No matter what, you'll always have a visual representation. Be it saved searches like GMail, or something else more like the Finder, you will always be able to visually navigate the data.
That's important. If you can't visually navigate it, then it's far too easy to lose stuff. It's just that the bulk of your organization is going to be done by a search engine. What's nice about that is that you can retroactively organize things. Ever had a pile of downloads and wish that you had organized them more? Well, now you can!
It can also be a tool for organization, not just the end of organization itself. Extending the cluttered dowload folder above, the first thing you could do is break the downloads up into groups ordered by date, broken by weeks. You could also search for things that have never been looked at (creation date is the same as modify date).
It also means that multiple people can share the exact same filesystem, but look at it many ways. Your children may only care about the games, email and webbrowser. You probaby care about these things, but you also care about your work.
It takes some abstract thought, since no one has a system that really makes it perfect yet, but Spotlight is a huge step in the right direction, and when we get there and polish ip up, it'll be a boon for everyone, from Grandma to Larry the Bitter IT guy.
The folder structure is very specific with a folder for the job, a subfolder for ad graphics, another subfolder for the layout and another subfolder for the maps. The jobs can be anywhere between 10 megs and a gig depending on the job. Dozens to hundreds of pictures (bmp, eps, tiff, etc), Indesign or Pagemaker layouts, Illustrator maps, etc. It would be chaos if an advertiser logo was out of place because that would crash the entire job when it went to be plated.
But, that's because you work around folders. It's not because folders are inherently superior or necessary to do this kind of work.
Imagine if every project had a tag, like, "USPS Job". Then, some files would have type tags like, "com.adobe.illustrator". You wouldn't visualize it as a flat space, of course. It might still look like the finder, sorta.
But it's tough to imagine, because spotlight doesn't yet have enough muscle and backup to pull it off. We're still few years out on that. Apple is positioning themselves for that, it seems, but they know they're not ready. Right now, Spotlight just lets you make new views of your data, in an ad hoc and semi-permanent fashion.
Awhile ago, I bought a Fingerworks Keyboard. These things use a heat-sensing technology to allow the same surface to detect gestures, button presses, and mousing without any "pushing" required. Contact is all it takes.
It's pretty slick, and it really helps me when I'm doing somethign that requires alot of transitioning from mouse to keyboard. It also adds gesturing to any application, which is pretty damn slick. Gestures can be even faster than keyboard input.
OS X throws away most of the Apple HIG, and is a delightful mish-mash of styles. It's awful. It makes a delightful demo, what with the Lickability Factor up around 11. But after having used it for a year (and, mind you, I've been a raving fanboy myself, first of System 7, then of OS/2) I have to say "no thanks". OS X tries to play close to the Mac ideal, but fails miserably.
The perfect OS doesn't exist yet. OS X tries very hard, and really succeeds quite well. There is no question that OS X represents a variety of control styles and organizational styles. It supports something like the spatial finder, with a browse view. With tiger, it supports something like "Your Computer is a Database" organizational memes (which some people like, and some people hate).
But many of these are practicalities. While a Interaction Designer may balk, they can still appreciate that OS X is closer to a usable and consistent system than nearly any other offering that is commercially viable.
We have to judge things according to their competition. Does OS X have usability problems that need attention? Of course! Do they do a better job than pretty much everyone else? Open for debate, but the general feeling is that OSX tries harder to do it right.
Time and time again, history has shown that incremental improvement is better than big-bang improvement. This effect is magnified many times in user interface, where beauty and utility must be balanced carefully and everyone's opinions taken into account.
I do like the.app application bundles (reminiscent of applications on Acorn's RISC OS!) but I wouldn't say everything was lovely - such drag-and-drop bundles don't work so well when something needs to install frameworks, drivers or whatever.
Well, the.app stuff is nice because it's trivial to include your required frameworks. There is a spot just for them. Never worry about versioning again, at the cost of your application size, of course.
What's even more impressive to me, as a developer, is the.framework scheme. Similar to the.app, the.framework allows libraries to consolidate their headers and binaries in one place. Key to this is that it also allows multiple versions to be held in one place. When upgrading a framework, the system should just update the framework internals without replacing your old version. It's extremely nice, and I really wish Linux would adopt it.
