Well, if they create webservices for wikipedia/wiktionary, everybody will be able to use it, not just KDE... I don't see what will prevent me to write a GNUstep client for example:-)
WebObjects was (and still is) one of the most powerful web application system. Much more sensible than a lot of stuff:-)
EOF -- an object relational mapper, providing isolation from the database and from the database model -- in particular is very, very nice. Not the final answer to everything, but still quite cool:-)
The sad thing with Apple's current WebObjects is that it's only java (it's even a J2EE environment), while originally (at NeXT) it was Objective-C based (plus WebScript, an ObjC-like script language). They dropped the Objective-C bit with WebObjects 5, sadly (4.5 had ObjC and Java). Well, ok, beeing a J2EE env has its own advantages, but still...
The documentation of WO 4.5 is here, the documentation for the current WO is here.
There is a free software implementation of WebObjects 4.5 from the GNUstep project, GNUstepWeb, which work well. OpenGroupware.org also has its own WO 4.5 implementation, NGObjWeb, which works very well too (it's the foundation of SOPE). I wrote an article showing how to do simple (html) components, but it's in french;-)
Though, if you want to discover a really interesting project, have a look to Seaside. It's inspired by WebObjects, with an excellent component model, but is even better (support of continuations, etc). And it's completely dynamic, letting you change things at runtime easily (Smalltalk rulez;-). It's one of the best thing I know:-)
The French are famous for being mean to the Quebecois ? what ? I am french, and frankly I think that most french people/love/ Quebec, and their fantastic accent (really!). Meeting a few assholes that treat you badly doesn't make a rule.
As far as I remember, it was not a proposal by the french government, it was the french academy, the ones that are in charge of the dictionaire. And it wasn't the word "courriel" (a rather nice word imho, invented by the Quebecois) but the word "mél" which is indeed, awful. I don't remember a government policy of using the word "courriel", but if that was the case, why not ? this word sounds better than "mail", frankly.
Trying to push "mél" and "cédérom" was quite stupid. "Courriel" on the other hand is nice and a good contraction from "Courrier électronique" -- like email.
In any cases, nobody is forced to use "Courriel" or "Mél" (berk!). But I for example ofent use "Courriel", just because I think it sounds better.
What you don't realize is that "learning Objective-C" is frankly not difficult or long. The longest thing is NOT to learn a language, particularly one as simple as Objective-C, but to learn the frameworks.. Learning Cocoa isn't immediate, but at least there's a nice payoff at the end -- Cocoa is easily one of the best GUI toolkit, very well designed (even if..) and powerful, and InterfaceBuilder in itself is worth it.
So, frankly, you won't save time by learning Cocoa#, to the contrary. Plus, Objective-C is a very cool object-oriented language (object-oriented in the sense of Smalltalk, not in the sense of C++ !), just the minimum additions to enable OOP on top of C. Neat.
By the way, don't forget the GNUstep project, a free software implementation of the OpenStep API, with Gorm, an equivalent to IB ! and all that runs on *nix, linux, mac and windows..
he came from the island of Corsica, which I believe was a French territory at the time.
You know, Corsica is still a french "territory" ? In fact it's just a french department like another.. (even if they have independentist, but frankly, the independentist movement in Corsica is more alike to mafia than anything at the moment.. and yes, part of my family is from Corsica).
The problem is, it's exactly BECAUSE there is not a strong EU, particularly a strong parliament, that we are in this mess. And why don't we have a strong parliament? because the people against a federal EU prevented it. The problem is not the EU, the problem is a weak parliament and powerful unelected bodies (council of ministers and european commission).
So now it's rather hypocritical to say that it's all EU's fault and thus people against EU are right. It's mostly their fault as well.
You should realize that in this case, the parliament did its job, listened to the citizen, and activaly "patched" the proposed directive (a better way than simply refusing it, as it wouldn't have stopped the patents proponents to submit again and again, like we saw..) to largelly remove all harm.
