IBM had trouble going from 180 nm fabrication to 90 nm fabrication. We announced 3 GHz by 2004, but internally we were predicting 5 GHz by this summer, based on IBM's projections.
They didn't pan out.
For the record, Steve Jobs copped to this publicly during one of his keynote addresses. WWDC 2004, I think it was. Funnily enough, though, it seems like Slashdotters like to remember the initial projection, not the retraction, which is strange because they happened in exactly the same forum attended by exactly the same group of people. It almost seems like... Slashdotters are less interested in the truth than they are in making a big stink! But surely that's not possible...
My big problem with iPhoto is the way it insists on moving all the JPEGs around and naming them according to its own whims.
That's iPhoto's purpose in life. Your objection is like saying, "My big problem with Excel is the way it adds all the numbers together."
Just out of my own curiosity... are you also one of those guys who complains that iTunes organizes your music for you? I ask because those guys are impossible for me to understand. We rolled out a product that was basically the holy grail of music management, and they complained that it managed their music for them. They insisted that they would rather name and sort each of 20,000 tiny files themselves, with all the obsessive focus of an autistic kid counting paper clips. It was freaky.
We actually have two low-end machines now. The low-end G5 is defined by the fact that it only has room for 4 GB of RAM and the absence of PCI-X. We ship a 1.8 GHz version for $1,499 and a dual 2.0 GHz version for $1,999.
The 1.8 GHz version is what we refer to as the "Power Mac mini."
Is anybody still wondering why the rumor mill is so damaging to Apple? This is a solid upgrade: faster CPUs, dual-layer 16X Superdrives, higher base RAM, same prices. But what does everybody say? "They're not dual core."
Guys, the dual-core CPUs are a myth. They exist only as prototypes. Apple has never even so much as implied that we plan to use them for anything, ever. Being disappointed when a product we never announced fails to materialize is, frankly, pretty bizarre.
(Incidentally, I don't know how wide-spread it was, but the dual-layer Superdrives were referred to by at least a few people internally as "Superduperdrives." I thought that was pretty funny.)
Maybe I'm off base here, but I can't recall ever seeing a mail program that would export messages to another format. However, every mail program can import messages from other formats.
The emlx format is incredibly simple. It's just your standard mail message smooshed up against a property list that contains easily parsed metadata. I wouldn't be surprised if other mail programs can already import that format.
Closer to A than you might think. Some chapters made the rounds at the office a few weeks ago. I haven't read the whole book, but I've read chunks of it.
And it's not so much that I've done thorough research on Steve as it is that I see him at work every day.
It showed that a cooperative development based on free will and open standards could lead to complex software projects
Sorry, I guess I didn't get the memo. Was that ever in doubt?
What this "cooperative development" thing cannot do is produce consistent results. Have there been success stories? You betcha. But they're so few and far between that all they really do is serve to emphasize the barrenness of the landscape between them.
Linux is, for all intents and purposes, dead as anything other than a server operating system. And frankly it was dead as a server operating system until SGI came along and dragged it into the 21st century. It has one and only one advantage that keeps it hanging on: It'll run on leftover hardware. Got a PC from 1995 that you're not using for anything? Put Linux on it and use it to serve files. It won't be easy, but if you can get it to work at all, it'll work well. That's an important niche to occupy. But it's fundamentally a stagnant one.
Linux contributed to bring several technologies that were only available on costly enterprise systems to the masses
Let's be fair here. Linux hasn't contributed a damn thing. It can't, because of its license. Nobody can actually use Linux code for anything unless they're willing to give up commercial control of their project. Nobody is willing to do that.
Want to talk about what the BSD guys contributed? Go right ahead. You'll get no argument from me. But Linux has contributed nothing. Rather, it cloned existing implementations and walled them off so nobody could actually build anything on them. Which I think goes a long way toward explaining why Linux has stagnated while Mac OS X has surged ahead. We can surge ahead because we're not constantly getting harassed by lawyers from the FSF.
I find pretty sane to see developers asking themselves "How can we make that software easier to use ?" all the time.