It's not as if applications don't use that system for installation, either - I recently installed iLife '05 (yeah, I'm slow) and it appears to have spread files out all over the place. Not as bad as your average Linux application admittedly, but I'm not sure how one would go about tracking them all down. (Okay, so I'm not likely to remove it, but I'm using it as an example, okay?)
The standard Installer.app keeps receipts of every.pkg it runs through/Library/Receipts. You can use that to remove them.
Almost all of your information is seriously outdated, or outright wrong. It's our job to stay current on stuff like this. Please consider that.
That is a bit like saying WINE has made massive strides in providing a Windows layer. It has made massive progress but it is barely close to functional except for a specific set of applications. Maybe that's a bad example, but what I'm saying is that dispite all the progress the problem persists (relative to other solutions BSD-on-Mach is slow).
"Barely functional?" I think you're overstating your case. I have successfully deployed XServes running OS X Server. They handle the apache load admirably. I have had some problems with MySQL (which the recent Anandtech article mentions), but this is because MySQL does more on OSX than it does on Linux to ensure transactions complete.
Do you have any real examples of this "Barely functional" behavior? Mach is probably slightly slower. I don't think it really makes a huge difference unless you're doing realtime programming (it sure hasn't made a difference for me).
No, [the Mach-O ABI speed problem] was still there.
No. It was not there on 68k. The Mach-O ABI assumes the programmatic availability of a program counter register. The PowerPC RISC architecture did not provide one, and was not written to leverage one. There was a logical and real disconnect in these ideologies which led to a noticable speed penalty on PowerPC machines.
GCC4.0 has a new option to help eliminate this flaw. But the flaw never existed on 68k.
Speaking as a former NeXTSTEP user, OS X performs nearly identical (if you take into account hardware advances and the monster that is Aqua). Except a few areas where there have been improvements (like pre-linking) it still has many of the old warts. A lot of the high performance neato things done back in the day where due to the NeXT hardware (because PC hardware completely sucked at the time). When NeXTSTEP finally went Intel it was a lot slower than Windows.
That's because at the time, its competition was vastly simpler and didn't offer nearly as much in terms of performance. Its competition was Windows 3.1.
Also, the state of x86 at the time was simply not caught up with its competitiors. The processors were just slower, and when you asked them to do more, they showed it. Blaming the OS because of the hardware is both unfair and incorrect. It ran great before the x86 switch.
NeXTSTEP had a great development environment. All object oriented and stuff. Objective-C is a cool language. However, there are serious penalties to be paid for that type of design. All method calls in Objective-C are virtual functions requiring a pointer dereference (possibly a hash lookup, I can't remember) and there is very limited opportunity to inline code.
Outdated. Not only does Apple have a lock-free implementation of their ObjC runtime, they also can aggressively leverage the static typing information.
Oh, and if you're really calling a function so often that you begin to see the dispatch latency, you just lookup and retain the C function pointer reference. But at that point, why were you using a non-inlined function in the first place?
The days where a virtual function pointer dereference was slow are gone, my friend, so even when it can't, there is no problem.
In nearly every college compiler course I've seen (and the one I took), they showed us how to make virtual functions very fast. Yes, there is overhead. No, it is not significant.
Indeed, the only time you can get anything even remotely resembling a performance hit with ObjC is when you use pure dynamic dispatch, using dynamic selectors. In that case, the system has to use extra work.
You can't do too much stuff at runtime without penalty (Smalltalk et al). I know someone will argue about computers b
There's no mention of what Apple will be using to boot their computers, just that as a programmer, you should not make assumptions. Apple has been preaching that for DECADES.
I didn't say it was inherent in the design. However, if it is so easily fixed then why is it still a problem? (This problem is not new, like I said, it has been around for nearly 20 years)
That's incorrect. XNU has made massive strides in performance. And the Mach-O problem? Didn't exist on 68k machines or Intel machines (from the NeXT days).
There is still work to do, but that doesn't mean you can invalidate the progress made.
Whatever... I have to work with what my customers are using. I can't just pick a platform and go with it.
Professionally, I'm a Linux developer. When I do GUI apps, I do them in Qt. But Apple acceptance is growing. I'm working a consulting job involvng lots of Apple dev (I may even have to get one of those transition kits).
