What is it with graphics card demos and fairies? I know they're written almost exclusively by young guys, but still - the Matrox G400 had a rather curvaceous fairy for their tech demo as well. OK, so obviously not quite as realistic as this one, but couldn't they be more imaginative?
I mean if you're going to have medieval fantasy characters in your demos, might as well go the whole way and have a proper Princess.
The main reason I think Red Hat have not shipped Java before is that Suns JVM is not free. There is a free implementation of java in the form of gcj and GNU Classpath, I'm guessing that is not being used because this is a Sun-specific version of RHL.
Nonetheless, considering that they've indirectly funded the development of a free Java implementation, it's a pity it wasn't used.
Yes, KIO is nice, certainly more useful to the end user than gnome-vfs. On the other hand, they both suck and should be deprecated in favour of a desktop-neutral solution that doesn't play hell with FS semantics, and isn't used as a generic tree view model plugin system.
I also think the concept of apps being desktop specific should die. GNOME/GTK is better in this respect, short of a few problems your average HIGified GTK app feels pretty integrated. Just a few details left to hammer out.....
I see your point, but I have quite a few issues with the DocBook DTD that make it harder to write, and ridiculously hard to produce decent stylesheets (and oh boy, does DocBook need decent stylesheets). The sheer scope of the thing is one problem, it's hard to figure out exactly which elements you should use. The other is the verbosity of the element names. Stuff like para instead of p, variablelistentry instead of li. HTML is far more pleasant to write.
Split pane support isn't really needed, Nautilus opens new windows fast enough that it's simpler to have two open at once and drag between them, as in MacOS. By fish I assume you mean ssh or scp, which there is in fact a gnome-vfs plugin for (dunno if redhat ship it as standard though).
If they don't start giving the users what they want, then I can see a fork coming right this way.
If there was going to be a fork, it'd have happened way before now.
FreeBSD is real life. Solaris is real life. AIX is real life. Mac OS X is real life. For the most part, they all work right out of the box. Why can't Linux?
It can, just seemingly not for you. Hence my assertion that life is not perfect.
Bull. Would you like to explain how I was supposed to get the NVidia drivers working without changing my kernel?
Sure, run the installer. That's it. It'll install the right module for your kernel, and if it's unusual it'll build one for you itself, nothing to it.
Alright, now I know you're trolling. *Any* major system upgrade is dangerous.
The difference between 8 and 9 is only six months. The only serious breakage was NPTL, which was well publicised beforehand.
Uh huh. Non sucky you say? Why doesn't it *#%$# work!? And more to the point, how do I get it all into a truly portable format (like text)?
It does work. And it IS text, go take a look in ~/.gconf - it's all XML, ready for your editing pleasure.
Heh, interesting. I'm exactly the other way around. When GNOME2 came out, I decided to compile it on my SuSE box using garnome (2 days! i was on dialup back then) and decided I much preferred it to KDE. Last week I moved to redhat 9, and am very happy with it.
Personally I don't see what the big deal over KDE was. Mandrake also use a global theme. The other changes they made were minor, except altering some apps to use the best, instead of whatever happens to use Qt or KDE. As I always had to do that myself with SuSE, it's nice for it to be done for me these days.
good luck brothers! i fear this battle will be the biggest linux has ever faced
.... and Saurons minions will not be allowed to triumph! To mount doom!
ALL: To Mount Doom!
Re:GNOME needs more user friendly documentation.
on
Glade 2 Tutorial
·
· Score: 2, Interesting
Well, personally I've found the GTK documentation alrightish, though gtk-doc pages take a bit of getting used to. Wading through a huge synopsis is a bit annoying at first, until you get more familiar with it, and then being able to rapidly jump to the function you want using Geckos type ahead find is just plain cool.
Unfortunately, it's still put to shame by MSDN. The docs are sometimes missing, sometimes wrong..... when I encounter these things I write a patch and submit it, it only takes a few minutes, and if more people did it the docs would improve faster.
The docs for the GNOME libs though are a bit poor, but the same could be said of KDE, a lot of it is out of date, too sparse etc.
