In my experience, people are fanatic about their way of doing something, even if it is completely illogical and not in their best interest.
In general I've found it's possible to get people to change their minds, if they are reasonably intelligent and don't have a large investment in one particular viewpoint (ie little emotional attachment to it). Even then, it's still possible to defeat people in argument, even if they don't change their minds straight away, sometimes it seems to plant the seeds of doubt in their minds - maybe, just maybe, this guy has a point.
Silly though it may sound, I've had my mind changed by what I've read on Slashdot for instance many times. A year ago, I was of the opinion that X sucked and Linus should make all the GNOME developers work on KDE:)
Why don't we let the good Professor be the public face of the FSF and banish the zealots to a sound-proof box?
Agreed. I thought the answers were tremendously cogent and well thought out. I didn't get the feeling any of the questions were being ducked, and I did get the feeling that he truly believes in what he is saying. I was able to understand a lot of what was said, despite the largely esotoric nature of the subject matter.
Basically, thanks Professor, and congratulations on an excellent interview.
Lack of user knowledge..... is a common problem in regards to bugs
This applies to everybody by the way, not just "lusers". I've had egg on my face several times when Gecko was doing seriously screwy rendering, the page worked fine in IE and Opera, I couldn't see the reason for it so I reported it as a bug. Normally, it was an obscure fault in my CSS, caused by lack of a good working knowledge of the spec. There was no bug.
I'm a lot more careful when reporting CSS bugs in Gecko these days.
Why not translate all that work to the desktop and start now on the plan of phasing out the X windowing system from unix GUIs. I'm not saying we take drastic steps now, but we'd be stupid to take no steps to transition the desktop to QT all the way down.
You're ignoring the fact that Qt is only GPLd when using X11. You're ignoring a lot of facts actually.
X windows reminds me of the space shuttle. It's big and old and we know it won't last forever, but we hide our heads in the sand and we don't want to hear about it. Well, that's a really stupid attitude, especially since there is such an inviting alternative.
If you're going to compare X to the space shuttle. then the Linux framebuffer would be a light Cessna aircraft. X has features, it has hardware support, it has apps. DirectFB (presumably what you are talking about) does not.
Finally, remember that it's actually GTK that works on the Linux framebuffer - not Qt. See for yourself. Qt is only GPLd when running on X, the code to make it work on the framebuffer isn't under the GPL afaik.
The trolls have given back a lot to the Linux community. KDE would be nowhere without Qt.
Don't forget the huge publicity boost KDE has given Qt. How many people here on Slashdot would have heard of Qt if it weren't for KDE? It works both ways.
GDI crashing, be it in the kernel or user mode, is basically a fatal system error.
Right. Ditto for X freezes. It always makes me laugh when people say "Oh, if X freezes just ssh in and kill it". Uh, right. X is dead man, you just lost all your apps. Getting back to the login screen by killing X is the same as rebooting effectively on a desktop machine, except faster.
That's not a rationale for moving X into the kernel of course. Having it as a separate userland process can be very useful - servers don't need a GUI. It can be upgraded much easier. If it can't start for some reason, you can simply failover to a command prompt. And so on.
I wonder if this is part of an attack against Apple?
I seriously doubt it. Apple is no threat to Microsoft, not even a remote one. They still make Mac software, including it's only real office suite. How can they be attacking Apple with that state of affairs?
Microsoft are shielded from Apple by simple economics, which haven't changed in a decade. Windows apps won't run on anything other than Windows - if you happen to run your copy of Windows on a Mac via Virtual PC, what do Microsoft care? You're still using their software, you're still paying them for the priviledge.
The only thing that is a real threat to Microsoft these days are Wine with Linux. Wine is obviously not really possible without the supporting base of free software it uses, hence the fact that it's a threat.
Wine on Linux/FreeBSD is basically the only way you can run Win32 programs without paying for Windows itself in a completely legal fashion. If Wine were to run on a Mac, maybe then it'd be an issue, but that's a large task technically. It wouldn't surprise me if Apple tried that at some point in the next few years if Virtual PC dies. Thank god Wine is under the LGPL.
