Shoot, the OLPC is just sitting there begging to be taken, and the licenses say go ahead. How hard can it be to put in a different CPU?
When your biggest, most powerful partner is Microsoft who has the power to kill you, it might be really hard to ship a Linux based system.
Unless iNTEL is so focused on high-end that it's going to take them months to build a real competitor in the ultra-low power range?
They don't have a cheap chip that fits the bill. They need to either sell a more expensive system, design a new chip, or subsidize the chips for that project. None are really good options.
Re:The best tools stay out of the way...
on
Goodbye Cruel Word
·
· Score: 1
I use Framemaker for almost everything - I find that you can export a PDF or RTF for when you need to talk to Word-bound folk.
I used to be a Framemaker guy before I diversified my skill set a bit. The 7.0 version was really, really slow to export large files to RTF, especially with big tables and images. I'm talking go take a coffee break long. I haven't used 8.0, except to play with it at a demo and I also got to talk to the development team, well sort of. They're all from India and the language barrier was an issue. They have basically convinced me not to hold out hope for the future. They couldn't even understand half of the major features I've seen requested on forums.
One surprise I had at the same conference, however, was a chat with the guys at MadCap. They're about to release a Framemaker competitor called "Blaze" which looked promising, aside from being Windows only, like Framemaker now is. (Actually it looked like every feature from Framemaker redone with a new code base and with modern features and using XML from the get go.) I think that is what has spurred Adobe to restart Framemaker development instead of telling everyone to switch to InDesign and wait a few years for them to re-implement the features and solve the crashing problems. Anyway, I'd check out Blaze when it is released or go sign up and snag a beta copy.
It seems to me that with minimal configuring, you can setup Microsoft Word 2007 to do exactly what you are asking...
Yeah, then you have a basic text editor that takes 30 seconds to start on a fast system, uses 120 Mb of RAM to open a bank page, and which randomly corrupts larger files on close so the next time you try to open them you're forced to revert to an earlier backup losing work on a regular basis. Sorry, for long documents Word is a non-starter for professional writers, not only because of the resources, but because it is unreliable.
YMMV, but splitting writing completely from layout/design has improved my work greatly.
That can certainly be effective, but I don't think it necessarily has to be that way. For example, InDesign includes a text editor mode that allows you to edit just the text or just XML in a customizable editing mode. It is a bit slow and buggy as of CS2 though. Other packages that integrate both workflows are Framemaker and Pages. I agree that writing and layout are different activities, but it is often nice to be able to switch back and forth to see how your writing or editing affects the layout and sometimes the layout influences what you want to write (references to pictures or embedded content for example).
None of the packages I've used do everything I want, and I'm still waiting for the perfect one. I do know it will be a cold day in hell when I go back to using Word:)
Re:The best tools stay out of the way...
on
Goodbye Cruel Word
·
· Score: 1
Been there, done that. TeX/LaTeX is nice and cross platform, but for real page layout, you should really try Pages...
I use LaTeX for some projects, but for the most part it is simply too out of date. Everything seems to be implemented as a hack, from inserting images, to hyperlinks, to color. It is useful for programatically laying out documents from XML or something or HTML to PDF transitions. Pages is pretty nice for small docs, but authoring/editing/laying out technical books is nicer in Framemaker. For other long works I prefer InDesign, and for catalogues or other works you want to update automatically, InDesign or Quark are both more featureful. I really wish someone would get off their butt and crete a layout/word processor that replicated all the features of Framemaker. Sadly the only people even trying (Madcap) are Windows only developers with no plan to make a Mac version and that means giving up the flexibility of OS X services.
Business is not based on good or evil but profit and loss. One should never expect business to do anything but maximize its profits. To control byuiness, one needs laws that make it profitable to do good and unprofitable to do evil. That means costs for business should include the externalities, such as production of greenhouse gases, now subsidized by government.
I thought this was pretty much a well accepted understanding of the interaction of business and the law. The responses to your post are sad. I guess I should expect as much, given how few people seem to understand that antitrust law does exactly that... attempt to make it illegal and a financial liability to undermine competition in a way that hurts consumers and progress. It is amazing how many people seem to think that companies deserve inherent rights, but should not have corresponding responsibilities.
OLPC has helped to define a market (actually, it seems more like they discovered it, the more I read), and now when someone else sets up shop next door, they cry foul.
We're not talking about someone else setting up shop next door. We're talking about a board member of the charity trying to undermine the work of that charity, for profit. Intel competing with OLPC project is fine, so long as they aren't part of the OLPC project and especially not trying to convince clients that because they're part of the project they have inside information the buyer does not.
Think of it this way: If a church has a homeless shelter, it is a good thing. If a businessman sees the chance to offer a flophouse for a few bucks a week, it is a bad thing.
No, that's not bad thing, but neither is it a good thing. In the first case the motivation is to help people (good). In the second case the motivation is to profit (not good and not evil by itself).
Either way, the homeless are off the streets at night, but because the businessman isn't doing it to get into heaven, but to make a buck or two, he's the worst kind of evil.
No, lying to people who come to the shelter and telling them the workers will rape you at night and claiming you know this because you're a board member for the charity and they should shell but a few bucks and stay at your roach infested flophouse instead. Sure there are roaches and it costs money, but surely you trust a board member at the charity to know what is going on there, right? Sorry, that is unethical.
But a profitable building is sustainable, a handout only lasts as long as the charitable continue to give.
Except many charities are sustainable, especially ones like the OLPC project where they are selling the machines and services at cost to underprivileged people and at a profit to others.
The thing that I find interesting is that Negroponte keeps pointing out all the faults with the Intel box to the press, like superior technology is always a no-brainer.
Any reasonable person who compares the two offerings with a view of what is better for children in the developing world, would choose the cheaper, better OLPC. I don't see any contest at all. That is why FUD from an Intel salesperson is so damning. They tried to compete not on merits, but by insinuating they had inside knowledge and the OLPC project was dying.
Maybe these countries have a hard time justifying a purchase that until a few months ago was vaporware.
OLPCs have had shipping demos for longer than Classmate PCs.
Maybe there's more PCs in these countries than Negroponte thinks, and these countries want to make sure their kids are able to use them.
Please, these are for small kids and they teach real, marketable skills and provide real functionality at a fraction of the price. Have you even played with an OLPC demo? It is amazing. I wish my Windows box had that much functionality for networking.
Within Intel, one part (the OLPC liaison) is pro OLPC, seeing it as a growth opportunity, while another (the field sales organization) is anti OLPC because it eats up into their potential sales. I am sure they never talked to each other. Even if they did, corporate politics and turf wars may have ensued, with sales winning this round.
This is entirely possible, but it doesn't much matter. Intel acted in bad faith and put immediate profits and hurting AMD above children's welfare. Whether some people Intel felt one way and some felt another (as I'm sure they did), Intel as an organization took action and are responsible for that.
When Negroponte made this public, it was embarrassing to Intel, and eventually the money balance tipped the scale and they withdrew from the OLPC.
More likely when they realized they were going to be forcibly expelled from a PR friendly nonprofit for unethically trying to undermine it, they decided to pull out first and try to minimize the fallout. Whatever the politics internal to Intel, it would behoove all of us to raise as big of a stink about this as possible. The more it hurts, the more likely Intel is to put in place policies to stop it from happening in the future... which should be the goal of society as a whole IMHO.
I was actually thinking of screwing AMD, whose processor was shipping in the OLPC. But yeah, the fact that they joined the project only to try to use that to try to backstab them.
Intel still sells the Classmate PC, and in the Peruvian case, the Intel machines it's trying to sell will still go to the same target audience as the OLPC units, it's not like they suddenly hate kids!
Intel put profits above the kids and they did so in a dishonest way. They tried to get Peru to buy classmate PCs instead of XO machines. Since Classmate cost twice as much, that would be getting PCs to half as many kids. Further, the classmate PC is just a really low end Windows box, not a system designed from scratch by a lot of very smart people to be perfect for kids to use for learning. I mean the Classmate PC doesn't even have a zeroconf implementation, while XO laptops use it to automatically create a mesh network and allow kids to collaborate using chat, sharing pictures, collaborating on musical compositions and games and school projects. How can anyone think a Classmate PC compares favorably, even at twice the price? This was Intel trying to make money and screw a competitor; ignoring the fact that they were also giving fewer kids and inferior solution.