Really, at some point Linux needs to leave its fascination with SGML/DocBook behind. I hate it for so many reasons. It really sucks. A custom solution I think could do what DocBook does better, faster and with less hassle.
It'd be nice to have better searching as cross referencing as well. DevHelp is OK but rather buggy (I'm thinking of trying to fix a few of those bugs this week). As ever, I'd love to try and fix them, I like the technologies, but I have little time and other projects take higher priority. Really somebody just needs to be contracted to work on them for a bit, a developer support role perhaps. If there are any companies out there who want to use GTK for their apps but want proper developer support, get in touch!
You really shouldn't be using the code generation, in fact in Glade 3 I think it's going to be ripped out. It's far, far, far easier to load.glade XML files with libglade. In C, it will even autoconnect all your signals for you!
Do you see a pattern emerging of "big problem, big fix, more big problems, more big fixes"? This stuff has been going on in Linux for YEARS now and it's driving me insane!
Tough. This is real life, things are not perfect. Luckily though, they improve. You don't seem to like this either - you bitch and moan, and then when things are improved to make them easy enough, you claim it's too late and Linux still sucks. Well, it seems you are impossible to please.
I have a 4.2 FreeBSD machine happily chugging away as a server/remote desktop. I have never felt the need to run around doing dangerous upgrades to half the system. WHY CAN'T LINUX DO THIS?
It can. Easily. You don't HAVE to upgrade you know. There's nothing magic about FreeBSD that means it's always got the latest features but you never need to upgrade. There's also nothing dangerous about it, going from RH8 to RH9 is mostly automatic.
GConf? Err... what does that have to do with desktop settings like icons? Those are configured through a Nautilus interface.
Yes, and Nautilus gets its settings from GConf. It's just a configuration system. Think registry, except non-sucky.
Fair enough. My largest complaint is these types of things *always* happen. I'm starting to understand why Gnome is the default for Linux. The two are in philosophical agreement.
Oh come on. You're just a BSD zealot in disguise, I can tell by the "linux is always broken but BSD is solid" attitude. For some reason many, many people don't seem to share that opinion. If you insist on trying to patch stuff together yourself, and can't do it, don't bitch on Slashdot when it breaks.
1.4 wasn't all that long ago.
It was over 2 and a half YEARS ago. That's a long time in free software land.
However, that would explain why Gnome can be an actually usable desktop if hacked like hell by the distro
The differences between GNOME as shipped by gnome.org and Redhat are minor - mostly a matter of artwork and default panel layout (obviously redhat provide some config tools as well).
This sounds like a stupid troll to me. Installing the nVidia drivers is easy, you just run the installer at runlevel 3 and then switch to the "NVidia" driver instead of "nv". Of course for 2D work it doesn't really matter, I tried both and found that in Redhat 9 the free driver felt just as fast, so I stuck with it. That's it. No RPMs, no screwing about with kernel modules.
OK, so lets see about the GNOME problems. As already pointed out, these problems seem to be caused by bad packaging. That goes for icons AND the menus. Nautilus themes are gone, Nautilus now synchs to the GNOME icon theme, to make it more integrated with the desktop. You CAN set this, or I can at any rate, so I don't really know what your problem is here, except perhaps a failure to read the release notes.
GConf does not "serialize objects", it's a simple way to store user settings in XML, with an auto-update API and such. Copying user settings involves copying a couple of directories. I've done it loads of times, it works fine.
System tray icons in GAIM should work fine in GNOME, they do for me etc etc. To be honest I think you just suffered from a screwed build or a bad setup. Yes, if you let people who know what they're doing tweak it, GNOME can really rock, but to be honest I used to have a garnome desktop and that worked fine too. I think you just had a bad experience, which doesn't generalise.
As for the last comment, Miguel de Icaza hasn't been involved with GNOME since the 1.4 days - if you're going to make dumb flamebait comments, at least make them relevant.
Last time I checked, Hailstorm was still very much alive, but buried deep within the company. It's still gestating, developing, until the market is ready. They aren't forgetting how lucrative the market for personal services could be. You just wait and see.
Of course, that doesn't mean that a toolkit will necessarily use those features, but it's a cool thing to have available.