There are quite a few in Wine also. I think they're primarily useful as an equivalent to try {} finally {} in languages that support exceptions (obviously C doesn't). If a condition fails just goto the end, where the cleanup code is guaranteed to run.
Used properly, gotos are no more harmful than any other construct.
A better comparison would be to a full gentoo compile, which apparently can take up to 24 hours on fairly normal hardware. Those build labs won't be running anything like normal hardware, so I'm guessing the size of Windows and Linux is roughly similar.
There's an interesting presentation on how they manage this here. Basically they have 5000 team members, not necessarily actual computer programmers....
The wierd thing about that is that a while ago there was a leaked presentation claiming that the development of W2K had 5000 developers and it was nearly unmanagable, and that in future they weren't going to have as many. So, either they improved their communications techniques considerably, or the numbers are kind of not quite accurate. Probably just internal politics.
I really like the NT Kernel. What driver developers do with it isn't the kernel's fault.
Well....... maybe.
I seem to recall an MS employee claiming that it was entirely Microsofts fault Windows was so unstable, even though crashes were normally caused by faulty drivers. His theory was that if MS were more open with the kernel code, driver manufacturers could work more closely and easier with them, and the overall stability would go up. Instead what happened (they claimed) was that they would investigate a crash, find that some dodgy driver was screwing about with the kernel and so they'd tighten up the interfaces, get even more secretive with the code. The driver developers, faced with a brick wall, would then invent even more elaborate (and fragile) hacks to do what they want, so the stability went down, not up.
So, you can't really blame the kernel as a thing per se, but perhaps you can blame the management of it. Linux is now facing a similar problem with the growth of binary only drivers - they tend to hook into the ksyms and cause extremely hard to track down bugs, which is why they are no longer allowed to use those hooks.
While it's possible there will be a few new readers, it's more likely the same people who already like/hate him who are reading this.
A few new readers? I think pretty much anybody outside of America won't have heard of him. I know I hadn't until this interview showed up. Dave Barry? Who?
Like hell the RIAA will let mp3s (or ogg) exist anymore, and if they do, I'll bet the default setting for any mp3 you record will be "don't copy this".
Yawn. More mindless paranoia. The RIAA have been trying to kill MP3 for ages. They've failed. You can still rip CDs easily, you can still trade MP3s with total strangers.
Here's what gets me though, why is MS the bad guy here? Obviously there's some demand for MS to fill here.
Yeah. Actually I've been told (by an MS exec) that the demand is mostly coming from normal business. They like the idea of keeping control of internal documents, keeping it secure, all the benefits of DRM etc. I've seen a roundtable discussion at a conference that was discussing the benefits a new age of DRM will bring, these guys were really enthusiastic but they weren't from the MPAA or RIAA. They were just business people (except the blonde in the short skirt, I think she was just there to distract the attendees).
Look, what you really want is a PowerBook. You know it, everyone here knows it. You just won't admit it.
What the hell is this, Freud? He has presumably evaluated the options, and decided he wanted a PC laptop without paying for Windows. Those were the requirements. Why are people having such a hard time dealing with this? You can tell there aren't many salespeople on this site, second guessing the customer is a quick way to lose sales.
So let's compare features. The Apple certainly has no potential whatsoever of running Microsoft Windows except through some complete emulation/virtualization software. Score one for Apple.
WTF? A lack of an option is a feature? Only a true Apple shill would claim that.
The Apple comes with an actual GUI far superior to Microsoft Windows (not even a contest) and much more polished than your typical UNIX GUI.
Want more? Well, your Mac is actually capable of running Microsoft Office should you later find yourself in a bind and be REQUIRED to deal with it to put food on the table.
Except that it's not Office. If you already own a copy (quite likely), you might as well just use it, and get perfect compatability. On the other hand, he already said quite clearly he didn't need any Windows software, which presumably includes Office. Piracy can be done on any OS btw, at least on Linux you don't actually need to.