Sure it is. But it's not sane to ask those questions to the exclusion of other areas of advancement. Oh, sure, those guys are tweaking the background color of the file browser, but they're overlooking the completely fucked up font rendering architecture, or the essential inability to localize the system. That kind of thing.
Linux isn't a "file-by-file" clone of Unix
Sure it is. All the run-time programs are basically file-by-file copies of Unix programs: init, inetd, the various networking daemons, all the command-line tools. Over the years people have sat down and copied Unix --Unix from the 1970s, remember -- file by file like the monks of old, dutifully reproducing everything even if it's just obviously stupid, like the init/inetd/cron/init.d/rc disaster.
And launchd didn't scrap all of them - it is just a system to unify daemons configuration and launching. Note that similar tools exist under Linux and the BSDs and that every major distribution features standardized configuration files.
Maybe you don't understand what launchd is. The launchd program replaces six entire subsystems. It gets rid entirely of init, rc, init.d, SystemStarter, cron and inetd. Makes them go away. No, no similar tools exist on another operating system.
And what if a critical service fails to start because of its malformed XML configuration file ?
Define "critical." We're talking about the difference between a bootable configuration and a non-bootable configuration here. On Linux, if init fails to start, the computer cannot be accessed by anybody, anywhere, via any method. One typo in the init configuration file can render the computer completely non-functional.
There's no such problem with launchd. If some configuration file gets fat-fingered, the service described by that file won't start. Maybe you'll be without SSH access, or without Open Directory. But the system will be running, and the problem can be fixed.
One of the features we've built into the preferences
Do that, and it would work over ANY file system... not just HFS+.
It does. You're uninformed. I'm gonna say it for what feels like the sixty-three millionth time: Go read the developer documentations.
Well, I'd really like to run with UFS instead of HFS+ under me.
There's absolutely no reason to do that. Zero. Zilch. None.
Since then I've had several occasions where HFS got so corrupted that fsck_hfs couldn't fix it.
That utility was deprecated two OS releases ago.
Between "Here's a bunch of technical mumbo-jumbo that isn't remotely true" and "HFS Extended sucks for reasons related to a three-year-old utility that seem strangely to be unreproducible by anybody else" I'm starting to wonder if you're not what they call around here a "troll." I think you're just posting nonsense for the heck of it.
Seriously: They're toolbar icons. If you don't like them, you're welcome to turn the toolbar off entirely, or to remove everything but the Spotlight search field. Whatever.
That's the point. It would have been so easy for them to get it right.
Why?
Because it makes your product look worse than it is --that's sort of being charitable here; it's entirely possible that the product really does suck and that a half-assed demo isn't doing it any harm --and because it misleads your customers.
Let's not move the goal posts, shall we? If the only politically acceptable option is a public-domain product or format, you're going to be disappointed a whole lot of the time.
Open doesn't mean "Whatever the politicos tell me it means this week." It means open. Documented. That's all.
Which is why Apple got their HTML-engine from Linux?
HTML rendering is a commodity, like a TCP stack. Why not take what's available for licensing so we can spend our time on other things? Meanwhile, let's talk about Core Data. Let's talk about launchd. Let's talk about Open Directory. Let's talk about Quartz. Let's talk about advanced typography. Let's talk about any of the technologies we've rolled out in the past five years that put us an entire generation ahead of Linux.
And what is this "Linux" you talk about?
Are you kidding?
Color-coding those close/maximise/minimise-buttons is just a big "fuck you!" to color-blind people.
Ever seen a traffic signal?
Spotlight is like Beagle on GNOME
No, it's not. Beagle is a content-indexing system similar to what every other operating system has. There's no relation between Spotlight and Beagle except in the very most superficial sense imaginable. All our developer docs on Spotlight go live on Friday. Please go read them. In particular, focus on the kernel-integration part so you can understand how file operations trigger import tasks. Focus on the extensible schema with support for arbitrary metadata attributes. Focus on the query language. These things are important.
Why is it that Apple "innovated" when they announced Spotlight, but Linux-guys did not when they demonstrated Beagle before Apple demonstrated their solution?