The problem then is that the minister's council and the commission (both unelected bodies) just decided to "forget" the parliament modifications. Unbelievable. I'm telling you, the parliament wasn't pleased !
Hopefully they will stop this shit, if only because they are pissed by the council/commission behavior.. and we can help convaincing them by contacting them:)
But you should realize that all "local" politicians use the EU as an excuse -- "hey it's not our fault, it's Brussels !!" -- while in fact, THEY are still in power (with the commission and the council), and for the parliament, well, you voted for it.
The EU is an excellent scapegoat for the national politicians, and nobody seems to realize that. Like if the national politicians were just unconnected to the delegation in brussels. Yeah, sure.
Uh, did you read the part where I said that if your "representatives" don't answer to the citizens that you have bigger problems than software patents? I know the EU is a fucked up idea. I am glad that some of the people in the EU are starting to realize this. Don't blame it on the US.
The EU is NOT a fucked up idea, the current organisation is. Of course, the US are not exactly pleased by the increasing power of the EU, but hey...
By the way, software patents are GOOD. They DO protect the small developer. As a small developer who has a couple of software patents that I have successfully licensed, I can PERSONALLY vouch for them. Of course, the slashbots don't want to hear this. The current issue with patents isn't the fact that there are software patents, but maybe the fact that there are cases where they have been granted without a good reason. Saying that "patents are bad" is just silly.
Bullshit. Even "normal" patents have bad side effects for the famous "little guy", and we're here talking about SOFTWARE patents, patents on ideas. That's the dumbest thing ever. I doubt that you're saying the truth with your "I can vouch for them" (ie, as an AC I think you're just pulling shit out of your ass), but even if that's the case, you should realize that your situation is the exception, not the norm, and by far. Software patents are used by big company to stiffle innovation. Ask bill gates, he wrote it black on white.
OpenStep isn't really a "frozen standard" any more, because the single widely-used implementation (OS X) continues to evolve.
The fact that Apple adds things to Cocoa (not that much by the way) doesn't suddenly transform OpenStep as a moving target. Cocoa is a moving target, not OpenStep. And guess what ? GNUstep already implements OpenStep (minus some rather unused classes).
The one big problem with GNUStep is that they don't care enough about replicating Apple's changes. : (
And how did you actually figured that ?
GNUstep primary goal was to implement OpenStep, yes. But Apple's changes are also added. Saying that "we don't care enough about replicating Apple's changes" is just plain wrong. If the changes are sound and, moreover, if somebody actually provides some code, I can guarantee you that it will be integrated. As always, it's a matter of coding. If you want to help, just start to code.. it's a free software project you know.
There is two free implementations of WebObjects -- one done by the GNUstep project, GNUstepWeb, and another that's done by OpenGroupware.org (they use it as their base for SOPE), called NGObjWeb.. both implements WebObjects 4.5 (ie, the version before java) and are used commercially.
Both works quite well, but they lack a scripting language (although the OGo guys works on a WebScript-like language, and the GNUstepWeb guys works on integrating StepTalk).
About EOF, the GNUstep project have an implementation named GDL2 (GNUstep Database 2) which implements EOF 4.5 -- but sadly we are missing an EOModeller (even if somebody is working on one). Here an article from Ludovic Marcotte about GDL2.
They are as ugly as the hind end of a dog with a hat on it. The first few minimal ones just make the clean GNUstep app look amateurish and childish with misaligned text in the buttons and garish coloured arrows to indicate the default button; the ones that attempt to imitate Aqua are actively ugly.