And to top all this off, they strongly implied they had insider knowledge because of their role as a member of the OLPC board and that the project was going under and the XO laptops would not be supported or as useful in the future. The people involved with this decision making are slime. No they don't hate kids, that would require human emotions. They just completely disregarded kids and ethics in their ruthless quest for a few more dollars.
Now regardless of who's making the machines and what OS, CPU blah blah they have in them, it's good that this device class actually exists and it's great that more people around the world get a chance to use devices that we take forgranted. OLPC and the Classmate are both doing a good job, and I'd love to see other devices like the EEE PC tailored towards developing nations in the near future.
I think this is completely untrue. The Classmate PC is an inferior grade Windows machine. The OLPC is an innovative new device targeted specifically with an OS, interface, hardware, services, support system, and internet service. It's the whole package done right by people who care. Have you even looked at the demos or VMs of it? It is technologically superior to Windows for this purpose, by a huge degree. In principal, competition in this area is just fine, but it has to be competition on a level playing field based upon the merits of the device, not a big corporation lying to try foist off a drastically inferior solution in the hopes that they can influence politicians to make what is clearly the wrong choice. When advocating competition, first check to see if Microsoft, Intel, the RIAA, the MPAA, or Bell is involved. These are all companies with long histories of being repeat offenders using anticompetitive tactics to undermine competition in the marketplace.
It's like if a construction company and Habitat for Humanity...
Actually, Mr. Negropnte himself had a better analogy, "They played another dirty trick in Peru," he said. "It's a little bit like McDonald's competing with the World Food Program."
I think that is a better analogy because the OLPC project designed an ideal system for these children with lots of cool, new features not available on regular computers. Intel didn't steal the plans, they just made a regular, really low end PC running windows. It doesn't have the cool software, doesn't auto-discover other machines and create a mesh network and allow kids to network applications together. It is really unsuited to the task, just as the food provided by McDonald's is largely unsuited to meeting the basic nutritional needs of children when compared to the offerings from the World Food Program.
Intel could have met their obligations and tried to pitch their new mobile, low power chipset for the next version of the OLPC. Instead they tried to be unethical and tried to poison deals with their competitor through deception, ignoring what is best for underprivileged children. This actually makes me more likely to buy an AMD processor for my next upgrade to my home server, but since this will not get any real press it will still probably make Intel money in the long run and they'll probably do it again next time they get a chance.
You're a waste of time because it became obvious you were bashing when it became apparent you were making out right lies.
Sigh. First, it's "outright." Second, I wasn't bashing or lying. I'm a Linux advocate and Linux on the desktop user. Pointing out real problems with the direction of Linux and its failure to move forward in ways that are not incremental, backwards compatible improvements on existing technologies is important for anyone interested in correcting or working around that problem.
Oh, I am certainly aware of issues on the Linux desktop because I actually use it.
So do I but I also use OS X and Windows so I can actually make educated points about where one is ahead of the other. You don't even seem to know what system services are and I notice you completely dropped that line of conversation when I pointed out you didn't know what you were talking about.
But you, you don't even give proper examples of issues...
I gave numerous examples, from the failure to standardize on a package format, to failure to copy major improvements from other vendors, to missing technologies. OpenStep, ZeroConf, system services, MACLs... how many more do you want?
...which ends up in you refuting my points with lies.
It's nice that you fail to specify anything I said which you specifically claim is a lie. It's nice that you ignore point after point from my previous posts. I say again, 8 years ago I could add a third party spelling checker or language translation or bibliography formatter or online reference lookup service to OS X that would instantly work in nearly all my programs. For 8 years I've been waiting for any Linux distro to add that same feature, with little hope. No KParts aren't even close. This is a fundamental technology that would require a real change and would mostly help Linux as a desktop OS. When should I be expecting it? I'm thinking never at this point and I explained my best guess as to why Linux is so slow to copy or adopt new features, especially from Apple. I cited numerous technologies OS X has integrated from Linux and other OS's showing how it is advancing and numerous technologies Linux has failed to copy from OS X.
I've been fairly patient here and gone out of my way to explain things I don't feel I should have to if you actually had regularly used both Linux and OS X as desktop systems for real work. I haven't seen much of anything back except empty rhetoric and misguided claims based on your failure to understand the features I reference. Then you turn to inflammatory, personal attacks. Fine. Put up or shut up. Show me the major advances in Linux as a desktop that aren't useful for Linux as a server and which actually affect my ability to get work done. I'm talking features, not prettier graphics or more stability. If I don't get a reply back with actual meat in it, I'm not going to bother to respond.
Yeah, cover your ears and yell, "la la la" while you're at it. Those who deny and refuse to listen to what their weaknesses are and try to improve them are doomed to failure. Let's hope Linux on the desktop is not dragged down by people trying to pretend there is nothing wrong.
Why? The thinkpad has all the three buttons, which work perfectly well.
Yeah, but more slowly, than one button with chording or 2/3 finger taps.
Like command-line history not working properly in Python
Python is a language. How can CLI history "not work?" You can run whatever shell/editor/etc. you feel like.
windows not maximizing to the whole of desktop
If you don't like this, complain to the application developer. Unlike Windows, OS X gives the developer the option as to which behavior happens, since for a lot of apps, maximizing results in a lot of empty whitespace and unlike Windows, you can actually use multiple applications at once usefully. When I maximize a browser window, I'd rather it not cover up my chat window with blank space to the side of the Web content I'm looking at.
F11 not working on Mozilla (yes, I know it's a conflict with standard Mac key), etc.
"System Preferences: Keyboard & Mouse: Keyboard Shortcuts" you can map any key or combo to any action in a given application or universally.
Well, the standard Python coming with the system was just bare-bones
You mean the dev tools I assume? You'll be happy to know the latest version of OS X comes with complete Python bridge and a nicer set of tools, but as a developer, most people want to download some specific tools they prefer anyway. I don't see this as an issue. OS X has some of the best Python support of any OS I can think ofand this is a one time config, since it will migrate to your next machine with zero effort.
The same process takes me a couple of hours max, including transferring all of my 20G home dir. Yes, for that reason, I tend to stick to defaults as much as I can, and to limit configuration choices to package selection. But the nice point about Linux is that for me, currently this resuls in a system that is quite usable.
We actually did some testing on this because on of our dev teams was looking at mandating macs for new developers because they were upset with how long it was taking users to get a functional Linux system installed and configured. The average time for new developers was 9 billable hours at the beginning, then another 6 hours over the first week, then another 14 hours in the first month assigned to setup and configuration time. In the end, we left the choice up to the individual, but a lot of us were pretty appalled. In any case, Linux is way behind in that regard. A couple hours of work is a lot more than a couple of clicks and then no need for the user to do anything. I love that my nicely customized and decked out linux install not goes with me from machine to machine, including all my accounts, tools, files, settings, etc. It sure saves me a lot more than a couple of hours each time.
Well this is the catch - *does* it provide useful new functionality to the average Linux user?
I know a lot of former Linux on the desktop users who are now OS X users, partly because it makes configuration so easy.
My experience of auto-discovery protocols is that they are not especially reliable and I would much prefer to just set stuff up manually.
Wow you must never have used Bonjour on a Mac. First, it always works. I've never seen it fail to find the printers or other services offered. You can literally just plug a printer into the network, or plug it into a Mac and click the "share" button and every mac on the LAN lists it as an option when you go to print, within 5 seconds. Second, how do "manually configure" your chat client to be able to hat with all the people at a conference you just went to? The last time I went to a conference, I opened up iChat and was able to have a Jabber IM conversation with every other Mac user on the local wireless. It is a great way to find things out and meet people. It's also great at the coffee shop and "manual configuration" for that would be onerous. And chat isn't the only application. At work there are shared streaming music services from people using iTunes and shared collaboration for text editing (SubEthaEdit rules for pair programming!) Have you seen all the cool applications that are collaborative for kids using the OLPC? Why can't regular Linux distros offer the same level of collaboration fro grown ups?
s an example, plugging 2 Windows machines into the same bit of network cable and giving them an IP address is supposed to let you access their network shares, but half the time one machine can't even find the other one (in my very limited Windows experience).
