Hah, Dvorak's old hat man! It's so 1990s. I personally use the Gentoo keyboard, where my typing's 5-10% faster than Dvorak or Qwerty. That's because instead of it being one keyboard layout for everyone, the keys are actually reordered for every application in the most optimal layout.
With Qwerty or Dvorak, you have to use the same keys regardless of what the program is you're using them with. The "Q", for example, on a QWERTY keyboard, is always in the top left (on English language layouts. It's "A" that's in the top left for French "AZERTY" keyboards.)
However, with Gentoo, the keys move around. So, for example, in OpenOffice.org, because I have to type "O" a lot, the "O" is right there where the "D" is in a QWERTY keyboard. The "Q", on the other hand, is assigned to F2, because I rarely need it.
Some have criticised the layout, arguing that the 5% efficiency increase is more than offset by the fact that you have to spend a day compi^H^H^H^H^Hlearning the new layout. This may be a problem for some people, but if you do a lot of typing, it's obvious that this is much more efficient. And besides, you can always let it run overnight, with you learning how to type using the new layout when you'd normally be wasting time asleep.
You should try it. I find the best performance is with -funroll-fingers -O102.
If you're an American, you might want to switch over to T-Mobile. Basic web access is free over GPRS, though they offer additional plans (T-Zones, which is exclusive content, and non-port blocked GPRS access for things like laptops) for an additional fee.
I wonder if it'd be worth the Konqueror people taking WebKit/etc and porting it back to KDE, rather than trying to keep up with WebKit in KHTML when the latter is obviously having problems because of slight architectural differences.
This way the three groups, Nokia, KDE, and Apple, will be working on making one browser engine perfect, rather than working on two very similar systems that, really, have no major advantages over one-another.
Symbian has little relationship with OS X/OpenStep. It strikes me if this was easy for Nokia to do, it should be architecturally reasonable to port it to a KDE environment.
I assume you mean Open Firmware, not OpenFireware (I take it they have to burn those EPROMs with that software! Hahaha I kill myself.)
Anyway, yes, they are dropping Open Firmware. They've confirmed that explicitly. So it's likely to be either BIOS or EFI in the final version. Please see the JE linked to my.sig for references.
I wish I could agree with the conclusion. I just don't think it's true.
If you want to market OS X to non-Mac users, you have to get it to them. Limiting yourself to a handful of technologically knowledgable geeks, and then only those who do not have a problem with blatant copyright infringment, is going to "get it to" a very small group of people, relative to the population Apple wants to get it to.
And yes, it's going to be a subset of geeks who do this. Why? Because relatively few people are willing to set up dual booting computers simply to test an operating system. Very few people are willing to do that and download what's presumably in the order of 2-3G (Panther is 3 CDs plus a developers CD, I don't have Tiger but I'm guessing it's larger) from the Internet to put themselves in a position of being able to try it out.
If Apple wants to get this out to non-geeks, they have to create something similar to "BeOS for Windows", where installation is a matter of downloading the system from the Internet or copying it from a CD, unzipping the file, and then running something, from Windows or maybe by booting the CD, that boots into BeOS. I think Apple knows that this is what would be necessary, together with a very substantial (far more than Be did) advertising campaign.
What this may, ultimately, boil down to is Apple putting a toe in the water to see how bad piracy might end up being. They're releasing a one-off unupdatable OS (as you've said) that has no restrictions built into it whatsoever. If it ends up on a few thousand unauthorized desktops, then Apple's going to breath a sigh of relief and incorporate minimal restrictions in the final releases. If it spreads like wildfire, with millions of unauthorized installations over the next few months, then Apple will feel obliged to put far tougher restrictions into the eventual release.
A demo though? They don't have to release a stealth demo. They can do so publically. It makes little sense that this is a demo. A test, perhaps, but not a demo.
BTW, I'd be surprised if the developers boxes do not have serial numbers in their operating system copies that can be traced. It'll be interesting to see if there's a lawsuit over this.
I'm pretty sure "Leia in the golden bikini" was an "on screen in Episodes I-VI" instance of it being obvious slavery was still very much around in the Empire. On top of which, it's reasonably fair to suggest the status quo existed except where noted, so if slavery existed in the Republic, it existed in the Empire because nobody clearly abolished it.
"I've just received word that the Emperor has dissolved the senate permanently. I've also just been told he's also abolished slavery, which strikes me as a good idea, hmmm? What do you think of that, Vader? I know you were once a slave so you must be dead pleased. Anyway, sounds like good news, pretty much yhe last remnants of the Old Republic have been swept away."
Re:Patent non Problem
on
Drafting GPL3
·
· Score: 2, Informative
What is the problem?
Two issues. First, the present wording needs to be a little more explicit when it comes to patents. Technically, most people reading the relevent part of the GPL would come to the conclusion that a patent holder cannot distribute a GPL'd program without agreeing to license, by implication, any patents covering the same program. Such a license would only apply to the program in question and direct derivations though that continue to use the same code. So if someone rewrites the code that the patent covers, writing new code that infringes the patent, the situation is legally ambigious.