Well, none of them are intended to be full-featured themes. Mostly, they are just snapshot of different tests and modifications. If you want a theme engine, you *need* to test things;-) -- but I surely won't advocate taking one of the ugly themes here as a good UI example. I quite like the general idea of this one for example, but it's obvious it needs works. On the other hand, when you say that the OSX themes looks bad, take a look at that one -- that's really close to OSX (well, minus wmaker titlebar, of course). Anyway, there are things to correct, and they don't look perfect (like for example, the misaligned text). But mostly it was to test different features and modifications... so yes, it was because I could:-)
Don't forget it's just a theme engine -- GNUstep will still have the NeXT look by default. And you know what ? personally I tend to prefer the NeXT UI (that's partly why I didn't work as much as I should have done on Camaelon). But as you can read it on slashdot, people won't use GNUstep without a "modern" look (it's perhaps sad, but that's the reality). On a pragmatic way, having a theme engine will, regardless of slashdot users, be useful for cross platform (on Windows, etc).
Vim in Terminal.app is effectively a customization -- you can set your editor in ProjectCenter, and you can possibly use the integrated editor. But, the integrated editor is a bit poor in features, and above all I'm too used to vim:-)
The thing is, lots of gnustep dev like 1) the floating menus 2) the list selector widgets (personally that's one of my favorite widgets -- and by the way they still exist on Cocoa, and are used..) 3) the "empty" scrollbars is NOT a bad idea. Again, it's the case on Cocoa as well.
So your "not even MacOSX" comment is way off base. The only thing OSX does NOT is the floating menu, and a lot of people regret them.
Sure, Objective-C is the root of all evil and slow a system to death. That's why, as we have all seen with the recent (!) steve job video, a workstation with a 68030 (how much slower than our current hardware ?)
was undoubtely unbearably slow...
Well, HURD was supposed to be the Kernel of the GNU OS, and GNUstep in the beginning was supposed to be the GUI of the GNU OS. As recently Hurd progressed with the first results of the L4 port, it was a reference to that..
Like I said in a previous comment, I'm working on Camaelon 2, a pixmap theme engine that lets you have pretty things.. it should be officially released before the end of this month (it already works, I just want to clean up things).
I didn't see any support for layout management in Gorm
Well, I didn't show that part, but that works exactly in Gorm like on InterfaceBuilder on OSX (and imho it's a better model most of the time than the springs). So of course you can have resizable widgets. In addition, GNUstep implements a couple of widgets implemeting the spring resizing model (that's used by Renaissance by the way, an XML framework for describing UI for GNUstep...), so if you *really* want the spring model, you can use it.
The thing is, the NeXT machine used to have gamma corrected display. And so, in real use, it wasn't as dark as it can appear on screenshots. Alas, GNUstep gui, by default, use the same colors value -- and so, if your display is not correctly calibrated, it will appear too gray, too dark.
You can change the colors with the Preferences.app though, if you're not fond of the default gray !
The other possibility is to wait for Camaelon 2, a pixmap theme engine I'm working on. It works at the moment, and if you really want it now you can drop me a mail, but I'd like to clean up the code and add a few things. I intend to release it before the Fosdem (so, before the end of this month). Some screenshots like this one or
that one show what can be done.
Is it possible to push the user interface experience of GNUStep out of the dark, muddled, inorganic mess that it is now and into something more appealing, something, dare I say, more feminine?
Well, basically, NeXT overcharged their hardware, then their software. For example, you probably never heard of WebObjects, even if it was (still is actually) one of the best technology to create a dynamic website... and it's no wonder considering they used to sell it at insanely huge prices. Now you can have it via Apple for 500$...
this is big enough stuff that they should have been able to get through a few lean years and sell the technology....
Well, they did:-) -- to Apple...
Actually, the problem they had, is that nearly nobody in the industry was used to OOP. Now it's easier to understand the brilliance of NeXTSTEP's concepts, but it was probably more difficult to convaince people at the time ? (check the real media video on openstep.se/next/videos , they take half the video to explain the interest of OOP before introducing IB..)
And of course, a NeXT Cube and even a NeXT station were extremely expensive... too bad, they were 15 years ahead of their time (yeah, OSX is not as clean as NeXTSTEP, partly because of the need to integrate all theses existing apps..)