Right, so from the DPS for X website:
The main problem with DPS is that it is a complex interface that requires, for efficient implementation, a fair amount of state and code to be resident in the server.
For reasons of both efficiency and functionality, it is often better to use DPS as a low-level rendering interface and put high-level libraries on the client side. Thus, the high-level features of DPS, while present in every implementation, are not actually used and are therefore unused baggage. This fact is explicitly acknowledged by recent implementations of DPS, which include features such as user paths that are not useful without an additional library layer above them.
Actually for pure 2D (ie desktop), I find the nv driver in XFree 4.3 to be as fast as the nvidia one subjectively speaking. For redhat 8 I could really tell the difference, but I installed the proprietary NVidia drivers under 9 and actually went back to the nv driver - I couldn't tell the difference, and the NVidia driver seemed to cause flickering on the mouse cursor, took ages to start up and so on.
Indeed, and according to dps.sf.net, it seems a lot of DPS was never actually used and apps did their own rendering anyway. The plethora of custom widgets in Apples own software would seem to affirm this belief.
Certainly there is nothing that says you can't run a proprietary windowing system on top of the kernel, is there?
Well, yeah, there's this thing called the GNU manifesto, maybe you should read it. It says things like, we should have a completely free OS, and that open standards should be used where possible. X11 is open. Aqua/Quartz is not. Nuff said.
Anyway, it's far from obvious to those in the display design system community that DPS is a superior system. In particular, the team working on adding it to Xfree stopped when RENDER was designed - they claim it's a far better solution to the problem than DPS ever was, and that it recognises real world needs more.
This is interesting, but hardly news. Take a look around - adverts for.NET and Visual Studio flood this place every single day. It's one of the biggest ironies of the open source movement IMHO. Even without odd corporate ties, Malda and the crew have been paid by Microsoft for a long time now.
So, that leaves us to speculate as to why. Well, I think that's pretty obvious too, nothing strange going on here. Slashdot is a concentrate of a huge number of IT/tech oriented readers. Not all of them hate Microsoft in fact - remember only 50% of slashdots readers even look at the comments, only perhaps 1-2% actually post. So, they get more "bang for their buck" by targetting Slashdot than most websites. They don't have to work hard to find people interested in programming - we're all right here.
And as for those who hate Microsoft? Well, MS are a monopoly, never forget that. There is only so much preaching to the choir that you can do. They want to sell to people who don't already use their products, as well as those who do. What better place to advertise than here? Why, it's almost as if everybody who is no longer their customer passes through these doors at some point. A marketing mans wet dream.
And so why wouldn't they want to keep Slashdot afloat? Letting it die would satisfy some Microsoft employees I'm sure - but they're geeks too you know, and some (many?) read it as well. Might as well keep all your potential customers in one place, rather than let them float away on the winds and currents of the net. Makes sense, to me at any rate. I'd do it if I were them.
And did you report this rep? What did Redhat say? Did they apologise?
I'm sure you're being honest, but given your track record as somebody rabidly pro-MS and often anti-Linux, allegations against random redhat employees carry less weight than they otherwise would.
Considering that basically the only message that resonated with MS customers about Linux was "you might get sued for using it", is it any wonder people are suspicious about SCO?
Blargh. It sucks for everybody who doesn't use Debian (ie most). And of course with Debian you're reliant upon a developer packaging it for you, keeping it up to date, and you have to use unstable if you want to stay up to date. I'm talking about a real solution that scales to all the software in the world. See sig;)
In the past, Microsoft has mainly concerned itself with positioning Windows NT based servers against the superior Linux-based products from Debian, Red Hat, and Caldera.
Windows NT is fading away. Win2003 is a good piece of work from what I've seen/heard - I wouldn't be so fast to declare Linux superior, not any more. If you think Microsoft are just going to sit still while Linux motors on, think again. They move fast too.
This memo demonstrates an important shift in their strategy: they are now in a position where they are competing against Linux on thedesktop, having lost many key battles on the server side.
I'm pretty sure Windows has a higher market share in the server side of things (still). Sure, Linux is growing quickly, and it's hurting Windows, but it's easy to forget amidst all the hype that Linux is still the little guy, even after all these years.