That means that you can create "lickable" GUIs using the X Window System.
This has got to be a troll. It lets you do what you have been able to do on every other OS for years, namely display X apps. It doesn't alter their appearance.
And on top of all of that...... *snip*
Why should he care about any of that? Oh, and that's a nice hidden font preview you've got there.
And you're not supporting a company that supports Microsoft. In fact, you are supporting a company that actually competes with Microsoft (on some small level).
This is definately a troll. Apple "competing" with Microsoft is an interesting parlour game, but no more. You were advertising the virtues of IE and Office earlier.
What is it around here? Do people get commission on every Mac sold or something?
Another benefit of the mac is that the linux distro's will work better and be less buggy then intel ones because of the limited hardware. They don't need to support 3,000 peripherals from god knows where.
Sorry, but that's not true in my experience. I've found PPC distros to be remarkably inferior to the PC equivalents, or at least for the fairly easy to use ones. Some of the big names don't release PPC versions at all. Common sense would tell you that Mac hardware would be easier to support, but apparently that isn't the case, and a minority market within a minority market means they get little testing and feedback compared with the PC equivalents. Anyway, that's my experience when a friend wanted Linux on his Mac, we went through about 5 distros, all of them sucked hard compared to the same on a PC.
Or if you hate ms and refuse to support them go buy a powerbook from apple. They are pricy but have been known for over a decade to be supperior quality.
He already said he didn't want to pay extra for an iBook, why would he want to pay even more for a PowerBook? He also noted performance was a concern, something that writes off Apple entirely.
You can run Linux on a mac as well as have a big selection of software to choose from with MacOSX. Adobe photoshop, IE, MS-Office, games, etc.
Except he already said all his software was available on Linux. Also, this definition of the word "big" must be different to the one I use, other than Photoshop, you can run the real IE and MSOffice on Linux via Wine. Quite why anybody would care about Mac/IE is beyond me, it doesn't have the rendering quirks that mean you have to have IE for Windows present in web design, and if he's happy with Linux why would he pay however many hundred dollars it is for Mac/Office? Games selection is weak on the Mac, if he really cared about that (which he doesn't), he could use WineX on Linux or use the growing collection of native games.
I have not seen one innovation from their research dept make it to market. Please indicate if you have contrary proof.
Interesting really. I remember an interview with some bright young graduate working at MSR in Cambridge. She was researching visual bookmarks, or something like that, draggable thumbnails of web pages. I think there was something like it in IE for the Mac, but this was a bit different (can't remember exactly how). Anyway, she created a prototype and integrated it with IE. She was asked what would be done with this system she'd created (and written) and the response was something along the lines of, "Oh well I'll forward it to the IE product team, but I very much doubt it'll actually end up in the product". She didn't elaborate, but seemed perfectly happy with the idea that despite this clearly being quite a good idea, it wouldn't make it into the product.
IE6 rolled around and what was new? Er, resizing images that were too big to fit on the screen. I wonder how much other stuff ends up like that. Researched, then forgotten about.
Perhaps Apple is still in the early stages of tweaking Mac OS X... maybe they're working on the demanding areas first and will eventually touch up the more minor performance issues (window resize, for example).
When they're doing things like pushing GUI compositing onto a 3D accelerator, I don't think they're in the "early stages" of optimization somehow.
The strange performance you're seeing is pretty interesting, and partly explains why there are so many "Macs are slow! No they're not! Yes they are!" type flamewars around here.
Basically I think that things which push the CPU/mobo hard, show up the fact that the Motorola CPUs are seriously underpowered. There are quite a few things however that can basically disguise that. Winding up the GUI server to 10,000 priority for instance. Having the window compositing done in hardware. I expect apps like Final Cut Pro and PhotoShop are pretty heavily optimized for that processor, stuff like altivec and so on are apparently good for image processing.