If you think I'm kidding, take a look at Nokia's supposedly generous offer to allow its patented technologies to be used in Linux, which somes with terms and conditions that imply Nokia could sue in the future over the same patents concerning future versions of the kernel. Work out under what circumstances they could do so (someone independently adding non-derived code that happens to infringe), and you suddenly have something that doesn't look as generous as it originally was.
The second issue is more of a political one. Patents suck. If someone chooses to go to war against a free software application by suing it for patent infringement, does that organization have any moral right to use free software at all?
There are few incentives that exist to discourage software patents. To be forced to never use free software again may be one of the few ways the free software community can fight back. With free software entrenched enough, this could be a good thing for those who believe in the freedom to program, and who oppose the notion of a "right" existing to implement a particular type of technology that can be expressed algorithmically.
Sigh. OK, lets see what percentage of mac users have the know-how to, and will be motivated to install Windows on a partition of their mac? I'd say maybe 10%, if that. What percentage of mac users buy games? I'd say maybe 70%. Thus, the vast majority of mac users won't have Windows but will want to run games. Do you see where I'm going with this?
It's not that difficult to install Windows. It's especially not difficult if you're using a standardized architecture and can get a friend to come over who "knows PCs". One thing I thought of after I wrote the response to you is that it doesn't even need repartioning - the user can get a cheap external drive and install Windows on that. We know that unless Apple goes out of its way to cripple the BIOS, their machines will be USB2 bootable.
As to the issue of whether or not there will be other hurdles to installing Windows on a mac, who knows, but the Apple VP said that apple would not actively try to stop users from running Windows, but will stop users from running OS X on another machine, that means a custom BIOS, chipset, a new kind of management system, or activating the DRM on the pentiums. The first two of those will require either user hacking or support from MS, which is unlikely.
That Mac today is an open system. It doesn't have a non-standard BIOS. It doesn't have a special chipset that makes OS X incompatable with anything else. It doesn't have any hardware DRM. Yet you can't install Mac OS X on the architecturally compatable IBM RS/6000. Why? Because the routines to boot OS X do some fairly trivial checks before loading it to make sure that OS X is running on "supported hardware".
Even so, your essential point is false. A custom BIOS is not enough to prevent an operating system like Windows from booting. Neither is a custom chipset. What would be required would be for those measures to be incompatable with Windows. That's far from likely. Apple wouldn't be saying what they're saying if they were going to do that.
Finally, as I've said before, Apple has said they'll not take any measures to prevent Windows from running. In fact, Schiller has outright said people "probably will" run Windows on their Macs.
Your ad hominem attack on me is weak, by the way. I really don't care if I have to run PC versions of games (which I rarely play) and I'll almost certainly be installing VMware on a new mac as soon as possible for testing reasons anyway.
I didn't say you cared about whether you can run PC versions of games. And it wasn't an ad hominem attack, it was an attempt to try to persuade you to see the obvious: I sincerely believe you're blinded by the fact you want the new Mac to not be a PC clone. I'm coming across so many Mac users who react with horror with every new revelation about it. I'm not calling upon others to dismiss your ideas because of that, I'm telling YOU to look into why you're arguing this is some kind of non-standard non-PC-clone box when EVERY SINGLE BIT OF NEWS RELEASED SO FAR says it is. If you're not of this mindset, and just haven't been following this in that much depth, then I apologise, but in that case, just to let you know, it's not a great day for people who do not like the PC architecture that much.
As a suggestion, take a look at the JE I've linked to from my.sig. I have a strong suspicion that you'll not like it, but I've sourced every single statement I've made as a statement of fact. Sure, there's some speculation, but even then, I've only posted that when it's highly improbable that it's false.
Your view on this situation is suffering from too much Slashdot. Please try to understand that 99.9% of computer users are not Slashdot geeks and wouldn't even attempt to re-install windows as the only operating system on their existing machine today. The vast majority of people who buy video games are included in that.
You are misinformed. The intel based macs will most likely be running a custom BIOS, not on compatible with Windows by default
You're accusing me of being misinformed?
Apple has said, outright, they intend to do nothing to prevent Windows from running. That's a fact. It's been widely quoted.
Most mac users (like most Windows users) do not dual boot their systems, or even have any idea how to partition and install a second OS
That's because there's no benefit to them as yet. What is the incentive to install OpenBSD, GNU/Linux, or AIX, on their other partition?
There's a huge incentive to install Windows on an ix86 based Mac. Out of the box, your choice of games, if you're interested in them, becomes pretty much the sum total of almost every game written since the mid nineties.
They certainly don't know how to, or don't care to hack their BIOS, which may be necessary.
