Cisco clearly owns the trademarked name, and in the phone category. And they've owned it sine long before the iPod, a name registered only in 2001. And they're using it. Sooner or later, probably sooner, Apple will have to pay money to Cisco. Lots of money.
I'm not sure about that. Apple owns the trademark in other countries. In the US, Linksys trademarked iPhone, but the product they sold died and they did nothing with the trademark for many, many years. It wasn't until the hype about Apple releasing a phone came out, with everyone calling it an "iPhone" that Cisco (who acquired Linksys) quickly rushed a product with that name to market. In fact, if you did a Google search a week after Cisco announced their product, 8 out of the top 10 results returned were talking about an unannounced product from Apple, one was cisco's marketing page and one was a page talking about how Cisco was trying to confuse customers by releasing something called an iPhone. If you don't defend a trademark and if the public perception of your trademark is that it is from another product, you lose that trademark. This is not like a copyright. The point of trademark law is to stop companies from tricking consumers into thinking a product is from one company when it is from another. Most people assume anything called "iPhone" is made by Apple, even before Apple announced it. This is no legal slam dunk for Cisco.
I think you might have misunderstood my original message. I wasn't suggesting charging for a "lite" version of the product. Rather, you'd maintain your free product with bug/security fixes and maybe the occasional small update - but put most of your innovative work into developing a "for pay" version of the same product.
I understood your message, I just disagreed with it. I've seen that business plan work when marketed properly, but that is not the case here. This is an established product with a user base. They are using it because it is free. Moving all innovation to a non-free version is unlikely to do anything other than steadily move your user base to Thunderbird and Evolution. A few might upgrade and become paying customers, but I doubt it would be enough to support the development.
And people who just want a free program to use can still do so with your original product.
Yeah, I've worked on just such a product. The non-free version falls behind the innovation curve very quickly. Thunderbird is not spending most of their time working on a pay version and it shows. If the client is to survive it needs to harness the user base and others interested in an alternative for a given niche, not drive them elsewhere. I just don't think your plan is workable for a product with an established user base of cheapskates.
Not at that price tag, they're not. That's with a two year agreement with Cingular. That's way, way beyond most peoples' price points, and with "only" 4 or 8 gigs of storage, it's roughly useless for the people who WOULD use it.
That has always been Apple's strategy. Skim the cream first. A lot of people I know paid that much or more for a PDA phone. This was really close to what I was guessing their prices would be, but a lot nicer looking than I anticipated.
It's a hybrid bastardization of several products that turned out to be a poor idea.
I have a cheap cell phone because I've never found an expensive one I liked. They all had crappy interfaces and were pretty indifferent at being phones. I don't own an iPod or any portable mp3 player because I never thought I would use it that much. I do own a GPS, but I rarely use it because I only have it on me when I'm hiking. I own a cheap digital camera I rarely use because I only have it with me on vacations. I figured Apple would be coming out with a phone, but I did not expect I would want one. But this is it. I'm willing to spend my money if I find a quality product that I think is worth it. My pocket knife cost $200, but it is not going to snap or fold while I'm using it. In general hybrid devices are not good at multiple things because the integration and interfaces suck. In principal they are great for items you carry with you because you actually are willing to carry them. The scissors on my old swiss army knife were not as good as a regular, full sized pair, but they got used because I had them with me. The same principal holds here. I'll buy a phone+PDA+mp3 player+GPS+camera because I will use those features if I have them with me all the time. I'll pay for them, if someone puts them together correctly, and it looks like someone finally has.
I predict Apple will not be able to keep these in stock and they will be the most popular an imitated device in years.
While DRM is intended to prevent or reduce casual copyright infringement...
I disagree with this. In my opinion DRM is intended to prevent lawful use of copyrighted material and motivate people to buy multiple copies of the same work by intentionally breaking interoperability with other devices. That is to say, content producers would like their customers to buy one copy for their home CD player another copy for the tape player in the car and another copy for their portable player. The industry is used to income from people periodically re-buying their favorite media in the new format or to replace the copy they have broken. They are terrified of the idea that a person could buy one copy and use it forever, handing it down to their children.
Media companies claim that they are trying to stop illegal copyright infringement, but they also claim accidentally posting a song on a file sharing network costs them hundreds of thousands of dollars in lost revenue, and if not for file sharing networks 90% of the gross national income would be spent on music. Why anyone would believe such obvious liars is beyond my understanding.