Yeah, and MS has refused to implement ZeroConf instead using their own, broken, proprietary crap. That doesn't mean it can't be done right or that it hasn't. I'm just tired of waiting for Kubuntu to catch up. I love Kopete. It makes setting up accounts simple and is more customizable than iChat, but I spend the time to manually configure accounts in iChat just so I can use local chat without having to go up to everyone I see and ask them to register a Jabber account.
In this case, the plaintiff is alleging that Apple violated the Sherman Antitrust Act which requires that a monopoly or cartel must first exist.
This is untrue. According to the Sherman antitrust act the company or cartel in question need only have what they call "monopoly influence" in a market, which is later defined as having undue influence due to market share, such that they can undermine free trade.
So legally, the plaintiff must prove that Apple has a monopoly in some way. Let's say that for the sake of argument that Apple has a monopoly. It would however only be in iPods/iTunes. It would not be in digital music players or digital media.
They need to prove Apple has the ability to undermine free trade in a market using their market share and it is by market, not product. They're claiming Apple has such influence in "online video market," "online music market," "hard-drive based music player market," and the "Flash-based music player market." The courts may or may not accept those market definitions, but that is what they're being asked to rule on.
Well that depends on your definition of non-trivial.
Absolutely anything that makes it easier for people to use Apple's devices instead of a competitors that are result of their market share are all that matters according to the law.
If I understand your statement correctly, Apple's large market share makes them somehow responsible for any 3rd party attachment I loaded work with any of their competitors?
No, it doesn't make them responsible for the third party attachments. It makes them responsible for not taking any action which capitalizes on what third parties are doing as a result of their large market share, in order to gain share in a different market (like online video sales).
Even if we accept that somehow Apple is a monopoly in violation of the Sherman Antitrust Act, Apple would be in violation if they prevented their competitors from working 3rd parties not if they did nothing.
You're fundamentally failing to understand our antitrust law. No company is in violation of the Sherman antitrust act for being a monopoly. They're in violation if they're a monopoly which takes any action to leverage that monopoly into another market. There are two things to consider:
Do they have monopoly influence? Do they have the power undermine free trade?
If they do, are they leveraging that monopoly to gain share in another market.
The behavior of third parties speaks to number 1. If people can't as easily go with a different vendor for their next device, this speaks to how much power Apple has, not to if they are abusing that power.
If they have monopoly influence, they are clearly abusing that influence. They are tying together their online store and their software with their player. Do you really think iTunes would be as popular if it did not ship with iPods? Do you really think they're sell as many songs from the iTunes store, if it was not the only store accessible with the software that ships with iPods? It is a clear cut case of tying, so the only question is "do they have enough influence to undermine the market" and that is the point in debate among people who have a clue about US or EU antitrust law. I'll say again, the ability to move music to a different device as easily as to a new iPod and the availability of third party add ons, both speak to how much power Apple has and those will be things the courts look at to make their determination. Apple is liable, under the law for tying, not for what third parties do, but what third parties do helps determine if what Apple has done (tying products) undermines the market and thus is illegal.
Even assuming that Apple is monopoly, unbundling iTunes does not alleviate the situation at all.
Yes it does. It removes Apple's ability to undermine the "music jukebox software" mark
Maybe you tend to work with more incompetent/lazy people than I do. I know at least one of these engineers spent a month writing a new kernel module and security application for OS X for his own personal use and which he open sourced. Four of these engineers (all mac users) were poached from us by Google and now have one day a week to work on a personal project. A large number of the others contribute to one or more OSS projects.
When a large change is needed, the distributions that target those uses implement it and use it.
No, they don't. When was the last time Linux fundamentally changed in some way? Being compatible with the other distributions is more important than any one new feature, so for the most part new features that require a major change don't happen.
I've seen distributions take their own way all the time. Anyone who runs a distribution can do what they want with that distribution, if you don't have enough influence. You can fork off and start your own. That is the beauty of it.
Forking Linux for a desktop version is a doomed venture and everyone knows it. If you can't get multiple, major distros on board you can't make a big change.
Believe it or not, there are Linux distributions that do exactly that too.
No they don't because they still have to worry about being compatible with the other major distributions. You can't just migrate Ubuntu to OpenStep because all the packages Redhat makes will still be RPMs and thus you end up managing multiple package formats (like people do now) and it is actually worse than it was before.
It would take a new LSB proposal and once accepted the distributions would add support for it.
Ha. LSB, what a joke. When was the last time there was an accepted LSB proposal to make Linux work better as a desktop instead of a server. Oh yeah, never.
*shrugs* They will get what they want, and the Linux desktop users will get what they want. This is why we have distributions.
You miss the point. They enjoy having OpenStep style packages on the desktop because they work better. They will fight, however, to make it harder for Linux on the desktop to get that same feature because they want Linux to be a server and it is just bloat for a server. They are actively holding Linux on the desktop back.
Ironically I don't see OS X keeping up with Kubuntu in the long term. Already Apple can't seem to follow their own HiG philosophy.
I just happen to have worked as a user interface designer and usability tester for a while, and got paid to go to conferences and take classes on the subject. Yeah, Apple makes some interface mistakes, but they're still way, way ahead of Kubuntu in that regard.
As for fundamental new features - What fundamental features could they add?
System Services would be one. Drag and drop packages would be another. Built in support for MACLs with mandatory MACLs for all programs would be one. A package format that is portable, may include source, includes a link for updates for software that isn't in a repository, and usable support for commercial licensing would be nice. A trust/signing framework with a protocol for checking the references for a given program would be a step forward. A proper method of migrating your system to new hardware would be a plus. Need I go on?
The desktop environments (Gnome, KDE) have such capability already. There just isn't a grammar checker available yet. Things like spell checkers are universal to both DEs though and you can even choose which spell checker you want to use.
Really. Where, exactly do I put the system service file that will allow me to translate my Kopete chats and Konquerer forms between chinese and English automatically? OS X only got a standard grammar checker in the latest
However, the feel just simply can not make up for design choices (like the menu bar at the top of the screen) that kept me guessing every step of the way. Basics - like which menu to use for what, how to really open certain applications, and how to close or minimize and even maximize them properly for starters.
Imagine for a moment you've never used a computer before and don't have a bunch of preconceptions and learned habits about how these features would be implemented. Imagine you're designing a new OS from scratch and have to decide how to design the GUI? Objectively, which is better, having one menu at the top, or one menu per window, with a single app maximized to fill the screen most of the time. With the menu always at the top, Fitt's law makes it an infinitely large target in one of the two dimensions and your testing shows that users can get to it faster and over time become faster yet because the menus are always in the same place instead of a different place depending on the window position. It also enables developers to make applications with menus, but no windows, when no window is needed. From a usability and design perspective, the OS X way is objectively better.
Now on to the minimize/maximize. What makes more sense a button that always makes an application fill the whole screen, or a button that the developers can set to fill the whole screen or grow just large enough to fit all the content, depending on the type of application and what the developer thinks is best? Surely giving developers the option is better no? The only reason it isn't better is if you are trained to always use every application in fullscreen mode because having multiple applications displayed simultaneously is too hard to use because of poor design choices (like Windows). Otherwise, having a window grow large enough to fit a Web page, but not automatically cover your chat client with blank window, makes a lot more sense for some types of windows.
Now imagine you build this OS and market it and years later someone else makes a different OS that does things differently. Should you revise your choices and make them objectively worse, pissing of all your current customers, simply to be more like the other OS and make things easier on people who want to switch from that OS?
Maybe - if I had to spend a whole year with this, I could learn to accept it over time... Mind you, this is personal taste.
Personal preference counts for a lot and if you're stuck in your ways, sometimes it is best to stick with what you know or a close clone thereof. It just annoys me, a person who spent a lot of time learning about usability, user interfaces, and human/computer interaction, that things Apple did correctly from a scientific perspective are so often cited as faults and reasons to avoid their products. I mean Apple has made numerous UI mistakes that no one ever mentions, but they insist on bringing up things they've trained themselves to use on a poorly designed UI, as faults for a correctly designed one.