For thoses who want to see how programming is done in GNUstep, there's this short flash demo here
GNUstep is a free software implementation of the OpenStep API (like Cocoa), and it provides development tools as well. The demo steve do is doable in GNUstep as well..
(Yes, it's flash... a mpeg version will probably be available next week... in the meantime, it's a good idea to check either swift tools or swfdec , if you don't want or can't use the Macromedia Flash player..)
by the way, for thoses who will go to the Fosdem next february, there is a GNUstep developer's room, and there will be presentations, discussions, etc.
Actually, I could see two developments of the mac mini as a server that would sell quite well:
- first and most obviously, a mac mini "home digital center" -- with 5.1 output, tv tuner, tv out.
- second, a cheap brick for clustering, with a gigaethernet card (that doesn't automatically conflic with the "home digital center" idea..).
Mind you, this is the most exciting idea imho -- not because it could be used for some "traditional" clustering (even if I can imagine some people using it that way), but because it could be used to redefine the whole concept of Personal Computer and Home Computing..
Instead of having "one" computer, you would have a bunch of mac mini servers, cooperating. For geeks it could be, one mini for the database, one mini for the web server, etc. But for lambda mac user, it could be a different separation : one mini for the photos, one mini for the movies, etc.
The point would be that OSX would abstract the cluster, so people would have nothing complex to do. Just imagine -- "Gosh imovie is slow... hey I go buy a new brick, plug it on my network, and it will be faster!".
That could not be _that_ hard to implement for Apple -- don't forget that they control everything (hardware and software), so for them it's easier to push for this kind of innovation. And even if just iLife and system apps fully use the cluster, it could be useful (and 3rd party developers would anyway use it too after a bit).
Don't forget that OSX already have RendezVous, simplifying the building of this kind of "home cluster", and XGrid... so what I'm describing could be seen as a mere improvement of what already exist...
On a side note, I wonder if it's possible to easily "stack" the mac mini -- could be nice to have external peripherics in the same kind of case as the mini, and you would build you tower by stacking your equipment (PVR box, Firewire drive, music I/O, etc..). Talk about evolution;-) [it's probably not doable with the current mini, but it would be great to have that for the next one;-)]
Slashdot Mirror
Well, if they create webservices for wikipedia/wiktionary, everybody will be able to use it, not just KDE ... I don't see what will prevent me to write a GNUstep client for example :-)
EOF -- an object relational mapper, providing isolation from the database and from the database model -- in particular is very, very nice. Not the final answer to everything, but still quite cool :-)
The sad thing with Apple's current WebObjects is that it's only java (it's even a J2EE environment), while originally (at NeXT) it was Objective-C based (plus WebScript, an ObjC-like script language). They dropped the Objective-C bit with WebObjects 5, sadly (4.5 had ObjC and Java). Well, ok, beeing a J2EE env has its own advantages, but still...
The documentation of WO 4.5 is here, the documentation for the current WO is here.
There is a free software implementation of WebObjects 4.5 from the GNUstep project, GNUstepWeb, which work well. OpenGroupware.org also has its own WO 4.5 implementation, NGObjWeb, which works very well too (it's the foundation of SOPE). I wrote an article showing how to do simple (html) components, but it's in french ;-)
Though, if you want to discover a really interesting project, have a look to Seaside. It's inspired by WebObjects, with an excellent component model, but is even better (support of continuations, etc). And it's completely dynamic, letting you change things at runtime easily (Smalltalk rulez ;-). It's one of the best thing I know :-)
The French are famous for being mean to the Quebecois ? what ? I am french, and frankly I think that most french people /love/ Quebec, and their fantastic accent (really!). Meeting a few assholes that treat you badly doesn't make a rule.
As far as I remember, it was not a proposal by the french government, it was the french academy, the ones that are in charge of the dictionaire. And it wasn't the word "courriel" (a rather nice word imho, invented by the Quebecois) but the word "mél" which is indeed, awful. I don't remember a government policy of using the word "courriel", but if that was the case, why not ? this word sounds better than "mail", frankly.