Oh goodness gracious! What are you talking about? To remind you, we're talking about Apple shipping PCs with one minor change - they'll come with a version of OS X installed by default that can detect whether it's running on Apple hardware. That's it. You don't have to "hack the BIOS" to dual boot on any PC made now. Why, oh why, are you pretending such a thing will be necessary just because it has the word "Apple" on the front?
A wine style port will have a very large footprint, but will probably be used by some game developers once the project is up and running well on OS X. Wine ports of some games may be fine, while others will suck very badly (depending on the type of game and controls).
Sure. But you're missing the point here. The point is that developers are no longer going to care about porting games. There's a fairly large expense associated with doing so, and those developers are going to see little point in doing this themselves when Codeweavers can do it for them, and if Codeweavers eff up, the serious gamers will probably already have Windows anyway.
I know why you're arguing against this. You're arguing because I'm telling you stuff you don't want to hear. That's why you're making stuff up, pretending that Apple is going to disable Windows, or that dual booting is somehow hard (it's not even hard on a real Mac, believe me.) You're saying "This is going to suck, I'll have to run Windows or use a poor clone of the Windows API that might not run the games I want the way I want them". And you know what, you're right. But you're failing to look at the bigger picture. The developers are interested in making games that people will buy. Serious gamers will get Windows. Non-serious gamers will probably get Windows anyway, if only to help see the occasional website under IE or use some bit of software from the office. For the rest, open sourcers can help. It's no longer a matter of emulating a CPU, just an API. No, it'll not be perfect, but hey, that's your fault, you should have installed Windows, right?
If you really think most users are going to install a second OS and reboot every time they want to play a game, well you're an idiot.
Thanks, but I think the truth is you're delusional. Most people will look at the relative merits, and realise they don't have a choice in the matter. Just as now, the majority of games will be Windows-only. Just as now, the Mac games that exist will generally cost more than their PC counterparts. Those sales can and will dwindle as the majority of people put two and two together and realise they're better off buying Windows games than Mac games. There will be multiple ways in which those games can be run, but at least one of them will involve dual booting.
Will 90% of the Mac market not install Windows? I don't know. But I can tell you that 90% will probably not be playing many games.
But HP is selling iPods. Sure, they're rebranded, but they are, actually, iPods, from Apple. They're made at the same factories, from the same designs.
This is different to the suggestion that a high quality clone manufacturer (hmmm, IBM/Lenovo?) would actually be permitted to ships its own PCs with OS X installed.
They're updating the CLI, not the Console Window. And you've always been able to copy and paste to and from console windows anyway, check the system menu (the menu on the top left of the title bar.)
If you can buy an emulator for $200 and Windows for another $100
An emulator? Why is someone going to have to spend $200 on an emulator? What emulator? What's it going to emulate?
The only money they'll have to spend, assuming they're not going to pirate a copy, is on Windows itself. Once they have Windows, it'll be a simple matter of partioning the disk and installing it. That's it. Potentially, the only third party software they'll need is the dual boot menu software, and those will be readily available, as they are anyway today.
Apple is moving from Open Firmware PowerPC boxes to IBM PC clones with the latest Intel processors and BIOSes. They've said you'll be able to install Windows on them, and you will. I don't know where you're getting it from that you'll need an "emulator", I'm sure a few VMWare things will appear for those who want to run two operating systems at once, but dual booting will satisfy the majority of those who want to play the full range of Windows games more readily than waiting for Aspyr and MacGames to port the three games they plan to port this year. Indeed, a Codeweavers-style Wine port will probably be the Mac gamer's second choice, which leaves relatively little market left for those who want to port the games properly.
First, Intel chips wont be replacing the G5's anytime soon. Benches of Tiger running on Pentium 4's with Rhapsody got destroyed by Tiger on a G5.
Do you have a link that backs up the latter claim? The article you link to is about Rosetta's emulation under Tiger on a P4 vs "native" G5, not anyone benchmarking "Tiger with Rhapsody" (however you'd run that, I assume you mean someone pulled out an old Intel Rhapsody disk running it on a P4 and compared it to Tiger on a G5)?
And the quoted figures are actually quite impressive, an emulated (not natively compiled) app running on a new P4 will probably run about as fast as it would on whatever Mac the purchaser of the P4 Mac was replacing, as speeds seemed to hover around the 30-60% of a G5 range. Most Mac users do not have a G5.
Meanwhile, according to xlr8yourmac, native app performance on the P4 developer's box looks pretty good, better than a dual 2.7GHz G5 in some cases.
Not a bad processor. I have problems with a lot of what Apple's doing, but it does look like they're picking a decent processor and making it work.
Apple has confirmed, in the universal binaries document, that Intel Macs will not have Open Firmware. Additionally, the beta-Macs going out to developers have BIOSes, and there's little reason to believe that Apple will not put BIOSes in the final versions, for two good reasons. The first is that they've said they'll not do anything that'd prevent Microsoft Windows from running. And the second is, if they're not porting Open Firmware (which is processor independent), then what's the point of choosing firmware other than BIOS? The latter at least guarantees compatability with off-the-shelf video cards, amongst other things.