I've mentioned the food-on-the-table issue before, and it's frustrating. But, let's be fair here, continued education is part of our responsibility as software developers. Even with this arch switch, Apple is in a very strong position and people are starting to realize the benefits.
If Mac prices drop 10% because of this switch, you and I may be getting a lot more mac platform jobs.
OS X is slow, bloated, and somewhat insecure. The slow and bloated parts are just a problem with the design. BSD on Mach is wasteful and they do way too much object-oriented stuff that is inefficient (not that OO is bad, just their design which has Smalltalk-like issues).
Err, you say you're a developer, but then you say things like this. The Mach/BSD issue is not the bottleneck. That Anandtech article was painfully innacurate and uninformed.
There are some bottlenecks in this region, but they are not inherent to the design. For example, the Mach-O ABI has a weakness on RISC machines, and the kernel resource locking needs to be more finely articulated.
OS X has a lightweight OO architecture for device drivers, but this is in C++, so it hardly matters once the code is compled. Mach itself is OO, but implements a very fast message passing algorithm. It is not the source of any performance woes, and opens up many possibilities in distribued computing.
And certainly, these issues are different from any Smalltalk VM performance issues I've seen. OS X isn't suffering because they refuse to let the OO metaphor go. It's suffering because as an OS its still growing. Linux had its fair share of problems and performance woes back in the 2.2 days. They were corrected fairly quickly.
XNU is showing a similar trend.
This goes way back the design of NextStep which had similar problems. As for the insecurity, it's the same problem I have with Windows. I don't have the source code to most of the system and there are is lot of legacy and convenience stuff in there that will eventually lead to insecurities just like on Windows (just wait and see when OS X is more pervasive).
Again, gah? You don't have the sourcecode to the windowing system and some of the applications. You have the code to all the services, the core of the OS and company.
There are a few holes here, and it'd be nice to see them filled, but they're not really in typical problem areas. The vast majority of problems exist in services or in the security architecture of the system. I can understand if you're upset that we don't have the code to the Keychain, that's something Apple needs to open so we can have some confidence about it.
But it's nowhere near as bad as Windows. Or many other commercial OSs, for that matter. To compare them this way does a major disservice to the app.
don't need an Aqua-like eye-candy system to do development on. I can chose to run GNOME, KDE, or something lightweight. I like that control because it keeps my system performance up in the places I need it (eg. I need to run VMware fast, I need to compile fast, etc.).
For me, one of the major attractions of OS X is how damn good Cocoa and its dev platform is. GNOME and KDE suck by comparison, in nearly every way you compare. At unlike KDE, at least Apple is honest about being proprietary.
This doesn't mean they will run a standard BIOS. Surely they will not. But it looks an awful lot like they want their solution to be an Intel showboat.
Also, given the fact that we have Apple on record saying that they will do nothing to stop people from running Windows on their new macs, I think that they're going to stake their Different-ness more on the speed and quality of their engineering.
8. Cost. If you expect an Apple box to cost significantly less with a different processor, you're smoking crack.
9. Performance. Anyone who wants serious power will still go with Linux, especially since Apple is inexplicably going from a 64-bit processor with a 128-bit memory bus to a 32-bit clunky piece of junk.
I have a funny feeling that one of the reasons that Intel was so eager to pull Apple into the fold was that they want a showboat for Intel research.
As much as AMD is favored around here, no one can deny that Intel has a lot of hot engineers with some damn good ideas. They have a new answer to the BIOS, they're beating the multi-core drum pretty hard, and they've got a lot of good architectural ideas.
Intel can make an entire platform for Apple, and it'll probably be a year or two ahead of the stock Dell line.
I can't help but get the feeling that Apple's decided to draw a line in the proverbial sand. The only way they can survive in the x86 world is by being very fast with their engineering and have a long lead on hardware.
While I grant that the mac market share is small, please don't hold off from learning Cocoa just because of that. Much like learning any computer language, Objective-C is full of interesting concepts that will broaden your mind as a programmer.
While it won't put food on the table, Cocoa is a beautiful framework in many ways. The language is a natural fit for GUI development. Technically speaking, Cocoa is vastly superior to nearly all its competition. GTK+ and Qt (ye gods, especially Qt) are just blown out of the water.