A cursory glance at the article would reveal that the spooks also work with Apple and that Novel also works with "somebody" in the govt.
It's not surprising that Apple would be partnering with the NSA. They briefly announced then removed all mention of a framework in Leopard that implements the mandatory access controls the NSA developed for SELinux. I have no doubt that they would be a valuable resource in auditing such an implementation.
You can have that now on linux, if you stick to KDE. (I imagine you can also have it if you stick to gnome) I have my mail client, web browser, IM program, word processor etc. with working spellcheck using the same dictionaries and, internally, the same component.
The problem is this only works for KDE applications that know about the component beforehand (called Kparts) and specifically include it. As a result, sure a few integrate spell checking, but there is no way to tailor this for other functions the user may want. On OS X these are drop in services (literally you drop them in the Services directory) and all programs that use the native APIs can use services that apply to those APIs. That means even if the creator of my IM program did not think I'd use a grammar checker, if I install one I still can use it in my IM client. The same goes for a dictionary/thesaurus lookup, online reference lookups, language translations, bibliography auto-formatting, etc. Application developers cannot know all the functions I will and won't use, nor even what functions will be available. Maybe some commercial software I buy includes excellent translation of English to and from Japanese. The creator of Kopete can't assume most people will have this software, and can't build it into Kopete. On the other hand, on OS X, Apple may know nothing about this commercial translation app, but if the app provides that translation as a service, iChat can use it anyway.
Kparts is better than nothing, but it is, in my opinion, insufficient. Building on the assumption that developers will know better than users what components each user will want, need, and have available is a very poor assumption.
I, personally, don't want spellchecking in my OS, unless its a subsystem/service that can be disabled. I almost never spellcheck (not much of what I write is important enough to justify it) and I don't want to incur any operating system overhead. I just want it to make sure my applications don't crash!
That's the point of implementing these at the OS level. One chunk of code running per function rather than on per function per application. Globally disabling and installing services is easy and if you don't use it in an app, it doesn't cost you anything. So maybe you couldn't care less about spellchecking. How about an MD5 checksum service? I have one of those installed. How about removing Windows style line endings? I have that too. How about looking up the highlighted function in the online docs? Application designers don't know everything you're going to do. By making the system modular everyone can install those services that they want/use and disable those they don't.
That much is true. However, commercial software does tend to package itself in such a way as to be usable anyway.
I haven't found that to be the case, much of the time. It certainly doesn't get the benefit of auto updates through the package manager, not the benefits of portability.
It might be nice to have the package manager be able to repackage things for easy transport, and include the dependencies there, but on a live system, I don't need that cruft.
Yeah power users often prefer a little saved disk, or RAM, especially on the server. That is why I doubt Linux (which face it is a great server but not ready for general purpose, novice desktop use) will get these usability features. Being able to uninstall by dragging to the trash and move to portable media or another computer with a simple drag operation, or update to a new machine on a new chip with one click are all features that benefit less expert users more than hardcore geeks. For those who scoff when they hear the mac "just works" this is what they're talking about.
Now, as for "my mom could do that", she certainly could do at least the part about copying a list of packages. She might have trouble copying config files over, but ideally, she won't have touched them much anyway, so she won't have to. And she can drag her home directory to an iPod, so that's not a big deal.
My mom can barely drag things to the trash. She doesn't have an iPod and if you said "config file" or "package list" to her she's politely nod then tell you to please write out a list of mouse clicks exactly what she should do, which would be too long and confusing and she would never actually even try it.
A significant amount of the software on my Mac cannot be uninstalled by dragging to trash. For instance: VPN software (menu + kernel extensions), Insomnia (kernel extension), a virtual desktop implementation, and a few other things.
Yeah, but that is all software that installs with an installer anyway, and hence cannot benefit from OS X style packages. You're probably in the 1% of users who would ever install such things. Normal people don't use VPNs or kernel extensions. Only power users do, and while they would benefit from a package manager (which I think Apple should add) they are not hindered significantly by OS X style packages for the rest of the software.
Regarding encryption: I don't use it, but I imagine it works fine -- there's at least a couple of solutions which use FUSE, which makes them just simple enough to seem bulletproof to me.
Others have said it worked fine for them, but it never has for me, and I tried twice. Installation was not as easy as clicking an "encrypt my account" and things broke. I'm willing to write this off as possibly bad distro choices or rotten luck.