Since Apple went intel the costs have been pretty much on par, sometimes a little cheaper and sometimes a little more, compared to competing systems from Dell and etc.
Actually, he said he went with a Thinkpad, which runs about 15-20% cheaper than the closest comparable Mac hardware (without any special discounts for students or the like). Of course it also has 2-3 times the hardware failure rate of an Apple laptop according to Consumer Reports.
Bonjour for services hasn't taken off in a big way on Linux (or Windows outside of Adobe Products/iTunes for that matter).
I understand lack of support on Windows. They have a competing, proprietary protocol and they want to use it as their normal lock-in, but why hasn't anyone picked up the ball yet on Linux? An open standard with multiple FOSS implementations to harvest that provides useful new functionality seems like something that should have been implemented a long time ago as a standard service, with an easy way for developers to add it into their application.
By the way I think someone said they might work on a Kopete bonjour plugin a few weeks ago.
That would be awesome and stop my GF complaining that I'm not "on IM" when I'm tinkering with my all Kubuntu box.
I'm also a little sad that OSX has dropped default support for printers annouced over CUPS broadcast but thems the breaks.
I didn't hear about that, but my Leopard machine still sees the old TCP/IP printer on my LAN and I thought that was just CUPS. I'll have to investigate.
By "Linux laptop", I mean a Thinkpad (T60p), which runs Ubuntu.
My old laptop ran Kubuntu just fine, and my new one does too, albeit in a VM now, since that is more convenient for me.
I'm sorry, I need my right and middle button even when I'm not using an external mouse.
You might want to consider breaking that habit. I was skeptical at first, but the usability studies are pretty clear on it and my personal experience baks them up. When using an external mouse, your hand is off the keyboard and it is faster for advanced users to have multiple buttons. When using a trackpad your hands remain on the keyboard, so having to move both to hit multiple buttons wile still moving the cursor is actually slower than just using one button in combination with either a modifier key near your other hand (depending on your handedness) or using the two or three finger tap method. Using a multi-button trackpad is easier to learn, but in the long run it is slower for advanced users because you have to take your fingers off the keys more.
What actually drove me nuts was the little things, like not having normal buttons for PgUp/PgDown, Delete etc.
I actually like the Thinkpad keyboard better too, both for the arrangement of keys and the responsiveness.
Also, many of the Unix-land progs had subtle annoyances like confusing modifier keys etc.
Really? This has never been a problem for me. Like what?
And for drag and drop install/uninstall - ha! 99.9% of what I need is available from Apt, and it's so much simpler than drag-and-drop install.
Ahh, see here's a problem with people who try to use OS X. They don't look for the best tool, they just stick with what they were using on Linux, which is often a crappy port because no one uses it; there being a better, native application. As for Apt, there are better package managers for OS X, including ones that have a working graphical interface as well as a CLI one and which handle OS X native software in addition to ports.
Now in Linux, if something is missing and I want to compile from source, I stand a good chance of it just working. In Mac, no.
This is true some of the time, although I notice more and more it going the other way, especially for security tools. This is simply a function of what the develop who made the tool was using, since things tend to be slightly broken everywhere else. With so many developers switching to OS X, I find a lot of tools won't compile easily under Linux, but will in OS X, but then again I'm in the security field which may not reflect the greater software ecosystem.
For instance, I've spent a day trying to find a python install for Mac that would include all the packages I need for my project, while in Linux, well, all of them are just there.
That's odd. I know a lot, and I mean a lot of Python programmers running OS X and no one ever mentioned a problem. Typing "OS X Python Tools" into Google did not pull up a bunch of sources for what you need? Going to Python.org, undefined.org, or macpython.com didn't have them?
I really, really tried to configure the Mac to my liking, and btw.
This might be your problem. Most all of the people I've talked to who actually tried OS X and did not like it are what I like to call "crusty old fellas" who are very set in their ways and don't want to try a new OS, they just want something that will be Linux/Solaris/IRIX and let them use exactly the same workflows and exactly the same software as their favorite UNIX flavor. If you're trying to make OS X into Linux, it will never be as good at being Linux as Linux. If you're not willing to adapt your workflows and learn new, sometimes better ways to do things, why try another OS in the first place?
I see many people around me using them just fine - but not for real work...
To make long story short, I returned it in disgust and got a Linux laptop instead. So far, all is fine.
Okay you were disgusted by... something and you got a "Linux laptop" by which I assume you mean you installed Linux on the MacBook, or was there some reason you couldn't get that to work?
Apple might be good for a grandma or for a graphic designer, but for a programmer it's an annoyance.
I know a hundred or so programmers who have switched in the last couple of years and one, just one who switched back. That one spent a lot of his free time developing a Linux for the desktop distribution and even dual booting could not find all the drivers he needed for all his peripherals and did not like switching between different UIs all the time.
I'm not saying your opinions aren't valid, but come on, give us some meat here. At least tell people why you felt OS X was an annoyance for you, as a programmer. I'll tell you what I don't like about it. It doesn't ship with a good package manager so developers don't code for one. It has a really polished UI, but only for one workflow per activity. The default applications are featureful, but not standards compliant enough and don't do a good enough job of interacting with more advanced functions from other programs.
On the other hand, Linux lacks a simple hardware upgrade mechanism, does not have drag and drop install/uninstall or portable application packages, lacks system services, lacks ubiquitous support for ZeroConf, and has a lot of user interface problems and spotty user testing.
I think you are actually saying "Linux is hurting Linux Desktop Development".
Well, sort of. The Linux development methodology does hurt it as a desktop in some ways, but the brain drain as people move to OS X hurts it more. If there were only Linux and Windows, a lot of people who now use OS X would choose Linux and devote time, money, and effort into fixing problems with Linux as a desktop and pushing for change. Instead, they mostly move to OS X for their desktop and actually argue against those changes because it is no longer in their best interests since they use Linux only as a server. This is nothing intentional on the part of Apple, but it is a real trend that I've noticed.
Disclaimer: I have been using OS X and Ubuntu on *daily support basis*, so I know what I am talking about.
I run OS X, Kubuntu, and WinXP daily.
OS X is nice, but it is black box, and in the end you will loath it.
I've been running it since version 10.0. It was pretty rough at first, but has really stabilized and continued to advance at a good pace. In fact, it is my OS of choice for applications, all other factors being equal. I don't loath it at all. More generically, I know literally a hundred or more engineers who have switched from Linux or a BSD in the last few years and only one who switched back and these are users with a choice of what they want to use for their daily work developing software for Linux and BSD based appliances.
As any other OS, it has lot of bugs, including VERY annoying ones which you can't fix even with having support contract with Apple and Adobe.
I've had my fair share of feature requests go unfulfilled, but all the actual bugs I filed in OS X have been fixed and I don't have a contract with Apple. Adobe is another story, but that is unrelated to the platform. I have the same problem with their Windows and Linux software.
However, Apple and OS X strenght[sic] is integration.
I'd say their strength is in their ability to make major changes that break things for the sake of advancing their platform and scrupulous user testing.
I just wish they would not be so annoyingly similar to Microsoft as they were in last year - all standard stuff, supporting OOXML, closing DAAP, etc.
They have read only support for OOXML, while TextEdit reads and writes ODF, seems pretty useful to me. They never closed DAAP because they never opened it in the first place. Some people reverse engineered it and things broke when it changed.
For Linux, problems are two - user base and apps.
I'd argue both of those are situational items that are contributed to by flaws in Linux. The first, is contributed to by Linux's commercial software unfriendly package management. Linux distros are varied and don't all use the same package manager or libraries, don't have support for software registration or software updates from a Website maintained by the distributor, who won't put it in a repository for technical and legal reasons. The second problem, user base, is partly because Linux does not do a very good job of catering to normal users, maintaining it's focus on current users who are mostly power users and CLI fans.
I'd also argue that while Linux is technologically ahead in a few ways, it is technologically behind in a lot of ways that matter to normal desktop users. There is no drag and drop package install/uninstall. Installed applications aren't easily portable. There are no OS X style system services. ZeroConf has not been ubiquitously integrated into standard applications. All of these (along with a good expose clone) are things I miss while using Linux.