The last bit of the rant I can't agree with either. Desktop Linux is not "ready", where by ready I mean I would be happy giving it to most reasonably intelligent computer users. We're not there yet, the software generally needs more spit and polish, and we need to get software installation really nailed. Too much stuff is just currently plain old broken (menus anybody?)
MacOSX is an alternative window manager sitting on top of BSD!
So why won't iTunes run on FreeBSD then? I tried it, it didn't work.
I think you get the point. Gross exaggurations are one thing. This is something else.
Re:Yeah. let's depend on IBM for our future
on
More on the PowerPC 970
·
· Score: 2, Interesting
Even Windows can't deliver consistently on its promise of universal ease-of-use because so many vendors have so much hardware that may or may not work with the system and its existing drivers.
Y'know, I have no idea why this keeps coming up. Assuming you buy from a reputable vendor, the only time you have hardware problems with a PC is if you do upgrades manually, or if something breaks (rare). The fact that Macs have a reputation for good integration is mostly marketing - if you go down to your local PC World and walk out with a new box, that'll work just fine too. If you then try and plug in ancient or super new hardware in an attempt to recycle stuff for instance, then yeah, you might get breakage. But that happens with Macs too, the only reason you don't see it is because there is practically no upgrade path outside of buying a new machine.
Whatever else you think about Apple's computers, they are without a doubt the easiest PCs on the planet if you're a neophyte.
No they aren't. Everybody I've seen who has been sat down in front of a Mac found it hopelessly confusing and non-intuitive. The only people I know who stuck with them, are those who bought them personally (they would, wouldn't they). I know I had to have the owner of said Mac sit next to me and guide me when I was trying to use his machine, and I'm far from being a neophyte. Stupid differences from Windows and idiotic conventions that had seemingly no basis in actual usability just pissed me off. Perhaps for people who have never used a computer before in their life it's easier than Windows (but I doubt it) - for people who have (the majority) it's just a pain in the ass.
Apple cares about making a good and easy-to-use product, or else they'd just be chasing Windows like (sorry, not trolling here, but it's true) GNOME and KDE are instead of constantly innovating their own hardware and interface designs.
Sorry, but you are trolling here. Apple aren't chasing Windows because regardless of what the majority of users want, their existance is justified in their customers eyes by the fact that they are different. People just take it as read that different equals better, despite a lack of compelling (objective) evidence to the contrary. Go read some usability reviews of MacOS X by long time users of the platform.
If you think it's about "more than just money" you need to wake up and smell the roses. Quite how a publically owned company, with a shrinking market share can be allowed to be motivated by anything other than money is beyond me. They're a business, their legal obligation is to their shareholders first and foremost. Wishy washy ideas about design purity might have had some merit back when it was just Jobs and Woz in their garage, but that Apple died years ago.
Anyway, if you want something that isn't motivated by money and is about building a quality product, Linux is about the only thing that qualifies. At the end of the day, the product and the ability for people to use it (in both a usability and a licensing sense) is everything. Go read and take part in the desktop discussion lists if you don't believe me. You might not like what it is today, but that's an entirely separate issue. Money isn't, cannot be, a serious motivation for these guys as the vast majority are not paid for it.
Targetting multiple architectures means that Apple's got to deal with unpredictable hardware configurations, cards, motherboards, drivers, all sorts of things that could cause inconvenient kernel panics, drive failures, or worse. Users are used to that with Windows, and they pretty much expect it with Linux. With Macs, they expect things to just work. Controlling the hardware is the best way for Apple to do that.
CPU architectures have nothing to do with driver instability, nor mixing hardware. That would only be an issue if Apple tried to write a PC version (as opposed to an x86 version) of MacOS, but hey, you know what? The world is a messy place. People don
Applied to the cyc project: a collection of facts is not necessarily intelligence.
No, but a collection of facts is required before you can integrate other facts into the system - ie learn things. It's also probably a lot more useful than a lot of the research into AI being done today.
Slashdot Mirror
I mean if you're going to have medieval fantasy characters in your demos, might as well go the whole way and have a proper Princess.