Trying to push "mél" and "cédérom" was quite stupid. "Courriel" on the other hand is nice and a good contraction from "Courrier électronique" -- like email.
In any cases, nobody is forced to use "Courriel" or "Mél" (berk!). But I for example ofent use "Courriel", just because I think it sounds better.
So, frankly, you won't save time by learning Cocoa#, to the contrary. Plus, Objective-C is a very cool object-oriented language (object-oriented in the sense of Smalltalk, not in the sense of C++ !), just the minimum additions to enable OOP on top of C. Neat.
By the way, don't forget the GNUstep project, a free software implementation of the OpenStep API, with Gorm, an equivalent to IB ! and all that runs on *nix, linux, mac and windows..
You know, Corsica is still a french "territory" ? In fact it's just a french department like another.. (even if they have independentist, but frankly, the independentist movement in Corsica is more alike to mafia than anything at the moment.. and yes, part of my family is from Corsica).
So now it's rather hypocritical to say that it's all EU's fault and thus people against EU are right. It's mostly their fault as well.
You should realize that in this case, the parliament did its job, listened to the citizen, and activaly "patched" the proposed directive (a better way than simply refusing it, as it wouldn't have stopped the patents proponents to submit again and again, like we saw..) to largelly remove all harm.
The problem then is that the minister's council and the commission (both unelected bodies) just decided to "forget" the parliament modifications. Unbelievable. I'm telling you, the parliament wasn't pleased !
Hopefully they will stop this shit, if only because they are pissed by the council/commission behavior .. and we can help convaincing them by contacting them :)
But you should realize that all "local" politicians use the EU as an excuse -- "hey it's not our fault, it's Brussels !!" -- while in fact, THEY are still in power (with the commission and the council), and for the parliament, well, you voted for it.
The EU is an excellent scapegoat for the national politicians, and nobody seems to realize that. Like if the national politicians were just unconnected to the delegation in brussels. Yeah, sure.
The EU is NOT a fucked up idea, the current organisation is. Of course, the US are not exactly pleased by the increasing power of the EU, but hey...
By the way, software patents are GOOD. They DO protect the small developer. As a small developer who has a couple of software patents that I have successfully licensed, I can PERSONALLY vouch for them. Of course, the slashbots don't want to hear this. The current issue with patents isn't the fact that there are software patents, but maybe the fact that there are cases where they have been granted without a good reason. Saying that "patents are bad" is just silly.Bullshit. Even "normal" patents have bad side effects for the famous "little guy", and we're here talking about SOFTWARE patents, patents on ideas. That's the dumbest thing ever. I doubt that you're saying the truth with your "I can vouch for them" (ie, as an AC I think you're just pulling shit out of your ass), but even if that's the case, you should realize that your situation is the exception, not the norm, and by far. Software patents are used by big company to stiffle innovation. Ask bill gates, he wrote it black on white.
OpenStep isn't really a "frozen standard" any more, because the single widely-used implementation (OS X) continues to evolve.
The fact that Apple adds things to Cocoa (not that much by the way) doesn't suddenly transform OpenStep as a moving target. Cocoa is a moving target, not OpenStep. And guess what ? GNUstep already implements OpenStep (minus some rather unused classes).
The one big problem with GNUStep is that they don't care enough about replicating Apple's changes. : (
And how did you actually figured that ?
GNUstep primary goal was to implement OpenStep, yes. But Apple's changes are also added. Saying that "we don't care enough about replicating Apple's changes" is just plain wrong. If the changes are sound and, moreover, if somebody actually provides some code, I can guarantee you that it will be integrated. As always, it's a matter of coding. If you want to help, just start to code.. it's a free software project you know.