Let's set aside for the moment the unarguable fact that Beagle has as much to do with Spotlight as a raven has to do with a writing desk. Let's just ignore that for a moment.
Spotlight is shipping. Now. Released. Working. In actual production. With full support and documentation. Third parties are already shipping Spotlight importers and Spotlight-savvy applications in advance of Tiger's release on Friday.
Beagle is a loose proof-of-concept that even the developers say is for experimental purposes only. It has no documentation whatsoever and the UI looks like it was cobbled together by a twelve-year-old using Microsoft Paint. And according to the Web site, the way to run it is to download source code from a source-control server.
I'm gonna say that again, because it matters. Spotlight is finished and shipping right now. The only way to run Beagle is to download the source code.
Tell me again how they're just alike?
For the simple reason that I think Linux/KDE is better.
Better at what? I'm having a hard time thinking of things that Linux can do at all, much less better than something else.
I can't give you details because I just don't have 'em, but there's going to be a pretty big rewrite of iPhoto hitting the street real soon now.
This is one of the dirty little secrets of Tiger: iPhoto is totally, 100% incompatible with Spotlight. We're gonna fix that, obviously, but it's a big job.
See, Spotlight calls for metadata to be stored inside files. That's why we changed the way Mail works, creating a new mail message file format (emlx) that's basically an mbox-style mail message concatenated with an XML property list. That way we can store a message and all relevant metadata in one file, making it trivial for Spotlight to index it.
iPhoto doesn't work like that. iPhoto stores all its metadata in a database, and generates a buttload of ancillary files for thumbnails and albums. That's very much not Spotlight-friendly. Plus, as you point out, it's got a big scaling problem.
So we're gonna be releasing a new version, referred to internally as 5.1 but that may not be the actual number, real soon now. When? Dunno. What specific features will it have? Dunno. But it's coming.
Dragging a disk from the desktop to a privacy pane shortly after I plug the device in doesn't stop it, though I assume it trashes the resulting index on the drive once it's finished, RIGHT?
Nope. It shuts the mdimport task down.
What happens if I plug that volume into a 10.3 system next, and move off all my sensitive data? Is the index still there?
I'm still not quite understanding the question. Are you asking me where the indices are stored? They're stored in a/.Spotlight folder on each volume.
What bothers me is the big fat grey border around all the windows using the 'brushed metal' UI.
Sorry. It's not something we're going to change.
Why the big thick grey border?
Why not? Seriously, man, you're way overthinking this. It's just a design element. It's there because that's what the guys in the art department put there. You shouldn't get apoplectic over it.
And will it go away, with a phasing out of Brushed Metal?
I'm not sure where you got the idea that anything was being phased out. It's not.
I was moving my old build of imapd back into/usr/local/sbin, after an archive-and-install.
I'm confused. I thought you were talking about moving something from one folder in/System to another folder in/System. Maybe you should explain in greater detail.
And if you want to file a bug, go right ahead. RadarWeb is open to anybody with an Apple ID. I'm not a programmer, so any bug I file will look exactly like a bug you file, and at this point I don't even understand the question.
Practically everything in Mac OS X can be compiled for Intel hardware with little or no modification. There are some huge exceptions, of course; all the vectorization would have to be pulled out, and there's an assload of that now. But while it's a lot of code, it's not a big fraction of the OS.
Hell, our core operating system, Darwin, has been available for Intel for some years now.
But from two messages up, the "Apple is switching to Intel" stuff is complete bullshit. The performance we're getting out of 2.5 GHz G5s on the shelves today is spectacular. It's still, even a year later, top-shelf performance.
Yes, IBM has had massive problems going to 90 nm fabrication. But so has everybody else. We would have been real happy if IBM could have been at 5 GHz now like they'd projected back in 2002-2003. Going on-stage at WWDC in the summer of '03 and saying "3 GHz in a year!" left us looking really stupid. But shit happens, you know? It's not like IBM is totally dropping the ball on us (yes, I'm talking about you, Motorola).