Nonetheless, considering that they've indirectly funded the development of a free Java implementation, it's a pity it wasn't used.
I also think the concept of apps being desktop specific should die. GNOME/GTK is better in this respect, short of a few problems your average HIGified GTK app feels pretty integrated. Just a few details left to hammer out.....
I see your point, but I have quite a few issues with the DocBook DTD that make it harder to write, and ridiculously hard to produce decent stylesheets (and oh boy, does DocBook need decent stylesheets). The sheer scope of the thing is one problem, it's hard to figure out exactly which elements you should use. The other is the verbosity of the element names. Stuff like para instead of p, variablelistentry instead of li. HTML is far more pleasant to write.
If they don't start giving the users what they want, then I can see a fork coming right this way.
If there was going to be a fork, it'd have happened way before now.
It can, just seemingly not for you. Hence my assertion that life is not perfect.
Bull. Would you like to explain how I was supposed to get the NVidia drivers working without changing my kernel?
Sure, run the installer. That's it. It'll install the right module for your kernel, and if it's unusual it'll build one for you itself, nothing to it.
Alright, now I know you're trolling. *Any* major system upgrade is dangerous.
The difference between 8 and 9 is only six months. The only serious breakage was NPTL, which was well publicised beforehand.
Uh huh. Non sucky you say? Why doesn't it *#%$# work!? And more to the point, how do I get it all into a truly portable format (like text)?
It does work. And it IS text, go take a look in ~/.gconf - it's all XML, ready for your editing pleasure.
Personally I don't see what the big deal over KDE was. Mandrake also use a global theme. The other changes they made were minor, except altering some apps to use the best, instead of whatever happens to use Qt or KDE. As I always had to do that myself with SuSE, it's nice for it to be done for me these days.
ALL: To Mount Doom!
Unfortunately, it's still put to shame by MSDN. The docs are sometimes missing, sometimes wrong..... when I encounter these things I write a patch and submit it, it only takes a few minutes, and if more people did it the docs would improve faster.
The docs for the GNOME libs though are a bit poor, but the same could be said of KDE, a lot of it is out of date, too sparse etc.
Really, at some point Linux needs to leave its fascination with SGML/DocBook behind. I hate it for so many reasons. It really sucks. A custom solution I think could do what DocBook does better, faster and with less hassle.
It'd be nice to have better searching as cross referencing as well. DevHelp is OK but rather buggy (I'm thinking of trying to fix a few of those bugs this week). As ever, I'd love to try and fix them, I like the technologies, but I have little time and other projects take higher priority. Really somebody just needs to be contracted to work on them for a bit, a developer support role perhaps. If there are any companies out there who want to use GTK for their apps but want proper developer support, get in touch!
You really shouldn't be using the code generation, in fact in Glade 3 I think it's going to be ripped out. It's far, far, far easier to load .glade XML files with libglade. In C, it will even autoconnect all your signals for you!
Tough. This is real life, things are not perfect. Luckily though, they improve. You don't seem to like this either - you bitch and moan, and then when things are improved to make them easy enough, you claim it's too late and Linux still sucks. Well, it seems you are impossible to please.
I have a 4.2 FreeBSD machine happily chugging away as a server/remote desktop. I have never felt the need to run around doing dangerous upgrades to half the system. WHY CAN'T LINUX DO THIS?
It can. Easily. You don't HAVE to upgrade you know. There's nothing magic about FreeBSD that means it's always got the latest features but you never need to upgrade. There's also nothing dangerous about it, going from RH8 to RH9 is mostly automatic.
GConf? Err... what does that have to do with desktop settings like icons? Those are configured through a Nautilus interface.
Yes, and Nautilus gets its settings from GConf. It's just a configuration system. Think registry, except non-sucky.
Fair enough. My largest complaint is these types of things *always* happen. I'm starting to understand why Gnome is the default for Linux. The two are in philosophical agreement.
Oh come on. You're just a BSD zealot in disguise, I can tell by the "linux is always broken but BSD is solid" attitude. For some reason many, many people don't seem to share that opinion. If you insist on trying to patch stuff together yourself, and can't do it, don't bitch on Slashdot when it breaks.