Once you learn the framework, you might find that keeping a mac version of your application out there is much easier than you thought. If you struture your designs right (and by that I mean a good Model-View-Controller division) you can do a lot of work without writing any code at all.
Apple already consolidated all this into their universally distributed Accelerate.framework, which provides a lot of fancy vector ops, a hand-tuned LAPACK and CBLAS library, and a high speed image processing kit.
When I was first going over them, I was confused why it was all written to be so platform-neutral, but now we know why. I'm sure that all their specially tuned code (like in Mail.app for search) runs just fine on Intel. Probably slower, SSE is not as fast as the AltiVec, but I doubt it required much of a rewrite.
Have you seen what happens when people try to bring Objective-C features into C++? TrollTech tried with Qt.
You get a complex meta-language layered over the top of Qt that involves a lot of complex memory semantics, another special compilation phaze that's obnoxious to deal with, special build tools that lack flexibility, and odd syntax that editors don't recognize.
It's a nightmare. Objective-C is a much better language all around for GUI programming. C++ has its place, and that's why ObjC and C++ can talk and play nice. But pure static typing (inferred or lexical) in Applications is going the way of the dodo, get on the bus now or be left behind.
I have no ability to find your old comments. This particular comment is linked by the journal entry under your name.
I like how you deny reading the old comment I made (which you clearly did, since you referenced it), but not the accusation that you're rooting around for ad hominem material. Cute.
The remainder of your post is not only offtopic, but it's outdated.
I'm am sorry my post was not kid tested and Minna-Kirai approved. Most other folks seemed to get my point, and I did get over 5 mod points, so I'm pretty sure I'm in the green. Don't you wish you had a -1 Not Kid Tested and Minna-Kirai Approved?
I still think you're being overly pedantic because you don't want to lose an argument you started on/., but it's really not worth my time (and I hope it's not worth yours) to continue.
No, that is false. It's completely false, because a "tracker" is not a human-viewable display. No person can watch a tracker looking for new files.
On IRC, I hear people called these sites "trackers". I'm told it's short for "torrent tracker trackers". They're release trackers. It's slightly confusing, but I didn't make up the term, nor am I the only one to use it.
Even if we correct your statement to be "tracker-associated websites", it's still wrong, as you presume knowledge about habits of other people that you don't actually possess. It is the factuality of your assertion that Animesuki et al are the most used that is being disputed. They are not the most popular, or else they would be higher up in Google results.
"anime bittorrent" and "anime torrent" show both of them in the top 10 for me, Anime suki is 3rd and 2nd respectively. Seems pretty damn popular and well linked. Scarywater is a little lower, but also in those links.
Do you have a special google that I am not privledged to? If so, share.
Those sites act as initial release points for fansubs. Once the series is licensed, they may take it down, but it stays up on all the slightly-more shady sites, such as those full of The Sopranos and "24" episodes.
I wasn't talking about those slightly shady sites. I was talking about sites that focus on anime, and I was talking about the typical ones that a cusory google examination would find.
Want to conduct another test? Go try to download a series which already has a major USA release, like GITS:SAC. Easy, huh? Because licensed anime is typical
When did I say that people didn't distribute, fansub, or watch licensed anime? Please point this out to me, because I was under the impression I was talking about said websites.
Crazy me, I guess.
PS. The following line is remarkably ignorant of Apple's policy regarding GUI consistency:
Apple cannot make a consistant and reasonable player app in Linux. If they choose toolkits and run with it, they get nailed. If they make their own toolkits, they get nailed.
Okay. First, are you reading my previous comments trying to find ad hominem material for your argument here? That's pathetic.
Second, I stand by that statement, and I've got a few years of OSX-based consulting to lend credence to my opinion about macs, for what it's worth.
Exhibiting the qualities, traits, or characteristics that identify a kind, class, group, or category: a typical suburban community.
Of or relating to a representative specimen; characteristic or distinctive.
Conforming to a type: a composition typical of the baroque period.
Please see definition one.
There are numerous online dictionaries to help you with those SAT words you don't know.
Most anime-oriented tracker sites do not carry series like Gundam Seed Destiny and Naruto. Some of the biggest and most popular tracker sites for anime follow this policy.
Well, this doesn't seem to me to be a great justification. You've been able to use a wide variety of peripherals with your consoles for years now. Even the gamecube supported mice and keyboards.