According to the rumor mill, it just gets into his personal life to a degree that Steve wasn't happy with. There's stuff in there about his parentage --Steve is said to not know who his biological father was --and his marriage. There's some stuff about his cancer, too, stuff that Steve didn't want talked about, and that certainly wasn't relevant to any subject the book could cover. Total National Enquirer stuff.
That's all office gossip. It's possible that it's all wrong.
Technically Apple is a public company, to the extent that shares are traded on the open market. But as a company and not (for example) an agency of the government, Apple is entitled to the same privileges and rights that a private citizen would enjoy, more or less.
In other words, you were 100% right in essence, but you got the terminology a little turned around. Just thought you'd like to know.
Steve Jobs is an unbelievable asshole. Seriously, you think you know assholes? Steve Jobs makes them all look like girl scouts. He's the biggest asshole I've ever encountered, ever.
He's also a hands-down, certifiable genius.
The fact that he's an asshole really pales in comparison to what he's capable of doing. The man is like some kind of magic crap detector. He can smell crap from ten miles away. And when he sees it, he can tell you exactly why it's crap, and exactly what needs to change to eliminate its crappiness.
Steve Jobs is the most arrogant man I know, but he's also the man most deserving of open, unapologetic arrogance I know.
At least for me, it's possible to personally dislike somebody and admire him at the same time.
How does that follow? We add and remove products from our online and brick-and-mortar stores every week. Severing the relationship with a vendor of books isn't going to have any effect at all on the bottom line.
To the contrary, I think that Steve (if this was actually his call; it's all rumor-mill to me) is doing exactly what he's supposed to do. He's protecting the company's brand, which is far and away our most valuable asset.
They're trying to forcefully strongarm the actions of another company.
That conclusion isn't really supported by the facts. There's a much simpler interpretation, and we all know the rule about simpler interpretations.
1. Wiley publishes a book that's frankly pretty disgusting in the way it seeks to drag Jobs' dirty laundry out into the public eye for nothing more than the prurient interest.
2. Apple, which sends a big fat check to Wiley every month for books to stock in our retail stores, decides to stop paying money to a publisher that would do something so low.
3. End of story.
It's not about strong-arming anybody. We're not trying to get Wiley to pull the book or anything. We just don't see the point in continuing to sell the products of a company that we don't have a good relationship with.
I would really love to know what is going through people's heads when they complain about something like this.
Me too.
(Yeah, I know I took your last out of context. But look closely at it and tell me if I didn't make my point.)
But ask yourself this: what good has ever come from governments or corporations bullying the press?
God forbid we should continue to live in a free market. Who the hell cares what good comes of it? Everybody involved in this mess is making decisions that they are 100% entitled to make. Wiley publishes a book Apple doesn't like? Apple ceases to line Wiley's pockets. Give and take, free market.
Will I have a better experience at The Apple Store because Apple has decided to pull some Mac books not because of their content but in retalitiation?
Are you going to stop buying Apple products or shopping at Apple stores over this? If so, then you're what I think the economists call a "fragile customer." (I'm not sure I have that term right. I'm remembering twenty-year-old lectures here.) In other words, your commitment to the vendor was so shallow that the slightest thing would have tipped you one way or the other. Price goes up $5? You're outta there. Traffic on the way to the store? You're outta there. Don't like Steve Jobs' latest haircut? You're outta there.
Companies write off customers like that all the time. It's part of the cost of doing business.
Apple's customers, however, are almost exclusively people who are not like that. Apple has spectacularly high brand loyalty according to market research. Sales don't fluctuate very much at all, not even when prices go up significantly. It's not an economic get-out-of-jail-free card, but it means that Apple, as a company, doesn't really need to give a shit about customers who storm out in a huff because we severed a relationship with a third party.
(Incidentally, Nikon has exactly the same kind of customer base.)
Slashdot Mirror
IBM had trouble going from 180 nm fabrication to 90 nm fabrication. We announced 3 GHz by 2004, but internally we were predicting 5 GHz by this summer, based on IBM's projections.
... Slashdotters are less interested in the truth than they are in making a big stink! But surely that's not possible ...
They didn't pan out.
