her crow's feet. Just you try it.)
Quite to the contrary. Film needs to create tonality from discrete grains of silver (converted into blobs of dye in color). That gives you anything but a "real" pixel. Film is more like a half-toned image: you only get one bit per pixel, but you can cleverly arrange those to give the appearance of smooth shades. But you can't have high resolution and smooth shades at the same time with film.
Good digital cameras have 12 bits per pixel and channel at full resolution. That means that you get nice gradations and tonality even at their highest resolutions. To get anything close to that with film, you have to sacrifice a lot of the resolution film has. That's why a 5 Mpixel camera is probably already significantly better for most kinds of pictures than a 35mm film.
You are right that the resolution numbers of digital cameras are cheating a bit. However, it's not as bad as you may think. You really don't need high resolution for color--the eye can't see it anyway, and the compression algorithms and printers will just throw it out anyway even if you capture it.
However, the imaging characteristics of film and digital are just different; resolution isn't everything and you can't compare them that way. For practical purposes, a 4 or 5 Megapixel gives most amateurs and professionals similar functionality to what they get with 35mm. Under many conditions, a 5 Mpixel digital camera produces nicer images than a 35mm film camera with any film, and under some other conditions, it's worse.
In fact, scaling up such estimates to medium format is particularly silly. People generally don't use MF or LF for higher resolution, but for characteristics like tonality and DOF. Even if you manage to get the right film and the right lens on your MF system, the longer shutter speeds and smaller DOF will likely counteract any theoretical gain in resolution under most conditions; and films for MF on the whole aren't as good for 35mm either. A Rolleiflex (maybe even loaded with Tri-X) or Pentax67 can't compete with modern 35mm SLR systems and lenses, but that's not the point.
So, don't wait for some big, high-resolution digital camera to do digital imaging. 12 Megapixels is nice, and some people may need, but most amateurs and professionals probably don't. The real news about the two latest SLRs is that they have no focal length multiplier, and that means that we finally get real wide-angles. That's been the real limitation of digital cameras so far.
Now, if only the price came down...
Do 802.11a or 802.11g fix the lousy security of 802.11b? I mean, no wireless network will ever be secure as a wired one because making it secure requires key management, but at least I should be able to expect that if I do my key management correctly, other people can't break in.
IBM does have a point, insofar that many enterprises have feature checklists that DB2 satisfies and MySQL doesn't, regardless of whether those enterprises actually need those features. And both DB2 and Oracle do have some features that a few customers actually need but MySQL doesn't provide.
I switched from Amazon to other on-line book sellers after their one-click patent; I recommend you do the same.