I'm surprised you find PhotoShop fast. I'd hazard a guess that this is because applying filters doesn't normally give you a time (ie "Done in 6.34seconds"), and using the tools is actually not that CPU intensive, so it's hard to "see" the slowness.
On the other hand, you'd be surprised at what is intensive. Window resizing is a good example, as with layout based GUI systems like Aqua and GTK, when opaque resizes are used the gui has to be reflowed and rerendered many many times a second for it to feel fast. This was a problem with Mozilla in the early days, I remember them trying to optimize Gecko to do opaque resizes well. It's very noticeable, because being effectively animation humans are sensitive to the "jerkyness". There is no good way to hardware accelerate GUI reflow. App startup would be another one, but they're now shifting the fixup processing (which is pretty heavy duty stuff) to application install time - you know when it says "Optimizing application"? Well, it's doing stuff that'd otherwise be done on application startup. I guess it's a kind of optimization, though moving stuff around doesn't actually make it any faster.
Other things that are probably slow are things that have to be done on the CPU - iCal calculations sounds right, compilation of large programs should show it up too. Large spreadsheet calculations. I don't really know how much of this stuff is accurate. Oh, obviously web page rendering as well, I think they are working hard on KHTML performance for this very reason.
So basically, the things where performance is critical to UI feel (because the user has to wait for it to finish) will feel slow, because, well, they are slow. Quite a lot of other stuff can either be made to feel fast by keeping the GUI responsive, or you can use hardware tricks to accelerate it.
Think about it - Microsoft is a monopoly, basically people have to buy their software. Yes, for geeks and highly non-demanding users stuff Windows lockin is less effective, but for the vast majority of people, they feel they have to use Windows.
So.... what does that leave left to advertise? It must get pretty boring working in Microsofts adverts department. I expect they've got bored of spamming OSDN, that was a good wheeze for a while, but now they have to do something to make the long winter days go past right?
Anyway, it's not like MS are actually threatened by Apple, anybody who runs the numbers can see that. It's just a side show, an entertaining game to try and give the surface appearance that there's actually competition in the markets.
Er, yes? Gecko is the best renderer out there by a LONG way. It's the de facto standard on Linux according to my site stats, and for good reason.
When Apple were looking for a browser core to use for Safari, they chose khtml over gecko, because it's cleaner.
I've said it before, and I'll say it again, I think that was a dumb mistake. OK, so they made that decision over a year ago, when Mozilla 1.0 wasn't yet out, and the code was much worse than it is now.
Nonetheless, there are a few things people should bear in mind about this:
Firstly, the idea that because Gecko was complex, it couldn't be used in a web browser, is a dumb one. Apple have put a lot of effort into bringing KHTML up to scratch, but Gecko was already there. So, if the Galeon, Epiphany, K-Meleon and Pheonix teams can make good browsers based on it, why can't Apple? Apple have way more resources than the Galeon team. Gecko is already one of the most advanced renders out there, they wouldn't even have needed to touch most of the code.
Secondly, KHTML is still way way immature compared to Gecko. It only recently got support for XML (basic support only). It's still catching up in terms of core standards compliance, and forget about stuff like XSLT, MathML, etc. That's not to bash KHTML, what's there is good, but in terms of usefulness in browsing the web, Gecko owns it. On my site, over 50% of the hits come from Gecko based browsers, something like 40% from IE and I think about 3-4% from Konqueror. KHTML and Gecko have been choices on Linux for a long time, yet most seem to use Gecko.
To be honest, I think they chose KHTML because it was hard to make Gecko efficiently use the unusual Mac rendering model. Web browsing was really showing up the fact that Macs are slow these days, in ways that can't be disguised using hardware acceleration, or windowing system tricks etc. KHTML could be more easily hacked to get raw speed, which is clearly more important to them than features or website compatability.
I use Mozilla, and its memory usage when I last looked (yesterday) was 81MB. In contrast Opera was sitting at 10MB, rendering pages faster and supporting CSS better (Moz still doesn't support CSS counters, so I can't number headings automatically, for example.)