For the record, Steve Jobs copped to this publicly during one of his keynote addresses. WWDC 2004, I think it was. Funnily enough, though, it seems like Slashdotters like to remember the initial projection, not the retraction, which is strange because they happened in exactly the same forum attended by exactly the same group of people. It almost seems like
My big problem with iPhoto is the way it insists on moving all the JPEGs around and naming them according to its own whims.
... are you also one of those guys who complains that iTunes organizes your music for you? I ask because those guys are impossible for me to understand. We rolled out a product that was basically the holy grail of music management, and they complained that it managed their music for them. They insisted that they would rather name and sort each of 20,000 tiny files themselves, with all the obsessive focus of an autistic kid counting paper clips. It was freaky.
That's iPhoto's purpose in life. Your objection is like saying, "My big problem with Excel is the way it adds all the numbers together."
Just out of my own curiosity
We actually have two low-end machines now. The low-end G5 is defined by the fact that it only has room for 4 GB of RAM and the absence of PCI-X. We ship a 1.8 GHz version for $1,499 and a dual 2.0 GHz version for $1,999.
The 1.8 GHz version is what we refer to as the "Power Mac mini."
Is anybody still wondering why the rumor mill is so damaging to Apple? This is a solid upgrade: faster CPUs, dual-layer 16X Superdrives, higher base RAM, same prices. But what does everybody say? "They're not dual core."
Guys, the dual-core CPUs are a myth. They exist only as prototypes. Apple has never even so much as implied that we plan to use them for anything, ever. Being disappointed when a product we never announced fails to materialize is, frankly, pretty bizarre.
(Incidentally, I don't know how wide-spread it was, but the dual-layer Superdrives were referred to by at least a few people internally as "Superduperdrives." I thought that was pretty funny.)
Maybe I'm off base here, but I can't recall ever seeing a mail program that would export messages to another format. However, every mail program can import messages from other formats.
The emlx format is incredibly simple. It's just your standard mail message smooshed up against a property list that contains easily parsed metadata. I wouldn't be surprised if other mail programs can already import that format.
Closer to A than you might think. Some chapters made the rounds at the office a few weeks ago. I haven't read the whole book, but I've read chunks of it.
And it's not so much that I've done thorough research on Steve as it is that I see him at work every day.
It showed that a cooperative development based on free will and open standards could lead to complex software projects
Sorry, I guess I didn't get the memo. Was that ever in doubt?
What this "cooperative development" thing cannot do is produce consistent results. Have there been success stories? You betcha. But they're so few and far between that all they really do is serve to emphasize the barrenness of the landscape between them.
Linux is, for all intents and purposes, dead as anything other than a server operating system. And frankly it was dead as a server operating system until SGI came along and dragged it into the 21st century. It has one and only one advantage that keeps it hanging on: It'll run on leftover hardware. Got a PC from 1995 that you're not using for anything? Put Linux on it and use it to serve files. It won't be easy, but if you can get it to work at all, it'll work well. That's an important niche to occupy. But it's fundamentally a stagnant one.
Linux contributed to bring several technologies that were only available on costly enterprise systems to the masses
Let's be fair here. Linux hasn't contributed a damn thing. It can't, because of its license. Nobody can actually use Linux code for anything unless they're willing to give up commercial control of their project. Nobody is willing to do that.
Want to talk about what the BSD guys contributed? Go right ahead. You'll get no argument from me. But Linux has contributed nothing. Rather, it cloned existing implementations and walled them off so nobody could actually build anything on them. Which I think goes a long way toward explaining why Linux has stagnated while Mac OS X has surged ahead. We can surge ahead because we're not constantly getting harassed by lawyers from the FSF.
I find pretty sane to see developers asking themselves "How can we make that software easier to use ?" all the time.
Sure it is. But it's not sane to ask those questions to the exclusion of other areas of advancement. Oh, sure, those guys are tweaking the background color of the file browser, but they're overlooking the completely fucked up font rendering architecture, or the essential inability to localize the system. That kind of thing.