However, I think it is not the end of desktop of Linux. For me, it's only now getting in shape that I have no shame to show to others.
I've been a desktop Linux user for years, but in my opinion it is falling behind rather than catching up. I've also seen some serious brain drain as Linux desktop user/developers move to OS X and stop contributing to Linux desktop efforts. There is hope, but OS X does seem to be a serious detriment to desktop Linux, even if it is not intentional on the part of Apple.
If there's one thing I dislike, it's "comparative" studies - inconclusive, all of them.
This article isn't really a comparative study, but the point of comparative studies isn't just to arbitrarily decide which is best, but to explore the ways in which each offering is better than others and thus inform people who care about those particulars.
This is mainly why I dislike the title "is apple 'killing' Linux on the desktop". If anything, Linux is GAINING ground on apple in recent years - thanks to distros such as Gnome and Ubuntu - so why the hyperactive titles? One would think/. too intelligent to fall for such techno-queeny cruft.
Yeah, the title is a bit overblown, but that is par for the course. I do think OS X as an offering is hurting Linux for the desktop by stealing developers away and allowing them to refocus their extra-curricular programming on things other than desktop Linux (like Linux as a server or applications on OS X).
Slashdot Mirror
When your biggest, most powerful partner is Microsoft who has the power to kill you, it might be really hard to ship a Linux based system.
Unless iNTEL is so focused on high-end that it's going to take them months to build a real competitor in the ultra-low power range?They don't have a cheap chip that fits the bill. They need to either sell a more expensive system, design a new chip, or subsidize the chips for that project. None are really good options.
I used to be a Framemaker guy before I diversified my skill set a bit. The 7.0 version was really, really slow to export large files to RTF, especially with big tables and images. I'm talking go take a coffee break long. I haven't used 8.0, except to play with it at a demo and I also got to talk to the development team, well sort of. They're all from India and the language barrier was an issue. They have basically convinced me not to hold out hope for the future. They couldn't even understand half of the major features I've seen requested on forums.
One surprise I had at the same conference, however, was a chat with the guys at MadCap. They're about to release a Framemaker competitor called "Blaze" which looked promising, aside from being Windows only, like Framemaker now is. (Actually it looked like every feature from Framemaker redone with a new code base and with modern features and using XML from the get go.) I think that is what has spurred Adobe to restart Framemaker development instead of telling everyone to switch to InDesign and wait a few years for them to re-implement the features and solve the crashing problems. Anyway, I'd check out Blaze when it is released or go sign up and snag a beta copy.
Yeah, then you have a basic text editor that takes 30 seconds to start on a fast system, uses 120 Mb of RAM to open a bank page, and which randomly corrupts larger files on close so the next time you try to open them you're forced to revert to an earlier backup losing work on a regular basis. Sorry, for long documents Word is a non-starter for professional writers, not only because of the resources, but because it is unreliable.
That can certainly be effective, but I don't think it necessarily has to be that way. For example, InDesign includes a text editor mode that allows you to edit just the text or just XML in a customizable editing mode. It is a bit slow and buggy as of CS2 though. Other packages that integrate both workflows are Framemaker and Pages. I agree that writing and layout are different activities, but it is often nice to be able to switch back and forth to see how your writing or editing affects the layout and sometimes the layout influences what you want to write (references to pictures or embedded content for example).
None of the packages I've used do everything I want, and I'm still waiting for the perfect one. I do know it will be a cold day in hell when I go back to using Word :)
I use LaTeX for some projects, but for the most part it is simply too out of date. Everything seems to be implemented as a hack, from inserting images, to hyperlinks, to color. It is useful for programatically laying out documents from XML or something or HTML to PDF transitions. Pages is pretty nice for small docs, but authoring/editing/laying out technical books is nicer in Framemaker. For other long works I prefer InDesign, and for catalogues or other works you want to update automatically, InDesign or Quark are both more featureful. I really wish someone would get off their butt and crete a layout/word processor that replicated all the features of Framemaker. Sadly the only people even trying (Madcap) are Windows only developers with no plan to make a Mac version and that means giving up the flexibility of OS X services.
I thought this was pretty much a well accepted understanding of the interaction of business and the law. The responses to your post are sad. I guess I should expect as much, given how few people seem to understand that antitrust law does exactly that... attempt to make it illegal and a financial liability to undermine competition in a way that hurts consumers and progress. It is amazing how many people seem to think that companies deserve inherent rights, but should not have corresponding responsibilities.
We're not talking about someone else setting up shop next door. We're talking about a board member of the charity trying to undermine the work of that charity, for profit. Intel competing with OLPC project is fine, so long as they aren't part of the OLPC project and especially not trying to convince clients that because they're part of the project they have inside information the buyer does not.
Think of it this way: If a church has a homeless shelter, it is a good thing. If a businessman sees the chance to offer a flophouse for a few bucks a week, it is a bad thing.No, that's not bad thing, but neither is it a good thing. In the first case the motivation is to help people (good). In the second case the motivation is to profit (not good and not evil by itself).
Either way, the homeless are off the streets at night, but because the businessman isn't doing it to get into heaven, but to make a buck or two, he's the worst kind of evil.No, lying to people who come to the shelter and telling them the workers will rape you at night and claiming you know this because you're a board member for the charity and they should shell but a few bucks and stay at your roach infested flophouse instead. Sure there are roaches and it costs money, but surely you trust a board member at the charity to know what is going on there, right? Sorry, that is unethical.
But a profitable building is sustainable, a handout only lasts as long as the charitable continue to give.Except many charities are sustainable, especially ones like the OLPC project where they are selling the machines and services at cost to underprivileged people and at a profit to others.
The thing that I find interesting is that Negroponte keeps pointing out all the faults with the Intel box to the press, like superior technology is always a no-brainer.Any reasonable person who compares the two offerings with a view of what is better for children in the developing world, would choose the cheaper, better OLPC. I don't see any contest at all. That is why FUD from an Intel salesperson is so damning. They tried to compete not on merits, but by insinuating they had inside knowledge and the OLPC project was dying.
Maybe these countries have a hard time justifying a purchase that until a few months ago was vaporware.OLPCs have had shipping demos for longer than Classmate PCs.
Maybe there's more PCs in these countries than Negroponte thinks, and these countries want to make sure their kids are able to use them.Please, these are for small kids and they teach real, marketable skills and provide real functionality at a fraction of the price. Have you even played with an OLPC demo? It is amazing. I wish my Windows box had that much functionality for networking.
This is entirely possible, but it doesn't much matter. Intel acted in bad faith and put immediate profits and hurting AMD above children's welfare. Whether some people Intel felt one way and some felt another (as I'm sure they did), Intel as an organization took action and are responsible for that.
When Negroponte made this public, it was embarrassing to Intel, and eventually the money balance tipped the scale and they withdrew from the OLPC.More likely when they realized they were going to be forcibly expelled from a PR friendly nonprofit for unethically trying to undermine it, they decided to pull out first and try to minimize the fallout. Whatever the politics internal to Intel, it would behoove all of us to raise as big of a stink about this as possible. The more it hurts, the more likely Intel is to put in place policies to stop it from happening in the future... which should be the goal of society as a whole IMHO.
I was actually thinking of screwing AMD, whose processor was shipping in the OLPC. But yeah, the fact that they joined the project only to try to use that to try to backstab them.
It seems overly generous to Intel in my opinion.
Intel still sells the Classmate PC, and in the Peruvian case, the Intel machines it's trying to sell will still go to the same target audience as the OLPC units, it's not like they suddenly hate kids!Intel put profits above the kids and they did so in a dishonest way. They tried to get Peru to buy classmate PCs instead of XO machines. Since Classmate cost twice as much, that would be getting PCs to half as many kids. Further, the classmate PC is just a really low end Windows box, not a system designed from scratch by a lot of very smart people to be perfect for kids to use for learning. I mean the Classmate PC doesn't even have a zeroconf implementation, while XO laptops use it to automatically create a mesh network and allow kids to collaborate using chat, sharing pictures, collaborating on musical compositions and games and school projects. How can anyone think a Classmate PC compares favorably, even at twice the price? This was Intel trying to make money and screw a competitor; ignoring the fact that they were also giving fewer kids and inferior solution.