1.4 wasn't all that long ago.
It was over 2 and a half YEARS ago. That's a long time in free software land. However, that would explain why Gnome can be an actually usable desktop if hacked like hell by the distro
The differences between GNOME as shipped by gnome.org and Redhat are minor - mostly a matter of artwork and default panel layout (obviously redhat provide some config tools as well).
OK, so lets see about the GNOME problems. As already pointed out, these problems seem to be caused by bad packaging. That goes for icons AND the menus. Nautilus themes are gone, Nautilus now synchs to the GNOME icon theme, to make it more integrated with the desktop. You CAN set this, or I can at any rate, so I don't really know what your problem is here, except perhaps a failure to read the release notes.
GConf does not "serialize objects", it's a simple way to store user settings in XML, with an auto-update API and such. Copying user settings involves copying a couple of directories. I've done it loads of times, it works fine.
System tray icons in GAIM should work fine in GNOME, they do for me etc etc. To be honest I think you just suffered from a screwed build or a bad setup. Yes, if you let people who know what they're doing tweak it, GNOME can really rock, but to be honest I used to have a garnome desktop and that worked fine too. I think you just had a bad experience, which doesn't generalise.
As for the last comment, Miguel de Icaza hasn't been involved with GNOME since the 1.4 days - if you're going to make dumb flamebait comments, at least make them relevant.
Last time I checked, Hailstorm was still very much alive, but buried deep within the company. It's still gestating, developing, until the market is ready. They aren't forgetting how lucrative the market for personal services could be. You just wait and see.
Right, so from the DPS for X website:
Actually for pure 2D (ie desktop), I find the nv driver in XFree 4.3 to be as fast as the nvidia one subjectively speaking. For redhat 8 I could really tell the difference, but I installed the proprietary NVidia drivers under 9 and actually went back to the nv driver - I couldn't tell the difference, and the NVidia driver seemed to cause flickering on the mouse cursor, took ages to start up and so on.
Indeed, and according to dps.sf.net, it seems a lot of DPS was never actually used and apps did their own rendering anyway. The plethora of custom widgets in Apples own software would seem to affirm this belief.
Well, yeah, there's this thing called the GNU manifesto, maybe you should read it. It says things like, we should have a completely free OS, and that open standards should be used where possible. X11 is open. Aqua/Quartz is not. Nuff said.
Anyway, it's far from obvious to those in the display design system community that DPS is a superior system. In particular, the team working on adding it to Xfree stopped when RENDER was designed - they claim it's a far better solution to the problem than DPS ever was, and that it recognises real world needs more.
So, that leaves us to speculate as to why. Well, I think that's pretty obvious too, nothing strange going on here. Slashdot is a concentrate of a huge number of IT/tech oriented readers. Not all of them hate Microsoft in fact - remember only 50% of slashdots readers even look at the comments, only perhaps 1-2% actually post. So, they get more "bang for their buck" by targetting Slashdot than most websites. They don't have to work hard to find people interested in programming - we're all right here.
And as for those who hate Microsoft? Well, MS are a monopoly, never forget that. There is only so much preaching to the choir that you can do. They want to sell to people who don't already use their products, as well as those who do. What better place to advertise than here? Why, it's almost as if everybody who is no longer their customer passes through these doors at some point. A marketing mans wet dream.
And so why wouldn't they want to keep Slashdot afloat? Letting it die would satisfy some Microsoft employees I'm sure - but they're geeks too you know, and some (many?) read it as well. Might as well keep all your potential customers in one place, rather than let them float away on the winds and currents of the net. Makes sense, to me at any rate. I'd do it if I were them.
I'm sure you're being honest, but given your track record as somebody rabidly pro-MS and often anti-Linux, allegations against random redhat employees carry less weight than they otherwise would.
Considering that basically the only message that resonated with MS customers about Linux was "you might get sued for using it", is it any wonder people are suspicious about SCO?
Blargh. It sucks for everybody who doesn't use Debian (ie most). And of course with Debian you're reliant upon a developer packaging it for you, keeping it up to date, and you have to use unstable if you want to stay up to date. I'm talking about a real solution that scales to all the software in the world. See sig ;)
Windows NT is fading away. Win2003 is a good piece of work from what I've seen/heard - I wouldn't be so fast to declare Linux superior, not any more. If you think Microsoft are just going to sit still while Linux motors on, think again. They move fast too.