There is two free implementations of WebObjects -- one done by the GNUstep project, GNUstepWeb, and another that's done by OpenGroupware.org (they use it as their base for SOPE), called NGObjWeb.. both implements WebObjects 4.5 (ie, the version before java) and are used commercially.
Both works quite well, but they lack a scripting language (although the OGo guys works on a WebScript-like language, and the GNUstepWeb guys works on integrating StepTalk).
About EOF, the GNUstep project have an implementation named GDL2 (GNUstep Database 2) which implements EOF 4.5 -- but sadly we are missing an EOModeller (even if somebody is working on one). Here an article from Ludovic Marcotte about GDL2.
I almost want to sob into my keyboard.
:-)
;-) -- but I surely won't advocate taking one of the ugly themes here as a good UI example. I quite like the general idea of this one for example, but it's obvious it needs works. On the other hand, when you say that the OSX themes looks bad, take a look at that one -- that's really close to OSX (well, minus wmaker titlebar, of course). Anyway, there are things to correct, and they don't look perfect (like for example, the misaligned text). But mostly it was to test different features and modifications... so yes, it was because I could :-)
Sorry...
They are as ugly as the hind end of a dog with a hat on it. The first few minimal ones just make the clean GNUstep app look amateurish and childish with misaligned text in the buttons and garish coloured arrows to indicate the default button; the ones that attempt to imitate Aqua are actively ugly.
Well, none of them are intended to be full-featured themes. Mostly, they are just snapshot of different tests and modifications. If you want a theme engine, you *need* to test things
Don't forget it's just a theme engine -- GNUstep will still have the NeXT look by default. And you know what ? personally I tend to prefer the NeXT UI (that's partly why I didn't work as much as I should have done on Camaelon). But as you can read it on slashdot, people won't use GNUstep without a "modern" look (it's perhaps sad, but that's the reality). On a pragmatic way, having a theme engine will, regardless of slashdot users, be useful for cross platform (on Windows, etc).
Vim in Terminal.app is effectively a customization -- you can set your editor in ProjectCenter, and you can possibly use the integrated editor. But, the integrated editor is a bit poor in features, and above all I'm too used to vim :-)
The thing is, lots of gnustep dev like 1) the floating menus 2) the list selector widgets (personally that's one of my favorite widgets -- and by the way they still exist on Cocoa, and are used..) 3) the "empty" scrollbars is NOT a bad idea. Again, it's the case on Cocoa as well.
So your "not even MacOSX" comment is way off base. The only thing OSX does NOT is the floating menu, and a lot of people regret them.
Er, NO, you SHOULD NOT USE IT.
:-)
...
Partly because Camaelon needed to patch gnustep, and that's not the case since a long time. So don't try to apply the patches
Camaelon 2 will provides a full pixmap theme engine, look up in http://www.roard.com/screenshots/ for latest themes sshot
should be released before the end of the month.
Sure, Objective-C is the root of all evil and slow a system to death. That's why, as we have all seen with the recent (!) steve job video, a workstation with a 68030 (how much slower than our current hardware ?) was undoubtely unbearably slow...
Well, HURD was supposed to be the Kernel of the GNU OS, and GNUstep in the beginning was supposed to be the GUI of the GNU OS. As recently Hurd progressed with the first results of the L4 port, it was a reference to that..
Firstly, please, please update the look-and-feel
Like I said in a previous comment, I'm working on Camaelon 2, a pixmap theme engine that lets you have pretty things.. it should be officially released before the end of this month (it already works, I just want to clean up things).
I didn't see any support for layout management in Gorm
Well, I didn't show that part, but that works exactly in Gorm like on InterfaceBuilder on OSX (and imho it's a better model most of the time than the springs). So of course you can have resizable widgets. In addition, GNUstep implements a couple of widgets implemeting the spring resizing model (that's used by Renaissance by the way, an XML framework for describing UI for GNUstep...), so if you *really* want the spring model, you can use it.