It creates a PS file, not a PDF, but I imagine that's only a short step away. So, not there yet, but close, probably a script away.
That sort of highlights my point though. I imagine a lot of users make PDFs this way and don't even think about it. What is easy for them is rocket science for the average user. It might as well require them to write the code themselves if they must use the CLI. They are that reluctant to mess with it. I've found a lot of Linux tasks that are like this. They just aren't quite usable because no one has bothered.
That's starting to get scary. What kind of security is there?
Very little I imagine. You can add a password and limit who has access, but the text going back and forth is probably there for anyone on the LAN to sniff. For work and collaborate coding (it rules for pair programming, a text editor with two cursors going at once) we run it over a VPN.
Most of the software I run on my Mac is open source, and Mac support can be very strange. I have a list of maybe 20-30 apps I'd have to check -- some of them have their own autoupdate, most of them don't, and almost all of them are a royal pain when it's actually t
Oh come on go there!...lame argument from someone who can't possibly be a web developer...and if by some chance he is a web developer, he must not be very good at it. Once you get better at your job you might be able to afford the right tools for your job.
There are a lot of people who end up doing some casual Web development as a part of their job or hobby. Personally, I do Web development as a minor part of my job and I test with IE7 in a VM. That does not mean everyone who makes Web pages is going to have the same setup all the time. A lot of people who work remotely, for example, may not have access to a Windows machine when using a laptop, and this sort of a solution could save them a lot of debugging time.
Then I'd bet you'd find that IE hovers around 80%-85% of all browsers.
IE 7 has about 55% of the market according to the best numbers I've seen. It was autoloaded on pretty much all IE6 machines except Win2K and corporate machines that blocked it.
What is so compelling about IE7 that I'd want to go through any effort at all? I'm using Firefox 2.0something, it meets my needs. If I were to jump through hoops to install this on my linux box, what would that get for me?
Assuming you, like half of the people here, end up doing some Web development at some point, you get the ability to test those pages in IE7, which has about 50% of the market right now. Being able to do that without having to buy a copy of Windows is a pretty big deal to a lot of people.
To overcome this, I think the real answer is to invest some money in advertising a newer, more advanced version of the software product.
That sounds like a terrible business plan. After it has already been free as in beer and when your client base is built up of those users, you need to run with that theme. Keep it free as in beer and open source the project. Then profit in the following ways:
Add features and customizations on contract for any company that wants them. You the number one expert on this. You have the reputation. Get this out there and in use by all those companies that want a free and open solution, but aren't afraid to pay for changes they need.
Support support support. This money train can be a trap as it leads the developer to have an incentive to keep a product hard to install and use, but you can make a lot of money providing support services and installation and setup consulting services if you really have features that differentiate your product from the herd.
Tertiary sources, put it up for download and slap some ads on that page. Put up some documentation and other info and slap on a few more ads. The more popular your software, the more money you make. Finally, you're the developer of some cool software and a million people just read your name. Congratulations you have a marketable asset, your reputation in the field. Get hired to do other contract work based upon this.
Please note, while I think this is a sound way to run a business, I'm not sure it will succeed in this case because it is very, very late to the game. Thunderbird, eVolution, and several others have been free and open for a long time and I'm not sure the feature set of Pegasus is strong enough to compete with them. They have the mind share right now and you really have to be better, at least in some ecosystem, as well as open and free if you want to be widely deployed. Still, it is a better shot than selling a "lite" version of a product everyone is used to having for free.
Now Thunderbird offers in-place spell-check and other features which were considered very advanced just a few years ago.
I can't help but express my disappointment that this is still an issue. Spellchecking should not be implemented on a per-application level. It should be implemented at an OS level and offered as a service to all applications (along with other such services). I mean sure it's nice to have spellchecking in your mail client and your word processor, but what about your chat client? What about in vi? What about in your Web browser and calendar and graphics program. More importantly, why should you have to train dictionaries for all of these programs separately? I already taught my layout program that MPLS isn't a misspelling, why should I have to do it again and again? And what about my grammar checker? Should I wait another four years until they add that to Thunderbird? What about online dictionary lookups, and thesaurus, and language translations, and bibliography references? All these things I can do today in most programs on Mac OS X, which is great, but it is high time Linux and Windows caught up. That would be evolution in action.
You would be right about "most" users except that the people who really, really need a Windows applicaton are also going to be able to install Parallels - I have seen it happen a lot, even non technical people do not really have trouble with the concept.