Do you think the games market is going to care HOW the Apple users get their copies of Windows? Or that those same Apple users run the very latest versions? Or that they're able to run Windows apps via some suddenly viable third party development like Wine for OS X?
No, it's a BIOS. And yes, he said Apple wouldn't do anything to preclude running Windows. Not having a BIOS would be a pretty significant preclusion.
It's a crying shame. Open Firmware on the Intel architecture would be a vast improvement. But it makes sense, with modern graphics cards requiring an Intel architecture and PC BIOS to start up. Apple would pretty much be guaranteeing a continuation of "Mac version" graphics cards if it went non-standard here. Apple would also not be able to take advantage of the huge amount of third party R&D in the PC market.
I assume you're one of the morons (hey, you started the insults) that was claiming Intel would be making PowerPCs right until the very end, correct? Look, sorry to tell you this, but Apple is going into the IBM PC clone business. It's sad. It's unfortunate. But that's what's happening. From now on, they're using two things to distinguish themselves from the herd: a different OS, and a passion for style. That's it.
A lot of this can be resolved by making intelligent decisions about what you're going to incorporate into the operating system and how you're going to structure it.
In the above case, you'd probably want to structure Apache as having a plug-in architecture, so the dependency on Apache is relatively obvious. I, personally, would also include it as part of the OS (as I said in my original comment, I think you should try to incorporate as much into the OS as possible within reason to remove the need for third party shared components.) PHP would be reasonable as part of the overall "Linux Web Services" package this'd be a part of.
MySQL I'd try to remove the dependency altogether, I'd rather the OS have a crude/generic DB (which could be MySQL) and ODBC-type framework built in for connecting to it or alternatives to it, and applications attempt to use it. I don't think I'm exaggerating much when I say the whole "Apps that need databases" thing is due for a serious rethink in general too.
Good example though, and one that anyone making a decision about how to put together a GNU/modified-Linux distro that integrates application management with object management should consider.
Actually, that's not really the problem, it would actually be wanted sometimes (eg a web developer having both FF1.0 and FF1.1 installed for compability checks)
This appears to be a reply to a, presumably unintended, quote out of context. I was talking about two apps with a shared component, where the shared component would end up being installed more than once if you go the monolithic, avoid-external-dependencies, route. Not two of the same app. ie in the example I give, both apps would have (statically linked) copies of the MPEG library. The intention behind package management systems was to get away from this, have the MPEG library a separate dependency, and have the two apps share the same copy, hence increasing efficiency.
I thought about this one for a while, and the phrase "sledgehammer to crack a nut" struck me as not quite appropriate for what I wanted to say. The gist though is that there are minor problems that aren't really problems, just inelegancies, and then there are "solutions" to those minor problems that are actually, totally, 100% awful. And when it comes down to it, I think the general direction of the packaging systems for *ix is a primary example.
Back in the late eighties and early nineties, for the most part, "installation" meant exactly what we're talking about is current in OS X, with the one exception that you generally had to tell "the system", somehow, where you'd copied the app. That is, under AmigaOS, you'd edit s:startup-sequence and put in an assign, under DOS you'd probably edit autoexec.bat and put in an entry in your set PATH= line, etc.
This latter bit of the process was, quite seriously, the only bad part of installations. Most packages came with instructions on how to do this last bit, usually in the form of running a quick program after you'd installed the app. The rest was easy. Apple solved this, at the time, even then, by including proper metadata in the operating system.
Time has passed, and now someone, somewhere, has declared this entire system, with or without the "you have to tell the OS where the app's installed" bit, fatally flawed. Why? Because if an application has a shared component, it might be installed more than once!
And is this important? Well, apparently so. You see, the additional library will take up more disk space, and more memory, and you know how short we are of both.
Now, there's room here for a little rant on how applications seem to have grown over the years disproportionally compared to (useful) functionality, but even leaving that aside, clearly memory and disk space are no longer considered the most important factors in computing right now. If I have two applications that manipulate MPEGs, and they both have a generic, half meg (let's be generous, 'cos it will not be that big), library for encoding MPEGs statically linked to them, is this a serious problem on any modern computer system? Or even any ten year old computer system?
The point is this isn't a problem. It's an inelegance. We can see, at a glance, that a fix for this issue would be a nice-to-have, but nothing more.
So what about the alternatives? Well, we now have half a dozen popular package management systems, varying from Slackware's "no dependency checking, read the documentation" installpkg system to Gentoo's "We're not just going to solve that minor issue, but we'll also solve the problem that your application is running 3-5% slower than it could do because it was compiled for the 80486" kitchen sink approach. What do these have in common? Are they elegant, trouble-free, approaches that result in applications running the way the developers intended?