Heck, the PS2 has successfully carried FFXI for quite awhile now, and PS2 users are in no way the minority on that system.
Now that modern consoles are just specialty PCs and have USB connectiors and enough RAM such that games can use USB HID drivers without sacrificing graphics, there isn't much argument.
I don't think PC gaming will ever dissapear, of course. But certainly, between the PSP and the next gen consoles, it's increasingly difficult to justify selling games on a PC. Especially once the XBox developers no longer get a sweet and short ride porting between Xbox and PC, you're going to see a precipitous drop in PC games.
And why would anyone buy $1k-$2k gaming machine when their speciality $400 system can push more polys and has nearly the same rendering abilities, uses any peripheral you want, makes online gaming a breeze, isn't going to get you an email virus, and has a wider selection of games?
Just because it's the next DragonBall doesn't mean it's bad.
Heck, for its time, Dragonball wasn't that bad. Dragonball only became bad once it was forced to continue on past the the series end that the creator envisioned (after the Frieza saga).
Naruto is still appearing in Jump and the original creator still has a story to tell. It is not the Eternal Series just yet. It's just long.
Typically? Bollocks. No one seems to care whether a series has been licensed or not.
Trackers like AnimeSuki and Scarywater care, and they're the way most people watch for anime. When a series goes off these trackers, they're harder to find.
Note, I did not say hard to find. I said harder. If someone wants to download Naruto, there are several groups that sub it.
Python is the new C++. It's easier than its predecessors, is not too adventurous or different, and has a lot of power and features. However, for web work it may be slightly out of its element. Yes, Zope and PEAK are cool tools, but they fill a niche that very few web applications fit well into.
If you need a small, fast, or medium sized webapp, no one challenges Ruby On Rails right now. Where that framework fails, Java based solutions start to shine. Subway might be able to get Python into that arena in the next few years, but not this year (and probably not next year either).
I'm a fan of Python, but it's not a terrific choice for web development these days in the face of the competition.
Typically, anime is only distributed via torrent when there is no american company planning to sell it. This policy is meant to help smooth frictions between american publishers and file sharers. It's hard to argue that money is lost when americans download episodes of an anime that may never even be shown anywhere but Japan, and if no money is lost then a lawsuit is rather pointless.
The day that Naruto got licensed for US distribution, the fanbase seemed to go completely crazy. No one wanted to stop watching. Several groups decided to take their effort "underground" (by which I mean not listed on popular anime tracker sites, only from IRC and obscure group webpages).
If anything, bittorrent is good for series like Naruto. Distribution companies get a free, zero-effort focus group for nearly every anime that comes out. By watching anime tracker stats, it's easy to see which series are a crazy success and which are bombs. This is also much more reliable than watching screening attendance at conventions (which tend to vary wildly by time and location).
It just goes to show that just because you can excercise your copyrights, it doesn't always mean you should. I seriously doubt an anime like Gantz (or even Midori No Hibi, although I think people would argue with me about that) would have ever seen american distribution without a lot of fan support from subbers and the thousands of people who download unreleased anime.
Slashdot Mirror
No matter what, you'll always have a visual representation. Be it saved searches like GMail, or something else more like the Finder, you will always be able to visually navigate the data.
That's important. If you can't visually navigate it, then it's far too easy to lose stuff. It's just that the bulk of your organization is going to be done by a search engine. What's nice about that is that you can retroactively organize things. Ever had a pile of downloads and wish that you had organized them more? Well, now you can!
It can also be a tool for organization, not just the end of organization itself. Extending the cluttered dowload folder above, the first thing you could do is break the downloads up into groups ordered by date, broken by weeks. You could also search for things that have never been looked at (creation date is the same as modify date).
It also means that multiple people can share the exact same filesystem, but look at it many ways. Your children may only care about the games, email and webbrowser. You probaby care about these things, but you also care about your work.
It takes some abstract thought, since no one has a system that really makes it perfect yet, but Spotlight is a huge step in the right direction, and when we get there and polish ip up, it'll be a boon for everyone, from Grandma to Larry the Bitter IT guy.
Imagine if every project had a tag, like, "USPS Job". Then, some files would have type tags like, "com.adobe.illustrator". You wouldn't visualize it as a flat space, of course. It might still look like the finder, sorta.