But Opera still has sucky DOM support (i think) etc etc. 81mb sounds very large indeed, I've never seen Moz use even half that. Bear in mind accurately measuring memory usage is hard with standard OS tools, as they usually don't adequately distinguish between shared libs etc.
AOL don't want to put themselves in a position where Microsoft can hold them over the barrel - sure, the ActiveX control can be used by them today, but what about tomorrow? Being independant for them makes sense.
Tunnelling doesn't allow us to deploy proper IP multicast however, which I think is one of the biggest advantages.
Having a proper, working multicast system well supported by the backbone routers would rewrite the rules of the net. Radio stations wouldn't be limited by needing constant donations of bandwidth, large file downloads could be put on a carousel saving more bandwidth, even websites if you were clever enough.
Using tunnelling doesn't give ISPs any incentive to move to IPv6 natively though, so we don't get the benefits of that.
Slashdot Mirror
In general I've found it's possible to get people to change their minds, if they are reasonably intelligent and don't have a large investment in one particular viewpoint (ie little emotional attachment to it). Even then, it's still possible to defeat people in argument, even if they don't change their minds straight away, sometimes it seems to plant the seeds of doubt in their minds - maybe, just maybe, this guy has a point.
Silly though it may sound, I've had my mind changed by what I've read on Slashdot for instance many times. A year ago, I was of the opinion that X sucked and Linus should make all the GNOME developers work on KDE :)
Agreed. I thought the answers were tremendously cogent and well thought out. I didn't get the feeling any of the questions were being ducked, and I did get the feeling that he truly believes in what he is saying. I was able to understand a lot of what was said, despite the largely esotoric nature of the subject matter.
Basically, thanks Professor, and congratulations on an excellent interview.
This applies to everybody by the way, not just "lusers". I've had egg on my face several times when Gecko was doing seriously screwy rendering, the page worked fine in IE and Opera, I couldn't see the reason for it so I reported it as a bug. Normally, it was an obscure fault in my CSS, caused by lack of a good working knowledge of the spec. There was no bug.
I'm a lot more careful when reporting CSS bugs in Gecko these days.
You're ignoring the fact that Qt is only GPLd when using X11. You're ignoring a lot of facts actually.
X windows reminds me of the space shuttle. It's big and old and we know it won't last forever, but we hide our heads in the sand and we don't want to hear about it. Well, that's a really stupid attitude, especially since there is such an inviting alternative.
If you're going to compare X to the space shuttle. then the Linux framebuffer would be a light Cessna aircraft. X has features, it has hardware support, it has apps. DirectFB (presumably what you are talking about) does not.
Finally, remember that it's actually GTK that works on the Linux framebuffer - not Qt. See for yourself. Qt is only GPLd when running on X, the code to make it work on the framebuffer isn't under the GPL afaik.
Don't forget the huge publicity boost KDE has given Qt. How many people here on Slashdot would have heard of Qt if it weren't for KDE? It works both ways.
Right. Ditto for X freezes. It always makes me laugh when people say "Oh, if X freezes just ssh in and kill it". Uh, right. X is dead man, you just lost all your apps. Getting back to the login screen by killing X is the same as rebooting effectively on a desktop machine, except faster.
That's not a rationale for moving X into the kernel of course. Having it as a separate userland process can be very useful - servers don't need a GUI. It can be upgraded much easier. If it can't start for some reason, you can simply failover to a command prompt. And so on.
I seriously doubt it. Apple is no threat to Microsoft, not even a remote one. They still make Mac software, including it's only real office suite. How can they be attacking Apple with that state of affairs?
Microsoft are shielded from Apple by simple economics, which haven't changed in a decade. Windows apps won't run on anything other than Windows - if you happen to run your copy of Windows on a Mac via Virtual PC, what do Microsoft care? You're still using their software, you're still paying them for the priviledge.
The only thing that is a real threat to Microsoft these days are Wine with Linux. Wine is obviously not really possible without the supporting base of free software it uses, hence the fact that it's a threat.