Well, by and large, not really. We now have DLL hell, half the installation systems require we run around trying to find obscure packages that we didn't realise we needed from the start, the systems aren't integrated with the file system, which supposedly is the gateway to the computer, so adding complexity to the entire process of understanding what you have on your machine, and for the most part we've also limited the flexibility of users when organizing their own machines. We've also added complexity to the physical system of storing applications, creating another central set of registries etc that can break. It used to be the worst anyone could do to a PC, forcing a reinstall of everything, was overwrite the FAT. Now the disk filling up during an app install can potentially cause the same problem.
This is a fix? Because it looks to me like it's introduced more problems than it's solved. In fact, it looks like it hasn't solved any problems, it's merely added an ounce of hardware efficiency in exchange for a high degree of complexity. It's so bad that, in my e
OS/2 also had pre-emptive multitasking. And it supported FAT filesystems too. Let's hope OS X doesn't have either!
OS/2 Warp (3) did actually take off, at least in Europe. According to Judge Jackson's Findings of Fact, IBM was forced by Microsoft to drop all marketing and bundling of the OS and Lotus Smartsuite in exchange for the support it needed to sell PCs with Windows 95.
Did I say they would? No. I said that if Apple computers can run Windows, it's more reasonable for a game developer to presume you will dual boot if you want to play their games than to spend the development time porting it. Take a look at the GNU/Linux situation with games. There are a few, but game developers generally ignore it, assuming that as you have a machine capable of running Windows, you'll happily boot into Windows to play your games.
Microsoft has commented, at the WWDC during the announcement, saying the MacBU will port Office to the new architecture. I think that's very much a vote of confidence, not something said by someone taken unawares, right? You think Apple sprung this on them? Even the Wall Street Journal knew about this a month ago.
After Jobs' presentation, Apple Senior Vice President Phil Schiller addressed the issue of running Windows on Macs, saying there are no plans to sell or support Windows on an Intel-based Mac. "That doesn't preclude someone from running it on a Mac. They probably will," he said. "We won't do anything to preclude that."
then I'm pretty sure most game developers will target Windows, and look at OS X much the same way as they do GNU/Linux. "We can sell to this crowd, they just have to set up their machines to dual boot."
Not that I don't hope I'm wrong, but I suspect this is why Microsoft isn't looking that upset about the switch right now.
The man's a loon. Aside from the gratuitious insults against his audience, there's also the small matter that his original "prediction" was that Apple would switch to Itanium within 18 months. More than two years later, they're actually switching to Intel's classic ix86 family.
The only thing "right" about Dvorak's article is the notion of a processor change at some point in the future that involves Intel somewhere. Great crystal ball you have there Dvorak!
Dude, if you're going to post this, you at least also add speculation that Intel is going to be producing PowerPCs. That was the other thing everyone got really obsessed with yesterday before Jobs announced the move to Pentiums, with the $1000 developers box you can buy right now (additional $500 may be required to join ADC at correct level) that has a Pentium in it.
Slashdot Mirror
With Qwerty or Dvorak, you have to use the same keys regardless of what the program is you're using them with. The "Q", for example, on a QWERTY keyboard, is always in the top left (on English language layouts. It's "A" that's in the top left for French "AZERTY" keyboards.)
However, with Gentoo, the keys move around. So, for example, in OpenOffice.org, because I have to type "O" a lot, the "O" is right there where the "D" is in a QWERTY keyboard. The "Q", on the other hand, is assigned to F2, because I rarely need it.
Some have criticised the layout, arguing that the 5% efficiency increase is more than offset by the fact that you have to spend a day compi^H^H^H^H^Hlearning the new layout. This may be a problem for some people, but if you do a lot of typing, it's obvious that this is much more efficient. And besides, you can always let it run overnight, with you learning how to type using the new layout when you'd normally be wasting time asleep.
You should try it. I find the best performance is with -funroll-fingers -O102.
If you're an American, you might want to switch over to T-Mobile. Basic web access is free over GPRS, though they offer additional plans (T-Zones, which is exclusive content, and non-port blocked GPRS access for things like laptops) for an additional fee.
This way the three groups, Nokia, KDE, and Apple, will be working on making one browser engine perfect, rather than working on two very similar systems that, really, have no major advantages over one-another.
Symbian has little relationship with OS X/OpenStep. It strikes me if this was easy for Nokia to do, it should be architecturally reasonable to port it to a KDE environment.
Anyway, yes, they are dropping Open Firmware. They've confirmed that explicitly. So it's likely to be either BIOS or EFI in the final version. Please see the JE linked to my .sig for references.
If you want to market OS X to non-Mac users, you have to get it to them. Limiting yourself to a handful of technologically knowledgable geeks, and then only those who do not have a problem with blatant copyright infringment, is going to "get it to" a very small group of people, relative to the population Apple wants to get it to.