Linux isn't a "file-by-file" clone of Unix
Sure it is. All the run-time programs are basically file-by-file copies of Unix programs: init, inetd, the various networking daemons, all the command-line tools. Over the years people have sat down and copied Unix --Unix from the 1970s, remember -- file by file like the monks of old, dutifully reproducing everything even if it's just obviously stupid, like the init/inetd/cron/init.d/rc disaster.
And launchd didn't scrap all of them - it is just a system to unify daemons configuration and launching. Note that similar tools exist under Linux and the BSDs and that every major distribution features standardized configuration files.
Maybe you don't understand what launchd is. The launchd program replaces six entire subsystems. It gets rid entirely of init, rc, init.d, SystemStarter, cron and inetd. Makes them go away. No, no similar tools exist on another operating system.
And what if a critical service fails to start because of its malformed XML configuration file ?
Define "critical." We're talking about the difference between a bootable configuration and a non-bootable configuration here. On Linux, if init fails to start, the computer cannot be accessed by anybody, anywhere, via any method. One typo in the init configuration file can render the computer completely non-functional.
There's no such problem with launchd. If some configuration file gets fat-fingered, the service described by that file won't start. Maybe you'll be without SSH access, or without Open Directory. But the system will be running, and the problem can be fixed.
One of the features we've built into the preferences
Do that, and it would work over ANY file system... not just HFS+.
It does. You're uninformed. I'm gonna say it for what feels like the sixty-three millionth time: Go read the developer documentations.
Well, I'd really like to run with UFS instead of HFS+ under me.
There's absolutely no reason to do that. Zero. Zilch. None.
Since then I've had several occasions where HFS got so corrupted that fsck_hfs couldn't fix it.
That utility was deprecated two OS releases ago.
Between "Here's a bunch of technical mumbo-jumbo that isn't remotely true" and "HFS Extended sucks for reasons related to a three-year-old utility that seem strangely to be unreproducible by anybody else" I'm starting to wonder if you're not what they call around here a "troll." I think you're just posting nonsense for the heck of it.
Seriously: They're toolbar icons. If you don't like them, you're welcome to turn the toolbar off entirely, or to remove everything but the Spotlight search field. Whatever.
What was wrong with the old ones?
They were ugly.
It's just a bloody dropshadow.
That's the point. It would have been so easy for them to get it right.
Why?
Because it makes your product look worse than it is --that's sort of being charitable here; it's entirely possible that the product really does suck and that a half-assed demo isn't doing it any harm --and because it misleads your customers.
To think that somebody would give up all the features of Tiger because the Mail toolbar buttons look different boggles the mind.
But whatever, dude. It's a free country.
Let's not move the goal posts, shall we? If the only politically acceptable option is a public-domain product or format, you're going to be disappointed a whole lot of the time.
Open doesn't mean "Whatever the politicos tell me it means this week." It means open. Documented. That's all.
Which is why Apple got their HTML-engine from Linux?
HTML rendering is a commodity, like a TCP stack. Why not take what's available for licensing so we can spend our time on other things? Meanwhile, let's talk about Core Data. Let's talk about launchd. Let's talk about Open Directory. Let's talk about Quartz. Let's talk about advanced typography. Let's talk about any of the technologies we've rolled out in the past five years that put us an entire generation ahead of Linux.
And what is this "Linux" you talk about?
Are you kidding?
Color-coding those close/maximise/minimise-buttons is just a big "fuck you!" to color-blind people.
Ever seen a traffic signal?
Spotlight is like Beagle on GNOME
No, it's not. Beagle is a content-indexing system similar to what every other operating system has. There's no relation between Spotlight and Beagle except in the very most superficial sense imaginable. All our developer docs on Spotlight go live on Friday. Please go read them. In particular, focus on the kernel-integration part so you can understand how file operations trigger import tasks. Focus on the extensible schema with support for arbitrary metadata attributes. Focus on the query language. These things are important.
Why is it that Apple "innovated" when they announced Spotlight, but Linux-guys did not when they demonstrated Beagle before Apple demonstrated their solution?