But it's tough to imagine, because spotlight doesn't yet have enough muscle and backup to pull it off. We're still few years out on that. Apple is positioning themselves for that, it seems, but they know they're not ready. Right now, Spotlight just lets you make new views of your data, in an ad hoc and semi-permanent fashion.
Awhile ago, I bought a Fingerworks Keyboard. These things use a heat-sensing technology to allow the same surface to detect gestures, button presses, and mousing without any "pushing" required. Contact is all it takes.
It's pretty slick, and it really helps me when I'm doing somethign that requires alot of transitioning from mouse to keyboard. It also adds gesturing to any application, which is pretty damn slick. Gestures can be even faster than keyboard input.
But many of these are practicalities. While a Interaction Designer may balk, they can still appreciate that OS X is closer to a usable and consistent system than nearly any other offering that is commercially viable.
We have to judge things according to their competition. Does OS X have usability problems that need attention? Of course! Do they do a better job than pretty much everyone else? Open for debate, but the general feeling is that OSX tries harder to do it right.
Time and time again, history has shown that incremental improvement is better than big-bang improvement. This effect is magnified many times in user interface, where beauty and utility must be balanced carefully and everyone's opinions taken into account.
What's even more impressive to me, as a developer, is the .framework scheme. Similar to the .app, the .framework allows libraries to consolidate their headers and binaries in one place. Key to this is that it also allows multiple versions to be held in one place. When upgrading a framework, the system should just update the framework internals without replacing your old version. It's extremely nice, and I really wish Linux would adopt it.
The standard Installer.app keeps receipts of every"Barely functional?" I think you're overstating your case. I have successfully deployed XServes running OS X Server. They handle the apache load admirably. I have had some problems with MySQL (which the recent Anandtech article mentions), but this is because MySQL does more on OSX than it does on Linux to ensure transactions complete.
Do you have any real examples of this "Barely functional" behavior? Mach is probably slightly slower. I don't think it really makes a huge difference unless you're doing realtime programming (it sure hasn't made a difference for me).
No. It was not there on 68k. The Mach-O ABI assumes the programmatic availability of a program counter register. The PowerPC RISC architecture did not provide one, and was not written to leverage one. There was a logical and real disconnect in these ideologies which led to a noticable speed penalty on PowerPC machines.
GCC4.0 has a new option to help eliminate this flaw. But the flaw never existed on 68k.
That's because at the time, its competition was vastly simpler and didn't offer nearly as much in terms of performance. Its competition was Windows 3.1.
Also, the state of x86 at the time was simply not caught up with its competitiors. The processors were just slower, and when you asked them to do more, they showed it. Blaming the OS because of the hardware is both unfair and incorrect. It ran great before the x86 switch.
Outdated. Not only does Apple have a lock-free implementation of their ObjC runtime, they also can aggressively leverage the static typing information.
Oh, and if you're really calling a function so often that you begin to see the dispatch latency, you just lookup and retain the C function pointer reference. But at that point, why were you using a non-inlined function in the first place?
The days where a virtual function pointer dereference was slow are gone, my friend, so even when it can't, there is no problem.
In nearly every college compiler course I've seen (and the one I took), they showed us how to make virtual functions very fast. Yes, there is overhead. No, it is not significant.
Indeed, the only time you can get anything even remotely resembling a performance hit with ObjC is when you use pure dynamic dispatch, using dynamic selectors. In that case, the system has to use extra work.
I've mentioned the food-on-the-table issue before, and it's frustrating. But, let's be fair here, continued education is part of our responsibility as software developers. Even with this arch switch, Apple is in a very strong position and people are starting to realize the benefits.
If Mac prices drop 10% because of this switch, you and I may be getting a lot more mac platform jobs.
There are some bottlenecks in this region, but they are not inherent to the design. For example, the Mach-O ABI has a weakness on RISC machines, and the kernel resource locking needs to be more finely articulated.
OS X has a lightweight OO architecture for device drivers, but this is in C++, so it hardly matters once the code is compled. Mach itself is OO, but implements a very fast message passing algorithm. It is not the source of any performance woes, and opens up many possibilities in distribued computing.