And to top all this off, they strongly implied they had insider knowledge because of their role as a member of the OLPC board and that the project was going under and the XO laptops would not be supported or as useful in the future. The people involved with this decision making are slime. No they don't hate kids, that would require human emotions. They just completely disregarded kids and ethics in their ruthless quest for a few more dollars.
Now regardless of who's making the machines and what OS, CPU blah blah they have in them, it's good that this device class actually exists and it's great that more people around the world get a chance to use devices that we take forgranted. OLPC and the Classmate are both doing a good job, and I'd love to see other devices like the EEE PC tailored towards developing nations in the near future.I think this is completely untrue. The Classmate PC is an inferior grade Windows machine. The OLPC is an innovative new device targeted specifically with an OS, interface, hardware, services, support system, and internet service. It's the whole package done right by people who care. Have you even looked at the demos or VMs of it? It is technologically superior to Windows for this purpose, by a huge degree. In principal, competition in this area is just fine, but it has to be competition on a level playing field based upon the merits of the device, not a big corporation lying to try foist off a drastically inferior solution in the hopes that they can influence politicians to make what is clearly the wrong choice. When advocating competition, first check to see if Microsoft, Intel, the RIAA, the MPAA, or Bell is involved. These are all companies with long histories of being repeat offenders using anticompetitive tactics to undermine competition in the marketplace.
Actually, Mr. Negropnte himself had a better analogy, "They played another dirty trick in Peru," he said. "It's a little bit like McDonald's competing with the World Food Program."
I think that is a better analogy because the OLPC project designed an ideal system for these children with lots of cool, new features not available on regular computers. Intel didn't steal the plans, they just made a regular, really low end PC running windows. It doesn't have the cool software, doesn't auto-discover other machines and create a mesh network and allow kids to network applications together. It is really unsuited to the task, just as the food provided by McDonald's is largely unsuited to meeting the basic nutritional needs of children when compared to the offerings from the World Food Program.
Intel could have met their obligations and tried to pitch their new mobile, low power chipset for the next version of the OLPC. Instead they tried to be unethical and tried to poison deals with their competitor through deception, ignoring what is best for underprivileged children. This actually makes me more likely to buy an AMD processor for my next upgrade to my home server, but since this will not get any real press it will still probably make Intel money in the long run and they'll probably do it again next time they get a chance.
Sigh. First, it's "outright." Second, I wasn't bashing or lying. I'm a Linux advocate and Linux on the desktop user. Pointing out real problems with the direction of Linux and its failure to move forward in ways that are not incremental, backwards compatible improvements on existing technologies is important for anyone interested in correcting or working around that problem.
Oh, I am certainly aware of issues on the Linux desktop because I actually use it.So do I but I also use OS X and Windows so I can actually make educated points about where one is ahead of the other. You don't even seem to know what system services are and I notice you completely dropped that line of conversation when I pointed out you didn't know what you were talking about.
But you, you don't even give proper examples of issues...I gave numerous examples, from the failure to standardize on a package format, to failure to copy major improvements from other vendors, to missing technologies. OpenStep, ZeroConf, system services, MACLs... how many more do you want?
...which ends up in you refuting my points with lies.It's nice that you fail to specify anything I said which you specifically claim is a lie. It's nice that you ignore point after point from my previous posts. I say again, 8 years ago I could add a third party spelling checker or language translation or bibliography formatter or online reference lookup service to OS X that would instantly work in nearly all my programs. For 8 years I've been waiting for any Linux distro to add that same feature, with little hope. No KParts aren't even close. This is a fundamental technology that would require a real change and would mostly help Linux as a desktop OS. When should I be expecting it? I'm thinking never at this point and I explained my best guess as to why Linux is so slow to copy or adopt new features, especially from Apple. I cited numerous technologies OS X has integrated from Linux and other OS's showing how it is advancing and numerous technologies Linux has failed to copy from OS X.
I've been fairly patient here and gone out of my way to explain things I don't feel I should have to if you actually had regularly used both Linux and OS X as desktop systems for real work. I haven't seen much of anything back except empty rhetoric and misguided claims based on your failure to understand the features I reference. Then you turn to inflammatory, personal attacks. Fine. Put up or shut up. Show me the major advances in Linux as a desktop that aren't useful for Linux as a server and which actually affect my ability to get work done. I'm talking features, not prettier graphics or more stability. If I don't get a reply back with actual meat in it, I'm not going to bother to respond.
Yeah, cover your ears and yell, "la la la" while you're at it. Those who deny and refuse to listen to what their weaknesses are and try to improve them are doomed to failure. Let's hope Linux on the desktop is not dragged down by people trying to pretend there is nothing wrong.
Yeah, but more slowly, than one button with chording or 2/3 finger taps.
Like command-line history not working properly in PythonPython is a language. How can CLI history "not work?" You can run whatever shell/editor/etc. you feel like.
windows not maximizing to the whole of desktopIf you don't like this, complain to the application developer. Unlike Windows, OS X gives the developer the option as to which behavior happens, since for a lot of apps, maximizing results in a lot of empty whitespace and unlike Windows, you can actually use multiple applications at once usefully. When I maximize a browser window, I'd rather it not cover up my chat window with blank space to the side of the Web content I'm looking at.
F11 not working on Mozilla (yes, I know it's a conflict with standard Mac key), etc."System Preferences: Keyboard & Mouse: Keyboard Shortcuts" you can map any key or combo to any action in a given application or universally.
Well, the standard Python coming with the system was just bare-bonesYou mean the dev tools I assume? You'll be happy to know the latest version of OS X comes with complete Python bridge and a nicer set of tools, but as a developer, most people want to download some specific tools they prefer anyway. I don't see this as an issue. OS X has some of the best Python support of any OS I can think ofand this is a one time config, since it will migrate to your next machine with zero effort.
The same process takes me a couple of hours max, including transferring all of my 20G home dir. Yes, for that reason, I tend to stick to defaults as much as I can, and to limit configuration choices to package selection. But the nice point about Linux is that for me, currently this resuls in a system that is quite usable.We actually did some testing on this because on of our dev teams was looking at mandating macs for new developers because they were upset with how long it was taking users to get a functional Linux system installed and configured. The average time for new developers was 9 billable hours at the beginning, then another 6 hours over the first week, then another 14 hours in the first month assigned to setup and configuration time. In the end, we left the choice up to the individual, but a lot of us were pretty appalled. In any case, Linux is way behind in that regard. A couple hours of work is a lot more than a couple of clicks and then no need for the user to do anything. I love that my nicely customized and decked out linux install not goes with me from machine to machine, including all my accounts, tools, files, settings, etc. It sure saves me a lot more than a couple of hours each time.
I know a lot of former Linux on the desktop users who are now OS X users, partly because it makes configuration so easy.
My experience of auto-discovery protocols is that they are not especially reliable and I would much prefer to just set stuff up manually.Wow you must never have used Bonjour on a Mac. First, it always works. I've never seen it fail to find the printers or other services offered. You can literally just plug a printer into the network, or plug it into a Mac and click the "share" button and every mac on the LAN lists it as an option when you go to print, within 5 seconds. Second, how do "manually configure" your chat client to be able to hat with all the people at a conference you just went to? The last time I went to a conference, I opened up iChat and was able to have a Jabber IM conversation with every other Mac user on the local wireless. It is a great way to find things out and meet people. It's also great at the coffee shop and "manual configuration" for that would be onerous. And chat isn't the only application. At work there are shared streaming music services from people using iTunes and shared collaboration for text editing (SubEthaEdit rules for pair programming!) Have you seen all the cool applications that are collaborative for kids using the OLPC? Why can't regular Linux distros offer the same level of collaboration fro grown ups?
s an example, plugging 2 Windows machines into the same bit of network cable and giving them an IP address is supposed to let you access their network shares, but half the time one machine can't even find the other one (in my very limited Windows experience).Yeah, and MS has refused to implement ZeroConf instead using their own, broken, proprietary crap. That doesn't mean it can't be done right or that it hasn't. I'm just tired of waiting for Kubuntu to catch up. I love Kopete. It makes setting up accounts simple and is more customizable than iChat, but I spend the time to manually configure accounts in iChat just so I can use local chat without having to go up to everyone I see and ask them to register a Jabber account.