Wine on Linux/FreeBSD is basically the only way you can run Win32 programs without paying for Windows itself in a completely legal fashion. If Wine were to run on a Mac, maybe then it'd be an issue, but that's a large task technically. It wouldn't surprise me if Apple tried that at some point in the next few years if Virtual PC dies. Thank god Wine is under the LGPL.
As did Netscape, who actually made their money in a rather nice server business
Used properly, gotos are no more harmful than any other construct.
A better comparison would be to a full gentoo compile, which apparently can take up to 24 hours on fairly normal hardware. Those build labs won't be running anything like normal hardware, so I'm guessing the size of Windows and Linux is roughly similar.
There's an interesting presentation on how they manage this here. Basically they have 5000 team members, not necessarily actual computer programmers....
The wierd thing about that is that a while ago there was a leaked presentation claiming that the development of W2K had 5000 developers and it was nearly unmanagable, and that in future they weren't going to have as many. So, either they improved their communications techniques considerably, or the numbers are kind of not quite accurate. Probably just internal politics.
Well....... maybe.
I seem to recall an MS employee claiming that it was entirely Microsofts fault Windows was so unstable, even though crashes were normally caused by faulty drivers. His theory was that if MS were more open with the kernel code, driver manufacturers could work more closely and easier with them, and the overall stability would go up. Instead what happened (they claimed) was that they would investigate a crash, find that some dodgy driver was screwing about with the kernel and so they'd tighten up the interfaces, get even more secretive with the code. The driver developers, faced with a brick wall, would then invent even more elaborate (and fragile) hacks to do what they want, so the stability went down, not up.
So, you can't really blame the kernel as a thing per se, but perhaps you can blame the management of it. Linux is now facing a similar problem with the growth of binary only drivers - they tend to hook into the ksyms and cause extremely hard to track down bugs, which is why they are no longer allowed to use those hooks.
A few new readers? I think pretty much anybody outside of America won't have heard of him. I know I hadn't until this interview showed up. Dave Barry? Who?
Yawn. More mindless paranoia. The RIAA have been trying to kill MP3 for ages. They've failed. You can still rip CDs easily, you can still trade MP3s with total strangers.
Yeah. Actually I've been told (by an MS exec) that the demand is mostly coming from normal business. They like the idea of keeping control of internal documents, keeping it secure, all the benefits of DRM etc. I've seen a roundtable discussion at a conference that was discussing the benefits a new age of DRM will bring, these guys were really enthusiastic but they weren't from the MPAA or RIAA. They were just business people (except the blonde in the short skirt, I think she was just there to distract the attendees).
Yes it has. Lots of times.
Look, what you really want is a PowerBook. You know it, everyone here knows it. You just won't admit it.
What the hell is this, Freud? He has presumably evaluated the options, and decided he wanted a PC laptop without paying for Windows. Those were the requirements. Why are people having such a hard time dealing with this? You can tell there aren't many salespeople on this site, second guessing the customer is a quick way to lose sales.
So let's compare features. The Apple certainly has no potential whatsoever of running Microsoft Windows except through some complete emulation/virtualization software. Score one for Apple.
WTF? A lack of an option is a feature? Only a true Apple shill would claim that.
The Apple comes with an actual GUI far superior to Microsoft Windows (not even a contest) and much more polished than your typical UNIX GUI.
Evidence please. Quite a few Mac users aren't so impressed.
Want more? Well, your Mac is actually capable of running Microsoft Office should you later find yourself in a bind and be REQUIRED to deal with it to put food on the table.
Except that it's not Office. If you already own a copy (quite likely), you might as well just use it, and get perfect compatability. On the other hand, he already said quite clearly he didn't need any Windows software, which presumably includes Office. Piracy can be done on any OS btw, at least on Linux you don't actually need to.
That means that you can create "lickable" GUIs using the X Window System.
This has got to be a troll. It lets you do what you have been able to do on every other OS for years, namely display X apps. It doesn't alter their appearance.