And certainly, these issues are different from any Smalltalk VM performance issues I've seen. OS X isn't suffering because they refuse to let the OO metaphor go. It's suffering because as an OS its still growing. Linux had its fair share of problems and performance woes back in the 2.2 days. They were corrected fairly quickly.
XNU is showing a similar trend.
Again, gah? You don't have the sourcecode to the windowing system and some of the applications. You have the code to all the services, the core of the OS and company.There are a few holes here, and it'd be nice to see them filled, but they're not really in typical problem areas. The vast majority of problems exist in services or in the security architecture of the system. I can understand if you're upset that we don't have the code to the Keychain, that's something Apple needs to open so we can have some confidence about it.
But it's nowhere near as bad as Windows. Or many other commercial OSs, for that matter. To compare them this way does a major disservice to the app.
For me, one of the major attractions of OS X is how damn good Cocoa and its dev platform is. GNOME and KDE suck by comparison, in nearly every way you compare. At unlike KDE, at least Apple is honest about being proprietary.According to Apple's Universal Binary Programming Guidelines, their machines will not run OpenFirmware.
This doesn't mean they will run a standard BIOS. Surely they will not. But it looks an awful lot like they want their solution to be an Intel showboat.
Also, given the fact that we have Apple on record saying that they will do nothing to stop people from running Windows on their new macs, I think that they're going to stake their Different-ness more on the speed and quality of their engineering.
As much as AMD is favored around here, no one can deny that Intel has a lot of hot engineers with some damn good ideas. They have a new answer to the BIOS, they're beating the multi-core drum pretty hard, and they've got a lot of good architectural ideas.
Intel can make an entire platform for Apple, and it'll probably be a year or two ahead of the stock Dell line.
I can't help but get the feeling that Apple's decided to draw a line in the proverbial sand. The only way they can survive in the x86 world is by being very fast with their engineering and have a long lead on hardware.
While I grant that the mac market share is small, please don't hold off from learning Cocoa just because of that. Much like learning any computer language, Objective-C is full of interesting concepts that will broaden your mind as a programmer.
While it won't put food on the table, Cocoa is a beautiful framework in many ways. The language is a natural fit for GUI development. Technically speaking, Cocoa is vastly superior to nearly all its competition. GTK+ and Qt (ye gods, especially Qt) are just blown out of the water.
Once you learn the framework, you might find that keeping a mac version of your application out there is much easier than you thought. If you struture your designs right (and by that I mean a good Model-View-Controller division) you can do a lot of work without writing any code at all.
Don't be so quick to dismiss ObjC/Cocoa.
Apple already consolidated all this into their universally distributed Accelerate.framework, which provides a lot of fancy vector ops, a hand-tuned LAPACK and CBLAS library, and a high speed image processing kit.
When I was first going over them, I was confused why it was all written to be so platform-neutral, but now we know why. I'm sure that all their specially tuned code (like in Mail.app for search) runs just fine on Intel. Probably slower, SSE is not as fast as the AltiVec, but I doubt it required much of a rewrite.
Have you seen what happens when people try to bring Objective-C features into C++? TrollTech tried with Qt.
You get a complex meta-language layered over the top of Qt that involves a lot of complex memory semantics, another special compilation phaze that's obnoxious to deal with, special build tools that lack flexibility, and odd syntax that editors don't recognize.
It's a nightmare. Objective-C is a much better language all around for GUI programming. C++ has its place, and that's why ObjC and C++ can talk and play nice. But pure static typing (inferred or lexical) in Applications is going the way of the dodo, get on the bus now or be left behind.
The remainder of your post is not only offtopic, but it's outdated.
You really are dedicated, aren't you?
/., but it's really not worth my time (and I hope it's not worth yours) to continue.
I'm am sorry my post was not kid tested and Minna-Kirai approved. Most other folks seemed to get my point, and I did get over 5 mod points, so I'm pretty sure I'm in the green. Don't you wish you had a -1 Not Kid Tested and Minna-Kirai Approved?
I still think you're being overly pedantic because you don't want to lose an argument you started on
Do you have a special google that I am not privledged to? If so, share.
I wasn't talking about those slightly shady sites. I was talking about sites that focus on anime, and I was talking about the typical ones that a cusory google examination would find. When did I say that people didn't distribute, fansub, or watch licensed anime? Please point this out to me, because I was under the impression I was talking about said websites.Crazy me, I guess.