This memo demonstrates an important shift in their strategy: they are now in a position where they are competing against Linux on thedesktop, having lost many key battles on the server side.
I'm pretty sure Windows has a higher market share in the server side of things (still). Sure, Linux is growing quickly, and it's hurting Windows, but it's easy to forget amidst all the hype that Linux is still the little guy, even after all these years.
The last bit of the rant I can't agree with either. Desktop Linux is not "ready", where by ready I mean I would be happy giving it to most reasonably intelligent computer users. We're not there yet, the software generally needs more spit and polish, and we need to get software installation really nailed. Too much stuff is just currently plain old broken (menus anybody?)
So why won't iTunes run on FreeBSD then? I tried it, it didn't work.
I think you get the point. Gross exaggurations are one thing. This is something else.
Y'know, I have no idea why this keeps coming up. Assuming you buy from a reputable vendor, the only time you have hardware problems with a PC is if you do upgrades manually, or if something breaks (rare). The fact that Macs have a reputation for good integration is mostly marketing - if you go down to your local PC World and walk out with a new box, that'll work just fine too. If you then try and plug in ancient or super new hardware in an attempt to recycle stuff for instance, then yeah, you might get breakage. But that happens with Macs too, the only reason you don't see it is because there is practically no upgrade path outside of buying a new machine.
Whatever else you think about Apple's computers, they are without a doubt the easiest PCs on the planet if you're a neophyte.
No they aren't. Everybody I've seen who has been sat down in front of a Mac found it hopelessly confusing and non-intuitive. The only people I know who stuck with them, are those who bought them personally (they would, wouldn't they). I know I had to have the owner of said Mac sit next to me and guide me when I was trying to use his machine, and I'm far from being a neophyte. Stupid differences from Windows and idiotic conventions that had seemingly no basis in actual usability just pissed me off. Perhaps for people who have never used a computer before in their life it's easier than Windows (but I doubt it) - for people who have (the majority) it's just a pain in the ass.
Apple cares about making a good and easy-to-use product, or else they'd just be chasing Windows like (sorry, not trolling here, but it's true) GNOME and KDE are instead of constantly innovating their own hardware and interface designs.
Sorry, but you are trolling here. Apple aren't chasing Windows because regardless of what the majority of users want, their existance is justified in their customers eyes by the fact that they are different. People just take it as read that different equals better, despite a lack of compelling (objective) evidence to the contrary. Go read some usability reviews of MacOS X by long time users of the platform.
If you think it's about "more than just money" you need to wake up and smell the roses. Quite how a publically owned company, with a shrinking market share can be allowed to be motivated by anything other than money is beyond me. They're a business, their legal obligation is to their shareholders first and foremost. Wishy washy ideas about design purity might have had some merit back when it was just Jobs and Woz in their garage, but that Apple died years ago.
Anyway, if you want something that isn't motivated by money and is about building a quality product, Linux is about the only thing that qualifies. At the end of the day, the product and the ability for people to use it (in both a usability and a licensing sense) is everything. Go read and take part in the desktop discussion lists if you don't believe me. You might not like what it is today, but that's an entirely separate issue. Money isn't, cannot be, a serious motivation for these guys as the vast majority are not paid for it.
Targetting multiple architectures means that Apple's got to deal with unpredictable hardware configurations, cards, motherboards, drivers, all sorts of things that could cause inconvenient kernel panics, drive failures, or worse. Users are used to that with Windows, and they pretty much expect it with Linux. With Macs, they expect things to just work. Controlling the hardware is the best way for Apple to do that.
CPU architectures have nothing to do with driver instability, nor mixing hardware. That would only be an issue if Apple tried to write a PC version (as opposed to an x86 version) of MacOS, but hey, you know what? The world is a messy place. People don
No, but a collection of facts is required before you can integrate other facts into the system - ie learn things. It's also probably a lot more useful than a lot of the research into AI being done today.