And on top of all of that...... *snip*
Why should he care about any of that? Oh, and that's a nice hidden font preview you've got there.
And you're not supporting a company that supports Microsoft. In fact, you are supporting a company that actually competes with Microsoft (on some small level).
This is definately a troll. Apple "competing" with Microsoft is an interesting parlour game, but no more. You were advertising the virtues of IE and Office earlier.
What is it around here? Do people get commission on every Mac sold or something?
Sorry, but that's not true in my experience. I've found PPC distros to be remarkably inferior to the PC equivalents, or at least for the fairly easy to use ones. Some of the big names don't release PPC versions at all. Common sense would tell you that Mac hardware would be easier to support, but apparently that isn't the case, and a minority market within a minority market means they get little testing and feedback compared with the PC equivalents. Anyway, that's my experience when a friend wanted Linux on his Mac, we went through about 5 distros, all of them sucked hard compared to the same on a PC.
Or if you hate ms and refuse to support them go buy a powerbook from apple. They are pricy but have been known for over a decade to be supperior quality.
He already said he didn't want to pay extra for an iBook, why would he want to pay even more for a PowerBook? He also noted performance was a concern, something that writes off Apple entirely.
You can run Linux on a mac as well as have a big selection of software to choose from with MacOSX. Adobe photoshop, IE, MS-Office, games, etc.
Except he already said all his software was available on Linux. Also, this definition of the word "big" must be different to the one I use, other than Photoshop, you can run the real IE and MSOffice on Linux via Wine. Quite why anybody would care about Mac/IE is beyond me, it doesn't have the rendering quirks that mean you have to have IE for Windows present in web design, and if he's happy with Linux why would he pay however many hundred dollars it is for Mac/Office? Games selection is weak on the Mac, if he really cared about that (which he doesn't), he could use WineX on Linux or use the growing collection of native games.
Interesting really. I remember an interview with some bright young graduate working at MSR in Cambridge. She was researching visual bookmarks, or something like that, draggable thumbnails of web pages. I think there was something like it in IE for the Mac, but this was a bit different (can't remember exactly how). Anyway, she created a prototype and integrated it with IE. She was asked what would be done with this system she'd created (and written) and the response was something along the lines of, "Oh well I'll forward it to the IE product team, but I very much doubt it'll actually end up in the product". She didn't elaborate, but seemed perfectly happy with the idea that despite this clearly being quite a good idea, it wouldn't make it into the product.
IE6 rolled around and what was new? Er, resizing images that were too big to fit on the screen. I wonder how much other stuff ends up like that. Researched, then forgotten about.
When they're doing things like pushing GUI compositing onto a 3D accelerator, I don't think they're in the "early stages" of optimization somehow.
The strange performance you're seeing is pretty interesting, and partly explains why there are so many "Macs are slow! No they're not! Yes they are!" type flamewars around here.
Basically I think that things which push the CPU/mobo hard, show up the fact that the Motorola CPUs are seriously underpowered. There are quite a few things however that can basically disguise that. Winding up the GUI server to 10,000 priority for instance. Having the window compositing done in hardware. I expect apps like Final Cut Pro and PhotoShop are pretty heavily optimized for that processor, stuff like altivec and so on are apparently good for image processing.
I'm surprised you find PhotoShop fast. I'd hazard a guess that this is because applying filters doesn't normally give you a time (ie "Done in 6.34seconds"), and using the tools is actually not that CPU intensive, so it's hard to "see" the slowness.
On the other hand, you'd be surprised at what is intensive. Window resizing is a good example, as with layout based GUI systems like Aqua and GTK, when opaque resizes are used the gui has to be reflowed and rerendered many many times a second for it to feel fast. This was a problem with Mozilla in the early days, I remember them trying to optimize Gecko to do opaque resizes well. It's very noticeable, because being effectively animation humans are sensitive to the "jerkyness". There is no good way to hardware accelerate GUI reflow. App startup would be another one, but they're now shifting the fixup processing (which is pretty heavy duty stuff) to application install time - you know when it says "Optimizing application"? Well, it's doing stuff that'd otherwise be done on application startup. I guess it's a kind of optimization, though moving stuff around doesn't actually make it any faster.