And yes, it's going to be a subset of geeks who do this. Why? Because relatively few people are willing to set up dual booting computers simply to test an operating system. Very few people are willing to do that and download what's presumably in the order of 2-3G (Panther is 3 CDs plus a developers CD, I don't have Tiger but I'm guessing it's larger) from the Internet to put themselves in a position of being able to try it out.
If Apple wants to get this out to non-geeks, they have to create something similar to "BeOS for Windows", where installation is a matter of downloading the system from the Internet or copying it from a CD, unzipping the file, and then running something, from Windows or maybe by booting the CD, that boots into BeOS. I think Apple knows that this is what would be necessary, together with a very substantial (far more than Be did) advertising campaign.
What this may, ultimately, boil down to is Apple putting a toe in the water to see how bad piracy might end up being. They're releasing a one-off unupdatable OS (as you've said) that has no restrictions built into it whatsoever. If it ends up on a few thousand unauthorized desktops, then Apple's going to breath a sigh of relief and incorporate minimal restrictions in the final releases. If it spreads like wildfire, with millions of unauthorized installations over the next few months, then Apple will feel obliged to put far tougher restrictions into the eventual release.
A demo though? They don't have to release a stealth demo. They can do so publically. It makes little sense that this is a demo. A test, perhaps, but not a demo.
BTW, I'd be surprised if the developers boxes do not have serial numbers in their operating system copies that can be traced. It'll be interesting to see if there's a lawsuit over this.
"I've just received word that the Emperor has dissolved the senate permanently. I've also just been told he's also abolished slavery, which strikes me as a good idea, hmmm? What do you think of that, Vader? I know you were once a slave so you must be dead pleased. Anyway, sounds like good news, pretty much yhe last remnants of the Old Republic have been swept away."
If you think I'm kidding, take a look at Nokia's supposedly generous offer to allow its patented technologies to be used in Linux, which somes with terms and conditions that imply Nokia could sue in the future over the same patents concerning future versions of the kernel. Work out under what circumstances they could do so (someone independently adding non-derived code that happens to infringe), and you suddenly have something that doesn't look as generous as it originally was.
The second issue is more of a political one. Patents suck. If someone chooses to go to war against a free software application by suing it for patent infringement, does that organization have any moral right to use free software at all?
There are few incentives that exist to discourage software patents. To be forced to never use free software again may be one of the few ways the free software community can fight back. With free software entrenched enough, this could be a good thing for those who believe in the freedom to program, and who oppose the notion of a "right" existing to implement a particular type of technology that can be expressed algorithmically.
It's not that difficult to install Windows. It's especially not difficult if you're using a standardized architecture and can get a friend to come over who "knows PCs". One thing I thought of after I wrote the response to you is that it doesn't even need repartioning - the user can get a cheap external drive and install Windows on that. We know that unless Apple goes out of its way to cripple the BIOS, their machines will be USB2 bootable.
That Mac today is an open system. It doesn't have a non-standard BIOS. It doesn't have a special chipset that makes OS X incompatable with anything else. It doesn't have any hardware DRM. Yet you can't install Mac OS X on the architecturally compatable IBM RS/6000. Why? Because the routines to boot OS X do some fairly trivial checks before loading it to make sure that OS X is running on "supported hardware".
Even so, your essential point is false. A custom BIOS is not enough to prevent an operating system like Windows from booting. Neither is a custom chipset. What would be required would be for those measures to be incompatable with Windows. That's far from likely. Apple wouldn't be saying what they're saying if they were going to do that.
Finally, as I've said before, Apple has said they'll not take any measures to prevent Windows from running. In fact, Schiller has outright said people "probably will" run Windows on their Macs.
I didn't say you cared about whether you can run PC versions of games. And it wasn't an ad hominem attack, it was an attempt to try to persuade you to see the obvious: I sincerely believe you're blinded by the fact you want the new Mac to not be a PC clone. I'm coming across so many Mac users who react with horror with every new revelation about it. I'm not calling upon others to dismiss your ideas because of that, I'm telling YOU to look into why you're arguing this is some kind of non-standard non-PC-clone box when EVERY SINGLE BIT OF NEWS RELEASED SO FAR says it is. If you're not of this mindset, and just haven't been following this in that much depth, then I apologise, but in that case, just to let you know, it's not a great day for people who do not like the PC architecture that much.
As a suggestion, take a look at the JE I've linked to from my .sig. I have a strong suspicion that you'll not like it, but I've sourced every single statement I've made as a statement of fact. Sure, there's some speculation, but even then, I've only posted that when it's highly improbable that it's false.
I'm not sure 9
Apple has said, outright, they intend to do nothing to prevent Windows from running. That's a fact. It's been widely quoted.
That's because there's no benefit to them as yet. What is the incentive to install OpenBSD, GNU/Linux, or AIX, on their other partition?There's a huge incentive to install Windows on an ix86 based Mac. Out of the box, your choice of games, if you're interested in them, becomes pretty much the sum total of almost every game written since the mid nineties.