Okay. First, are you reading my previous comments trying to find ad hominem material for your argument here? That's pathetic.Second, I stand by that statement, and I've got a few years of OSX-based consulting to lend credence to my opinion about macs, for what it's worth.
Step two, examine the top ten results. Note that scarywater and animesuki feature prominently amongst these results.
Step three, count how many of them allow Naruto to be listed.
Step three-point-five, count how many explcitly say in their policy that they do not track Naruto torrents.
Step four, fuck off.
Thank you, come again.
- Exhibiting the qualities, traits, or characteristics that identify a kind, class, group, or category: a typical suburban community.
- Of or relating to a representative specimen; characteristic or distinctive.
- Conforming to a type: a composition typical of the baroque period.
Please see definition one.There are numerous online dictionaries to help you with those SAT words you don't know.
Most anime-oriented tracker sites do not carry series like Gundam Seed Destiny and Naruto. Some of the biggest and most popular tracker sites for anime follow this policy.
Got anymore nits to pick, in true slashdot style?
Well, this doesn't seem to me to be a great justification. You've been able to use a wide variety of peripherals with your consoles for years now. Even the gamecube supported mice and keyboards.
Heck, the PS2 has successfully carried FFXI for quite awhile now, and PS2 users are in no way the minority on that system.
Now that modern consoles are just specialty PCs and have USB connectiors and enough RAM such that games can use USB HID drivers without sacrificing graphics, there isn't much argument.
I don't think PC gaming will ever dissapear, of course. But certainly, between the PSP and the next gen consoles, it's increasingly difficult to justify selling games on a PC. Especially once the XBox developers no longer get a sweet and short ride porting between Xbox and PC, you're going to see a precipitous drop in PC games.
And why would anyone buy $1k-$2k gaming machine when their speciality $400 system can push more polys and has nearly the same rendering abilities, uses any peripheral you want, makes online gaming a breeze, isn't going to get you an email virus, and has a wider selection of games?
What economic sense does that make?
The apple Bluetooth keyboard is really a beautiful thing. It's all the nice slim lines of the keyboard but no cord, and a nice key response rate.
That and a normal-sized bluetooth Kensigton mouse really help reduce desk clutter.
Just because it's the next DragonBall doesn't mean it's bad.
Heck, for its time, Dragonball wasn't that bad. Dragonball only became bad once it was forced to continue on past the the series end that the creator envisioned (after the Frieza saga).
Naruto is still appearing in Jump and the original creator still has a story to tell. It is not the Eternal Series just yet. It's just long.
Note, I did not say hard to find. I said harder. If someone wants to download Naruto, there are several groups that sub it.
Python is the new C++. It's easier than its predecessors, is not too adventurous or different, and has a lot of power and features. However, for web work it may be slightly out of its element. Yes, Zope and PEAK are cool tools, but they fill a niche that very few web applications fit well into.
If you need a small, fast, or medium sized webapp, no one challenges Ruby On Rails right now. Where that framework fails, Java based solutions start to shine. Subway might be able to get Python into that arena in the next few years, but not this year (and probably not next year either).
I'm a fan of Python, but it's not a terrific choice for web development these days in the face of the competition.
Typically, anime is only distributed via torrent when there is no american company planning to sell it. This policy is meant to help smooth frictions between american publishers and file sharers. It's hard to argue that money is lost when americans download episodes of an anime that may never even be shown anywhere but Japan, and if no money is lost then a lawsuit is rather pointless.
The day that Naruto got licensed for US distribution, the fanbase seemed to go completely crazy. No one wanted to stop watching. Several groups decided to take their effort "underground" (by which I mean not listed on popular anime tracker sites, only from IRC and obscure group webpages).
If anything, bittorrent is good for series like Naruto. Distribution companies get a free, zero-effort focus group for nearly every anime that comes out. By watching anime tracker stats, it's easy to see which series are a crazy success and which are bombs. This is also much more reliable than watching screening attendance at conventions (which tend to vary wildly by time and location).
It just goes to show that just because you can excercise your copyrights, it doesn't always mean you should. I seriously doubt an anime like Gantz (or even Midori No Hibi, although I think people would argue with me about that) would have ever seen american distribution without a lot of fan support from subbers and the thousands of people who download unreleased anime.