The thing is, the NeXT machine used to have gamma corrected display. And so, in real use, it wasn't as dark as it can appear on screenshots. Alas, GNUstep gui, by default, use the same colors value -- and so, if your display is not correctly calibrated, it will appear too gray, too dark.
You can change the colors with the Preferences.app though, if you're not fond of the default gray !
The other possibility is to wait for Camaelon 2, a pixmap theme engine I'm working on. It works at the moment, and if you really want it now you can drop me a mail, but I'd like to clean up the code and add a few things. I intend to release it before the Fosdem (so, before the end of this month). Some screenshots like this one or that one show what can be done.
A lot of "C"-isms showing up in the process of defining "look" and "behaviour".
Hm, can you explain what you mean ? I don't see what kind of "C"-isms are shown in it..
Something like this ? or that ?
tell me where NS is "pathetically primitive" ?
All I can say is, what the heck happened
Well, basically, NeXT overcharged their hardware, then their software. For example, you probably never heard of WebObjects, even if it was (still is actually) one of the best technology to create a dynamic website... and it's no wonder considering they used to sell it at insanely huge prices. Now you can have it via Apple for 500$ ...
this is big enough stuff that they should have been able to get through a few lean years and sell the technology....
Well, they did :-) -- to Apple ...
Actually, the problem they had, is that nearly nobody in the industry was used to OOP. Now it's easier to understand the brilliance of NeXTSTEP's concepts, but it was probably more difficult to convaince people at the time ? (check the real media video on openstep.se/next/videos , they take half the video to explain the interest of OOP before introducing IB..)
And of course, a NeXT Cube and even a NeXT station were extremely expensive... too bad, they were 15 years ahead of their time (yeah, OSX is not as clean as NeXTSTEP, partly because of the need to integrate all theses existing apps..)
For thoses who want to see how programming is done in GNUstep, there's this short flash demo here
GNUstep is a free software implementation of the OpenStep API (like Cocoa), and it provides development tools as well. The demo steve do is doable in GNUstep as well..
(Yes, it's flash... a mpeg version will probably be available next week... in the meantime, it's a good idea to check either swift tools or swfdec , if you don't want or can't use the Macromedia Flash player..)
by the way, for thoses who will go to the Fosdem next february, there is a GNUstep developer's room, and there will be presentations, discussions, etc.
Actually, I could see two developments of the mac mini as a server that would sell quite well:
- first and most obviously, a mac mini "home digital center" -- with 5.1 output, tv tuner, tv out.
- second, a cheap brick for clustering, with a gigaethernet card (that doesn't automatically conflic with the "home digital center" idea..).
Mind you, this is the most exciting idea imho -- not because it could be used for some "traditional" clustering (even if I can imagine some people using it that way), but because it could be used to redefine the whole concept of Personal Computer and Home Computing..
Instead of having "one" computer, you would have a bunch of mac mini servers, cooperating. For geeks it could be, one mini for the database, one mini for the web server, etc. But for lambda mac user, it could be a different separation : one mini for the photos, one mini for the movies, etc.
The point would be that OSX would abstract the cluster, so people would have nothing complex to do. Just imagine -- "Gosh imovie is slow... hey I go buy a new brick, plug it on my network, and it will be faster!".
That could not be _that_ hard to implement for Apple -- don't forget that they control everything (hardware and software), so for them it's easier to push for this kind of innovation. And even if just iLife and system apps fully use the cluster, it could be useful (and 3rd party developers would anyway use it too after a bit).
Don't forget that OSX already have RendezVous, simplifying the building of this kind of "home cluster", and XGrid ... so what I'm describing could be seen as a mere improvement of what already exist ...
On a side note, I wonder if it's possible to easily "stack" the mac mini -- could be nice to have external peripherics in the same kind of case as the mini, and you would build you tower by stacking your equipment (PVR box, Firewire drive, music I/O, etc..). Talk about evolution ;-) [it's probably not doable with the current mini, but it would be great to have that for the next one ;-)]