In this case, the plaintiff is alleging that Apple violated the Sherman Antitrust Act which requires that a monopoly or cartel must first exist.
This is untrue. According to the Sherman antitrust act the company or cartel in question need only have what they call "monopoly influence" in a market, which is later defined as having undue influence due to market share, such that they can undermine free trade.
So legally, the plaintiff must prove that Apple has a monopoly in some way. Let's say that for the sake of argument that Apple has a monopoly. It would however only be in iPods/iTunes. It would not be in digital music players or digital media.
They need to prove Apple has the ability to undermine free trade in a market using their market share and it is by market, not product. They're claiming Apple has such influence in "online video market," "online music market," "hard-drive based music player market," and the "Flash-based music player market." The courts may or may not accept those market definitions, but that is what they're being asked to rule on.
Well that depends on your definition of non-trivial.
Absolutely anything that makes it easier for people to use Apple's devices instead of a competitors that are result of their market share are all that matters according to the law.
If I understand your statement correctly, Apple's large market share makes them somehow responsible for any 3rd party attachment I loaded work with any of their competitors?
No, it doesn't make them responsible for the third party attachments. It makes them responsible for not taking any action which capitalizes on what third parties are doing as a result of their large market share, in order to gain share in a different market (like online video sales).
Even if we accept that somehow Apple is a monopoly in violation of the Sherman Antitrust Act, Apple would be in violation if they prevented their competitors from working 3rd parties not if they did nothing.
You're fundamentally failing to understand our antitrust law. No company is in violation of the Sherman antitrust act for being a monopoly. They're in violation if they're a monopoly which takes any action to leverage that monopoly into another market. There are two things to consider:
The behavior of third parties speaks to number 1. If people can't as easily go with a different vendor for their next device, this speaks to how much power Apple has, not to if they are abusing that power.
If they have monopoly influence, they are clearly abusing that influence. They are tying together their online store and their software with their player. Do you really think iTunes would be as popular if it did not ship with iPods? Do you really think they're sell as many songs from the iTunes store, if it was not the only store accessible with the software that ships with iPods? It is a clear cut case of tying, so the only question is "do they have enough influence to undermine the market" and that is the point in debate among people who have a clue about US or EU antitrust law. I'll say again, the ability to move music to a different device as easily as to a new iPod and the availability of third party add ons, both speak to how much power Apple has and those will be things the courts look at to make their determination. Apple is liable, under the law for tying, not for what third parties do, but what third parties do helps determine if what Apple has done (tying products) undermines the market and thus is illegal.
Even assuming that Apple is monopoly, unbundling iTunes does not alleviate the situation at all.
Yes it does. It removes Apple's ability to undermine the "music jukebox software" mark
Knowing how people are, probably three.
Maybe you tend to work with more incompetent/lazy people than I do. I know at least one of these engineers spent a month writing a new kernel module and security application for OS X for his own personal use and which he open sourced. Four of these engineers (all mac users) were poached from us by Google and now have one day a week to work on a personal project. A large number of the others contribute to one or more OSS projects.
When a large change is needed, the distributions that target those uses implement it and use it.
No, they don't. When was the last time Linux fundamentally changed in some way? Being compatible with the other distributions is more important than any one new feature, so for the most part new features that require a major change don't happen.
I've seen distributions take their own way all the time. Anyone who runs a distribution can do what they want with that distribution, if you don't have enough influence. You can fork off and start your own. That is the beauty of it.
Forking Linux for a desktop version is a doomed venture and everyone knows it. If you can't get multiple, major distros on board you can't make a big change.
Believe it or not, there are Linux distributions that do exactly that too.
No they don't because they still have to worry about being compatible with the other major distributions. You can't just migrate Ubuntu to OpenStep because all the packages Redhat makes will still be RPMs and thus you end up managing multiple package formats (like people do now) and it is actually worse than it was before.
It would take a new LSB proposal and once accepted the distributions would add support for it.
Ha. LSB, what a joke. When was the last time there was an accepted LSB proposal to make Linux work better as a desktop instead of a server. Oh yeah, never.
*shrugs* They will get what they want, and the Linux desktop users will get what they want. This is why we have distributions.
You miss the point. They enjoy having OpenStep style packages on the desktop because they work better. They will fight, however, to make it harder for Linux on the desktop to get that same feature because they want Linux to be a server and it is just bloat for a server. They are actively holding Linux on the desktop back.
Ironically I don't see OS X keeping up with Kubuntu in the long term. Already Apple can't seem to follow their own HiG philosophy.
I just happen to have worked as a user interface designer and usability tester for a while, and got paid to go to conferences and take classes on the subject. Yeah, Apple makes some interface mistakes, but they're still way, way ahead of Kubuntu in that regard.
As for fundamental new features - What fundamental features could they add?
System Services would be one. Drag and drop packages would be another. Built in support for MACLs with mandatory MACLs for all programs would be one. A package format that is portable, may include source, includes a link for updates for software that isn't in a repository, and usable support for commercial licensing would be nice. A trust/signing framework with a protocol for checking the references for a given program would be a step forward. A proper method of migrating your system to new hardware would be a plus. Need I go on?
The desktop environments (Gnome, KDE) have such capability already. There just isn't a grammar checker available yet. Things like spell checkers are universal to both DEs though and you can even choose which spell checker you want to use.
Really. Where, exactly do I put the system service file that will allow me to translate my Kopete chats and Konquerer forms between chinese and English automatically? OS X only got a standard grammar checker in the latest
Imagine for a moment you've never used a computer before and don't have a bunch of preconceptions and learned habits about how these features would be implemented. Imagine you're designing a new OS from scratch and have to decide how to design the GUI? Objectively, which is better, having one menu at the top, or one menu per window, with a single app maximized to fill the screen most of the time. With the menu always at the top, Fitt's law makes it an infinitely large target in one of the two dimensions and your testing shows that users can get to it faster and over time become faster yet because the menus are always in the same place instead of a different place depending on the window position. It also enables developers to make applications with menus, but no windows, when no window is needed. From a usability and design perspective, the OS X way is objectively better.
Now on to the minimize/maximize. What makes more sense a button that always makes an application fill the whole screen, or a button that the developers can set to fill the whole screen or grow just large enough to fit all the content, depending on the type of application and what the developer thinks is best? Surely giving developers the option is better no? The only reason it isn't better is if you are trained to always use every application in fullscreen mode because having multiple applications displayed simultaneously is too hard to use because of poor design choices (like Windows). Otherwise, having a window grow large enough to fit a Web page, but not automatically cover your chat client with blank window, makes a lot more sense for some types of windows.
Now imagine you build this OS and market it and years later someone else makes a different OS that does things differently. Should you revise your choices and make them objectively worse, pissing of all your current customers, simply to be more like the other OS and make things easier on people who want to switch from that OS?
Maybe - if I had to spend a whole year with this, I could learn to accept it over time... Mind you, this is personal taste.Personal preference counts for a lot and if you're stuck in your ways, sometimes it is best to stick with what you know or a close clone thereof. It just annoys me, a person who spent a lot of time learning about usability, user interfaces, and human/computer interaction, that things Apple did correctly from a scientific perspective are so often cited as faults and reasons to avoid their products. I mean Apple has made numerous UI mistakes that no one ever mentions, but they insist on bringing up things they've trained themselves to use on a poorly designed UI, as faults for a correctly designed one.
Actually, he said he went with a Thinkpad, which runs about 15-20% cheaper than the closest comparable Mac hardware (without any special discounts for students or the like). Of course it also has 2-3 times the hardware failure rate of an Apple laptop according to Consumer Reports.
I understand lack of support on Windows. They have a competing, proprietary protocol and they want to use it as their normal lock-in, but why hasn't anyone picked up the ball yet on Linux? An open standard with multiple FOSS implementations to harvest that provides useful new functionality seems like something that should have been implemented a long time ago as a standard service, with an easy way for developers to add it into their application.