Oh goodness gracious! What are you talking about? To remind you, we're talking about Apple shipping PCs with one minor change - they'll come with a version of OS X installed by default that can detect whether it's running on Apple hardware. That's it. You don't have to "hack the BIOS" to dual boot on any PC made now. Why, oh why, are you pretending such a thing will be necessary just because it has the word "Apple" on the front? Sure. But you're missing the point here. The point is that developers are no longer going to care about porting games. There's a fairly large expense associated with doing so, and those developers are going to see little point in doing this themselves when Codeweavers can do it for them, and if Codeweavers eff up, the serious gamers will probably already have Windows anyway.I know why you're arguing against this. You're arguing because I'm telling you stuff you don't want to hear. That's why you're making stuff up, pretending that Apple is going to disable Windows, or that dual booting is somehow hard (it's not even hard on a real Mac, believe me.) You're saying "This is going to suck, I'll have to run Windows or use a poor clone of the Windows API that might not run the games I want the way I want them". And you know what, you're right. But you're failing to look at the bigger picture. The developers are interested in making games that people will buy. Serious gamers will get Windows. Non-serious gamers will probably get Windows anyway, if only to help see the occasional website under IE or use some bit of software from the office. For the rest, open sourcers can help. It's no longer a matter of emulating a CPU, just an API. No, it'll not be perfect, but hey, that's your fault, you should have installed Windows, right?
Thanks, but I think the truth is you're delusional. Most people will look at the relative merits, and realise they don't have a choice in the matter. Just as now, the majority of games will be Windows-only. Just as now, the Mac games that exist will generally cost more than their PC counterparts. Those sales can and will dwindle as the majority of people put two and two together and realise they're better off buying Windows games than Mac games. There will be multiple ways in which those games can be run, but at least one of them will involve dual booting.Will 90% of the Mac market not install Windows? I don't know. But I can tell you that 90% will probably not be playing many games.
This is different to the suggestion that a high quality clone manufacturer (hmmm, IBM/Lenovo?) would actually be permitted to ships its own PCs with OS X installed.
They're updating the CLI, not the Console Window. And you've always been able to copy and paste to and from console windows anyway, check the system menu (the menu on the top left of the title bar.)
The only money they'll have to spend, assuming they're not going to pirate a copy, is on Windows itself. Once they have Windows, it'll be a simple matter of partioning the disk and installing it. That's it. Potentially, the only third party software they'll need is the dual boot menu software, and those will be readily available, as they are anyway today.
Apple is moving from Open Firmware PowerPC boxes to IBM PC clones with the latest Intel processors and BIOSes. They've said you'll be able to install Windows on them, and you will. I don't know where you're getting it from that you'll need an "emulator", I'm sure a few VMWare things will appear for those who want to run two operating systems at once, but dual booting will satisfy the majority of those who want to play the full range of Windows games more readily than waiting for Aspyr and MacGames to port the three games they plan to port this year. Indeed, a Codeweavers-style Wine port will probably be the Mac gamer's second choice, which leaves relatively little market left for those who want to port the games properly.
And the quoted figures are actually quite impressive, an emulated (not natively compiled) app running on a new P4 will probably run about as fast as it would on whatever Mac the purchaser of the P4 Mac was replacing, as speeds seemed to hover around the 30-60% of a G5 range. Most Mac users do not have a G5.
Meanwhile, according to xlr8yourmac, native app performance on the P4 developer's box looks pretty good, better than a dual 2.7GHz G5 in some cases.
Not a bad processor. I have problems with a lot of what Apple's doing, but it does look like they're picking a decent processor and making it work.
Apple has confirmed, in the universal binaries document, that Intel Macs will not have Open Firmware. Additionally, the beta-Macs going out to developers have BIOSes, and there's little reason to believe that Apple will not put BIOSes in the final versions, for two good reasons. The first is that they've said they'll not do anything that'd prevent Microsoft Windows from running. And the second is, if they're not porting Open Firmware (which is processor independent), then what's the point of choosing firmware other than BIOS? The latter at least guarantees compatability with off-the-shelf video cards, amongst other things.
Do you think the games market is going to care HOW the Apple users get their copies of Windows? Or that those same Apple users run the very latest versions? Or that they're able to run Windows apps via some suddenly viable third party development like Wine for OS X?
It's a crying shame. Open Firmware on the Intel architecture would be a vast improvement. But it makes sense, with modern graphics cards requiring an Intel architecture and PC BIOS to start up. Apple would pretty much be guaranteeing a continuation of "Mac version" graphics cards if it went non-standard here. Apple would also not be able to take advantage of the huge amount of third party R&D in the PC market.
I assume you're one of the morons (hey, you started the insults) that was claiming Intel would be making PowerPCs right until the very end, correct? Look, sorry to tell you this, but Apple is going into the IBM PC clone business. It's sad. It's unfortunate. But that's what's happening. From now on, they're using two things to distinguish themselves from the herd: a different OS, and a passion for style. That's it.