Let's set aside for the moment the unarguable fact that Beagle has as much to do with Spotlight as a raven has to do with a writing desk. Let's just ignore that for a moment.
Spotlight is shipping. Now. Released. Working. In actual production. With full support and documentation. Third parties are already shipping Spotlight importers and Spotlight-savvy applications in advance of Tiger's release on Friday.
Beagle is a loose proof-of-concept that even the developers say is for experimental purposes only. It has no documentation whatsoever and the UI looks like it was cobbled together by a twelve-year-old using Microsoft Paint. And according to the Web site, the way to run it is to download source code from a source-control server.
I'm gonna say that again, because it matters. Spotlight is finished and shipping right now. The only way to run Beagle is to download the source code.
Tell me again how they're just alike?
For the simple reason that I think Linux/KDE is better.
Better at what? I'm having a hard time thinking of things that Linux can do at all, much less better than something else.
Same reason we stock all the third-party products in our retail stores: To bring in customers.
Oh, I thought he was asking about the toolbar itself. Buttons are just bitmaps, nothing special.
Why? In what way would that make your computer easier to use? Or even prettier, for that matter?
Do you happen to work for Microsoft? Because that sounds like a very Microsoft-y idea. Superficially kinda neat, fundamentally very stupid.
I can't give you details because I just don't have 'em, but there's going to be a pretty big rewrite of iPhoto hitting the street real soon now.
This is one of the dirty little secrets of Tiger: iPhoto is totally, 100% incompatible with Spotlight. We're gonna fix that, obviously, but it's a big job.
See, Spotlight calls for metadata to be stored inside files. That's why we changed the way Mail works, creating a new mail message file format (emlx) that's basically an mbox-style mail message concatenated with an XML property list. That way we can store a message and all relevant metadata in one file, making it trivial for Spotlight to index it.
iPhoto doesn't work like that. iPhoto stores all its metadata in a database, and generates a buttload of ancillary files for thumbnails and albums. That's very much not Spotlight-friendly. Plus, as you point out, it's got a big scaling problem.
So we're gonna be releasing a new version, referred to internally as 5.1 but that may not be the actual number, real soon now. When? Dunno. What specific features will it have? Dunno. But it's coming.
Dragging a disk from the desktop to a privacy pane shortly after I plug the device in doesn't stop it, though I assume it trashes the resulting index on the drive once it's finished, RIGHT?
/.Spotlight folder on each volume.
/usr/local/sbin, after an archive-and-install.
/System to another folder in /System. Maybe you should explain in greater detail.
Nope. It shuts the mdimport task down.
What happens if I plug that volume into a 10.3 system next, and move off all my sensitive data? Is the index still there?
I'm still not quite understanding the question. Are you asking me where the indices are stored? They're stored in a
What bothers me is the big fat grey border around all the windows using the 'brushed metal' UI.
Sorry. It's not something we're going to change.
Why the big thick grey border?
Why not? Seriously, man, you're way overthinking this. It's just a design element. It's there because that's what the guys in the art department put there. You shouldn't get apoplectic over it.
And will it go away, with a phasing out of Brushed Metal?
I'm not sure where you got the idea that anything was being phased out. It's not.
I was moving my old build of imapd back into
I'm confused. I thought you were talking about moving something from one folder in
And if you want to file a bug, go right ahead. RadarWeb is open to anybody with an Apple ID. I'm not a programmer, so any bug I file will look exactly like a bug you file, and at this point I don't even understand the question.
Practically everything in Mac OS X can be compiled for Intel hardware with little or no modification. There are some huge exceptions, of course; all the vectorization would have to be pulled out, and there's an assload of that now. But while it's a lot of code, it's not a big fraction of the OS.
Hell, our core operating system, Darwin, has been available for Intel for some years now.
But from two messages up, the "Apple is switching to Intel" stuff is complete bullshit. The performance we're getting out of 2.5 GHz G5s on the shelves today is spectacular. It's still, even a year later, top-shelf performance.