By the way I think someone said they might work on a Kopete bonjour plugin a few weeks ago.That would be awesome and stop my GF complaining that I'm not "on IM" when I'm tinkering with my all Kubuntu box.
I'm also a little sad that OSX has dropped default support for printers annouced over CUPS broadcast but thems the breaks.I didn't hear about that, but my Leopard machine still sees the old TCP/IP printer on my LAN and I thought that was just CUPS. I'll have to investigate.
By "Linux laptop", I mean a Thinkpad (T60p), which runs Ubuntu.
My old laptop ran Kubuntu just fine, and my new one does too, albeit in a VM now, since that is more convenient for me.
I'm sorry, I need my right and middle button even when I'm not using an external mouse.
You might want to consider breaking that habit. I was skeptical at first, but the usability studies are pretty clear on it and my personal experience baks them up. When using an external mouse, your hand is off the keyboard and it is faster for advanced users to have multiple buttons. When using a trackpad your hands remain on the keyboard, so having to move both to hit multiple buttons wile still moving the cursor is actually slower than just using one button in combination with either a modifier key near your other hand (depending on your handedness) or using the two or three finger tap method. Using a multi-button trackpad is easier to learn, but in the long run it is slower for advanced users because you have to take your fingers off the keys more.
What actually drove me nuts was the little things, like not having normal buttons for PgUp/PgDown, Delete etc.
I actually like the Thinkpad keyboard better too, both for the arrangement of keys and the responsiveness.
Also, many of the Unix-land progs had subtle annoyances like confusing modifier keys etc.
Really? This has never been a problem for me. Like what?
And for drag and drop install/uninstall - ha! 99.9% of what I need is available from Apt, and it's so much simpler than drag-and-drop install.
Ahh, see here's a problem with people who try to use OS X. They don't look for the best tool, they just stick with what they were using on Linux, which is often a crappy port because no one uses it; there being a better, native application. As for Apt, there are better package managers for OS X, including ones that have a working graphical interface as well as a CLI one and which handle OS X native software in addition to ports.
Now in Linux, if something is missing and I want to compile from source, I stand a good chance of it just working. In Mac, no.
This is true some of the time, although I notice more and more it going the other way, especially for security tools. This is simply a function of what the develop who made the tool was using, since things tend to be slightly broken everywhere else. With so many developers switching to OS X, I find a lot of tools won't compile easily under Linux, but will in OS X, but then again I'm in the security field which may not reflect the greater software ecosystem.
For instance, I've spent a day trying to find a python install for Mac that would include all the packages I need for my project, while in Linux, well, all of them are just there.
That's odd. I know a lot, and I mean a lot of Python programmers running OS X and no one ever mentioned a problem. Typing "OS X Python Tools" into Google did not pull up a bunch of sources for what you need? Going to Python.org, undefined.org, or macpython.com didn't have them?
I really, really tried to configure the Mac to my liking, and btw.
This might be your problem. Most all of the people I've talked to who actually tried OS X and did not like it are what I like to call "crusty old fellas" who are very set in their ways and don't want to try a new OS, they just want something that will be Linux/Solaris/IRIX and let them use exactly the same workflows and exactly the same software as their favorite UNIX flavor. If you're trying to make OS X into Linux, it will never be as good at being Linux as Linux. If you're not willing to adapt your workflows and learn new, sometimes better ways to do things, why try another OS in the first place?
I see many people around me using them just fine - but not for real work...
Okay you were disgusted by... something and you got a "Linux laptop" by which I assume you mean you installed Linux on the MacBook, or was there some reason you couldn't get that to work?
Apple might be good for a grandma or for a graphic designer, but for a programmer it's an annoyance.I know a hundred or so programmers who have switched in the last couple of years and one, just one who switched back. That one spent a lot of his free time developing a Linux for the desktop distribution and even dual booting could not find all the drivers he needed for all his peripherals and did not like switching between different UIs all the time.
I'm not saying your opinions aren't valid, but come on, give us some meat here. At least tell people why you felt OS X was an annoyance for you, as a programmer. I'll tell you what I don't like about it. It doesn't ship with a good package manager so developers don't code for one. It has a really polished UI, but only for one workflow per activity. The default applications are featureful, but not standards compliant enough and don't do a good enough job of interacting with more advanced functions from other programs.
On the other hand, Linux lacks a simple hardware upgrade mechanism, does not have drag and drop install/uninstall or portable application packages, lacks system services, lacks ubiquitous support for ZeroConf, and has a lot of user interface problems and spotty user testing.
Well, sort of. The Linux development methodology does hurt it as a desktop in some ways, but the brain drain as people move to OS X hurts it more. If there were only Linux and Windows, a lot of people who now use OS X would choose Linux and devote time, money, and effort into fixing problems with Linux as a desktop and pushing for change. Instead, they mostly move to OS X for their desktop and actually argue against those changes because it is no longer in their best interests since they use Linux only as a server. This is nothing intentional on the part of Apple, but it is a real trend that I've noticed.
I run OS X, Kubuntu, and WinXP daily.
OS X is nice, but it is black box, and in the end you will loath it.I've been running it since version 10.0. It was pretty rough at first, but has really stabilized and continued to advance at a good pace. In fact, it is my OS of choice for applications, all other factors being equal. I don't loath it at all. More generically, I know literally a hundred or more engineers who have switched from Linux or a BSD in the last few years and only one who switched back and these are users with a choice of what they want to use for their daily work developing software for Linux and BSD based appliances.
As any other OS, it has lot of bugs, including VERY annoying ones which you can't fix even with having support contract with Apple and Adobe.I've had my fair share of feature requests go unfulfilled, but all the actual bugs I filed in OS X have been fixed and I don't have a contract with Apple. Adobe is another story, but that is unrelated to the platform. I have the same problem with their Windows and Linux software.
However, Apple and OS X strenght[sic] is integration.I'd say their strength is in their ability to make major changes that break things for the sake of advancing their platform and scrupulous user testing.
I just wish they would not be so annoyingly similar to Microsoft as they were in last year - all standard stuff, supporting OOXML, closing DAAP, etc.They have read only support for OOXML, while TextEdit reads and writes ODF, seems pretty useful to me. They never closed DAAP because they never opened it in the first place. Some people reverse engineered it and things broke when it changed.
For Linux, problems are two - user base and apps.I'd argue both of those are situational items that are contributed to by flaws in Linux. The first, is contributed to by Linux's commercial software unfriendly package management. Linux distros are varied and don't all use the same package manager or libraries, don't have support for software registration or software updates from a Website maintained by the distributor, who won't put it in a repository for technical and legal reasons. The second problem, user base, is partly because Linux does not do a very good job of catering to normal users, maintaining it's focus on current users who are mostly power users and CLI fans.
I'd also argue that while Linux is technologically ahead in a few ways, it is technologically behind in a lot of ways that matter to normal desktop users. There is no drag and drop package install/uninstall. Installed applications aren't easily portable. There are no OS X style system services. ZeroConf has not been ubiquitously integrated into standard applications. All of these (along with a good expose clone) are things I miss while using Linux.
However, I think it is not the end of desktop of Linux. For me, it's only now getting in shape that I have no shame to show to others.I've been a desktop Linux user for years, but in my opinion it is falling behind rather than catching up. I've also seen some serious brain drain as Linux desktop user/developers move to OS X and stop contributing to Linux desktop efforts. There is hope, but OS X does seem to be a serious detriment to desktop Linux, even if it is not intentional on the part of Apple.
This article isn't really a comparative study, but the point of comparative studies isn't just to arbitrarily decide which is best, but to explore the ways in which each offering is better than others and thus inform people who care about those particulars.
This is mainly why I dislike the title "is apple 'killing' Linux on the desktop". If anything, Linux is GAINING ground on apple in recent years - thanks to distros such as Gnome and Ubuntu - so why the hyperactive titles? One would thinkYeah, the title is a bit overblown, but that is par for the course. I do think OS X as an offering is hurting Linux for the desktop by stealing developers away and allowing them to refocus their extra-curricular programming on things other than desktop Linux (like Linux as a server or applications on OS X).