A lot of this can be resolved by making intelligent decisions about what you're going to incorporate into the operating system and how you're going to structure it.
In the above case, you'd probably want to structure Apache as having a plug-in architecture, so the dependency on Apache is relatively obvious. I, personally, would also include it as part of the OS (as I said in my original comment, I think you should try to incorporate as much into the OS as possible within reason to remove the need for third party shared components.) PHP would be reasonable as part of the overall "Linux Web Services" package this'd be a part of.
MySQL I'd try to remove the dependency altogether, I'd rather the OS have a crude/generic DB (which could be MySQL) and ODBC-type framework built in for connecting to it or alternatives to it, and applications attempt to use it. I don't think I'm exaggerating much when I say the whole "Apps that need databases" thing is due for a serious rethink in general too.
Good example though, and one that anyone making a decision about how to put together a GNU/modified-Linux distro that integrates application management with object management should consider.
Back in the late eighties and early nineties, for the most part, "installation" meant exactly what we're talking about is current in OS X, with the one exception that you generally had to tell "the system", somehow, where you'd copied the app. That is, under AmigaOS, you'd edit s:startup-sequence and put in an assign, under DOS you'd probably edit autoexec.bat and put in an entry in your set PATH= line, etc.
This latter bit of the process was, quite seriously, the only bad part of installations. Most packages came with instructions on how to do this last bit, usually in the form of running a quick program after you'd installed the app. The rest was easy. Apple solved this, at the time, even then, by including proper metadata in the operating system.
Time has passed, and now someone, somewhere, has declared this entire system, with or without the "you have to tell the OS where the app's installed" bit, fatally flawed. Why? Because if an application has a shared component, it might be installed more than once!
And is this important? Well, apparently so. You see, the additional library will take up more disk space, and more memory, and you know how short we are of both.
Now, there's room here for a little rant on how applications seem to have grown over the years disproportionally compared to (useful) functionality, but even leaving that aside, clearly memory and disk space are no longer considered the most important factors in computing right now. If I have two applications that manipulate MPEGs, and they both have a generic, half meg (let's be generous, 'cos it will not be that big), library for encoding MPEGs statically linked to them, is this a serious problem on any modern computer system? Or even any ten year old computer system?
The point is this isn't a problem. It's an inelegance. We can see, at a glance, that a fix for this issue would be a nice-to-have, but nothing more.
So what about the alternatives? Well, we now have half a dozen popular package management systems, varying from Slackware's "no dependency checking, read the documentation" installpkg system to Gentoo's "We're not just going to solve that minor issue, but we'll also solve the problem that your application is running 3-5% slower than it could do because it was compiled for the 80486" kitchen sink approach. What do these have in common? Are they elegant, trouble-free, approaches that result in applications running the way the developers intended?
Well, by and large, not really. We now have DLL hell, half the installation systems require we run around trying to find obscure packages that we didn't realise we needed from the start, the systems aren't integrated with the file system, which supposedly is the gateway to the computer, so adding complexity to the entire process of understanding what you have on your machine, and for the most part we've also limited the flexibility of users when organizing their own machines. We've also added complexity to the physical system of storing applications, creating another central set of registries etc that can break. It used to be the worst anyone could do to a PC, forcing a reinstall of everything, was overwrite the FAT. Now the disk filling up during an app install can potentially cause the same problem.
This is a fix? Because it looks to me like it's introduced more problems than it's solved. In fact, it looks like it hasn't solved any problems, it's merely added an ounce of hardware efficiency in exchange for a high degree of complexity. It's so bad that, in my e
OS/2 Warp (3) did actually take off, at least in Europe. According to Judge Jackson's Findings of Fact, IBM was forced by Microsoft to drop all marketing and bundling of the OS and Lotus Smartsuite in exchange for the support it needed to sell PCs with Windows 95.
Microsoft has commented, at the WWDC during the announcement, saying the MacBU will port Office to the new architecture. I think that's very much a vote of confidence, not something said by someone taken unawares, right? You think Apple sprung this on them? Even the Wall Street Journal knew about this a month ago.
Not that I don't hope I'm wrong, but I suspect this is why Microsoft isn't looking that upset about the switch right now.
I'm impressed. I take it you upgraded to the Oric Atmos when you decided it was time to learn Perl... ;-)
The only thing "right" about Dvorak's article is the notion of a processor change at some point in the future that involves Intel somewhere. Great crystal ball you have there Dvorak!
Dude, if you're going to post this, you at least also add speculation that Intel is going to be producing PowerPCs. That was the other thing everyone got really obsessed with yesterday before Jobs announced the move to Pentiums, with the $1000 developers box you can buy right now (additional $500 may be required to join ADC at correct level) that has a Pentium in it.