Other things that are probably slow are things that have to be done on the CPU - iCal calculations sounds right, compilation of large programs should show it up too. Large spreadsheet calculations. I don't really know how much of this stuff is accurate. Oh, obviously web page rendering as well, I think they are working hard on KHTML performance for this very reason.
So basically, the things where performance is critical to UI feel (because the user has to wait for it to finish) will feel slow, because, well, they are slow. Quite a lot of other stuff can either be made to feel fast by keeping the GUI responsive, or you can use hardware tricks to accelerate it.
So.... what does that leave left to advertise? It must get pretty boring working in Microsofts adverts department. I expect they've got bored of spamming OSDN, that was a good wheeze for a while, but now they have to do something to make the long winter days go past right?
Anyway, it's not like MS are actually threatened by Apple, anybody who runs the numbers can see that. It's just a side show, an entertaining game to try and give the surface appearance that there's actually competition in the markets.
Is Gecko actually a good thing?
Er, yes? Gecko is the best renderer out there by a LONG way. It's the de facto standard on Linux according to my site stats, and for good reason.
When Apple were looking for a browser core to use for Safari, they chose khtml over gecko, because it's cleaner.
I've said it before, and I'll say it again, I think that was a dumb mistake. OK, so they made that decision over a year ago, when Mozilla 1.0 wasn't yet out, and the code was much worse than it is now.
Nonetheless, there are a few things people should bear in mind about this:
Firstly, the idea that because Gecko was complex, it couldn't be used in a web browser, is a dumb one. Apple have put a lot of effort into bringing KHTML up to scratch, but Gecko was already there. So, if the Galeon, Epiphany, K-Meleon and Pheonix teams can make good browsers based on it, why can't Apple? Apple have way more resources than the Galeon team. Gecko is already one of the most advanced renders out there, they wouldn't even have needed to touch most of the code.
Secondly, KHTML is still way way immature compared to Gecko. It only recently got support for XML (basic support only). It's still catching up in terms of core standards compliance, and forget about stuff like XSLT, MathML, etc. That's not to bash KHTML, what's there is good, but in terms of usefulness in browsing the web, Gecko owns it. On my site, over 50% of the hits come from Gecko based browsers, something like 40% from IE and I think about 3-4% from Konqueror. KHTML and Gecko have been choices on Linux for a long time, yet most seem to use Gecko.
To be honest, I think they chose KHTML because it was hard to make Gecko efficiently use the unusual Mac rendering model. Web browsing was really showing up the fact that Macs are slow these days, in ways that can't be disguised using hardware acceleration, or windowing system tricks etc. KHTML could be more easily hacked to get raw speed, which is clearly more important to them than features or website compatability.
I use Mozilla, and its memory usage when I last looked (yesterday) was 81MB. In contrast Opera was sitting at 10MB, rendering pages faster and supporting CSS better (Moz still doesn't support CSS counters, so I can't number headings automatically, for example.)
But Opera still has sucky DOM support (i think) etc etc. 81mb sounds very large indeed, I've never seen Moz use even half that. Bear in mind accurately measuring memory usage is hard with standard OS tools, as they usually don't adequately distinguish between shared libs etc.
A netscape ISP? You mean, like this one?
AOL don't want to put themselves in a position where Microsoft can hold them over the barrel - sure, the ActiveX control can be used by them today, but what about tomorrow? Being independant for them makes sense.
Having a proper, working multicast system well supported by the backbone routers would rewrite the rules of the net. Radio stations wouldn't be limited by needing constant donations of bandwidth, large file downloads could be put on a carousel saving more bandwidth, even websites if you were clever enough.
Using tunnelling doesn't give ISPs any incentive to move to IPv6 natively though, so we don't get the benefits of that.