Yes, IBM has had massive problems going to 90 nm fabrication. But so has everybody else. We would have been real happy if IBM could have been at 5 GHz now like they'd projected back in 2002-2003. Going on-stage at WWDC in the summer of '03 and saying "3 GHz in a year!" left us looking really stupid. But shit happens, you know? It's not like IBM is totally dropping the ball on us (yes, I'm talking about you, Motorola).
According to the rumor mill, it just gets into his personal life to a degree that Steve wasn't happy with. There's stuff in there about his parentage --Steve is said to not know who his biological father was --and his marriage. There's some stuff about his cancer, too, stuff that Steve didn't want talked about, and that certainly wasn't relevant to any subject the book could cover. Total National Enquirer stuff.
That's all office gossip. It's possible that it's all wrong.
Technically Apple is a public company, to the extent that shares are traded on the open market. But as a company and not (for example) an agency of the government, Apple is entitled to the same privileges and rights that a private citizen would enjoy, more or less.
In other words, you were 100% right in essence, but you got the terminology a little turned around. Just thought you'd like to know.
Steve Jobs is an unbelievable asshole. Seriously, you think you know assholes? Steve Jobs makes them all look like girl scouts. He's the biggest asshole I've ever encountered, ever.
He's also a hands-down, certifiable genius.
The fact that he's an asshole really pales in comparison to what he's capable of doing. The man is like some kind of magic crap detector. He can smell crap from ten miles away. And when he sees it, he can tell you exactly why it's crap, and exactly what needs to change to eliminate its crappiness.
Steve Jobs is the most arrogant man I know, but he's also the man most deserving of open, unapologetic arrogance I know.
At least for me, it's possible to personally dislike somebody and admire him at the same time.
How does that follow? We add and remove products from our online and brick-and-mortar stores every week. Severing the relationship with a vendor of books isn't going to have any effect at all on the bottom line.
To the contrary, I think that Steve (if this was actually his call; it's all rumor-mill to me) is doing exactly what he's supposed to do. He's protecting the company's brand, which is far and away our most valuable asset.
They're trying to forcefully strongarm the actions of another company.
That conclusion isn't really supported by the facts. There's a much simpler interpretation, and we all know the rule about simpler interpretations.
1. Wiley publishes a book that's frankly pretty disgusting in the way it seeks to drag Jobs' dirty laundry out into the public eye for nothing more than the prurient interest.
2. Apple, which sends a big fat check to Wiley every month for books to stock in our retail stores, decides to stop paying money to a publisher that would do something so low.
3. End of story.
It's not about strong-arming anybody. We're not trying to get Wiley to pull the book or anything. We just don't see the point in continuing to sell the products of a company that we don't have a good relationship with.
I would really love to know what is going through people's heads when they complain about something like this.
Me too.
(Yeah, I know I took your last out of context. But look closely at it and tell me if I didn't make my point.)
But ask yourself this: what good has ever come from governments or corporations bullying the press?
God forbid we should continue to live in a free market. Who the hell cares what good comes of it? Everybody involved in this mess is making decisions that they are 100% entitled to make. Wiley publishes a book Apple doesn't like? Apple ceases to line Wiley's pockets. Give and take, free market.
Will I have a better experience at The Apple Store because Apple has decided to pull some Mac books not because of their content but in retalitiation?
Are you going to stop buying Apple products or shopping at Apple stores over this? If so, then you're what I think the economists call a "fragile customer." (I'm not sure I have that term right. I'm remembering twenty-year-old lectures here.) In other words, your commitment to the vendor was so shallow that the slightest thing would have tipped you one way or the other. Price goes up $5? You're outta there. Traffic on the way to the store? You're outta there. Don't like Steve Jobs' latest haircut? You're outta there.
Companies write off customers like that all the time. It's part of the cost of doing business.
Apple's customers, however, are almost exclusively people who are not like that. Apple has spectacularly high brand loyalty according to market research. Sales don't fluctuate very much at all, not even when prices go up significantly. It's not an economic get-out-of-jail-free card, but it means that Apple, as a company, doesn't really need to give a shit about customers who storm out in a huff because we severed a relationship with a third party.
(Incidentally, Nikon has exactly the same kind of customer base.)