Most people don't even know what an OS is. It's not that only really technical people need/want to run Windows apps, it is just that really technical people are the only people that know they can, are capable of installing it, and are willing to shell out the required money. My mom would like to her old, Windows only genealogy program rather than retype all the info, but she sure isn't willing to pay $100 bucks for Parallels, $120 for Windows, and however much it costs to find someone to install it.
CAD users, and a few other speciailzed applications - primarily technical. Anything non-technical is covered for sure, as are most technical areas at this point.
Games are not technical. Lots of random hardware devices that interface with computers (like kids toys) are not technical, realty listing software is not technical. You are, I believe, very much mistaken.
As far as gamers go, I completely disregard that argument since a gamer can either get a console or go with Bootcamp - gamers are those highly technical people that are installing Windows anyway to boost performance (at least the ones that "must" play HL2 are).
Maybe you've been hanging out with hardcore gamers too much, but most gamers are not computer people. They play games their friend brings over. Not all of these are available on the console and even if they are their friend may not own a console version. They are not going to be installing Bootcamp because they don't know it exists and don't know how to install Windows anyway. Expert users will find a way. Average users will not and I almost guarantee they will run into some software or task, even if it is just the IE only Website at a bank or other company they do business with.
When I upgraded my desktop from x86 to amd64 Linux, I had to install a new OS anyway -- and I'm assuming you would, too.
Upgrading OS X from an old machine to a new one consists of plugging a firewire cable between them and clicking the "upgrade from previous computer" option. It handles moving all the user accounts, settings, applications, files, certificates, etc. Sure you have to install a new OS, but installing a new OS does not me re-downloading and registering all my software.
Oh, and I thought I'd mention: I suppose you have unlimited RAM, also? VLC is a small-ish example, but just keep in mind, every library that you're not sharing for this odd kind of convenience is not just costing you disk and bandwidth, it's costing you RAM and execution speed.
RAM is always an issue, but I think you're overstating the case here. Larger downloads do not necessarily mean larger RAM fingerprints. Much of the download size is because it is a FAT binary with 32 and 64 bit PPC and Intel versions of the binary. As for shared Libraries, dynamic linking works just fine as do the included libraries so I'm not sure why it would be an issue.
What, you never uninstall programs?
Sure I do, but I rarely, if ever, want to remove the preference file that went with that application, which is the only thing with regard to uninstalling that is not accomplished by dragging it to the trash.
have apps that install things into the menubar, which won't go away just by throwing away the app.
Umm, what broken apps would those be?
I have apps that install kernel extensions... Basically, anything that comes in a.mpkg file is exactly the kind of thing which should have an uninstaller, but never does.
Oh you mean OS X style packages that don't come in the style of package we were talking about? So your problems with them are an argument against not using them? How interesting. Your arguments are for package managers, not against OS X style packages.
This puts OS X years behind Windows, by the way.
Heh. If being years behind means I never have any problems while being years ahead means I regularly have to run untrusted binaries, which sometimes don't work, and which are horribly inconvenient, I'll stick with years behind.
As for dependencies, I'll refer you back to VLC. Nobody, but nobody, is on such a fast connection that they don't notice the difference between 2 megs and 20 megs. 2 megs is "I want my VLC... Oh cool, it's here." 20 megs is go get some coffee. So, while you may prefer to waste everybody's resources so you can spend a half-second less time thinking, I'll stick with real package management.
20 megs is pretty fast for me and that 20 megs is not all wasted space. It adds functionality, even if it is functionality that only becomes apparent when I move between platforms and machines. As for an order of magnitude difference, that is also highly unusual. Most OS X applications bundles are not significantly different in size.
For the reasons I've stated above. Basically, if you have a good package manager, there isn't much of a reason for the OpenStep style package.
I strongly disagree. Being able to easily move applications between local drives, portable drives, network drives, and machines is a big win in my mind. Being able to treat them just like files really helps novice users and being able to delete by dragging to the trash has no real negative consequences for them, makes the system much more learnable. My mother can't go to add/remove programs and hunt through a list. It is just too complex. She does know how to drag files to the trash. I've seen Windows users drag applications to the trash in an attempt to uninstall them. I think what you really mean is you don't care about the benefits of these packages, so like most Linux dev people you ignore them.
I mean, everyone loves to talk about how they could put an app on their iPod,
Perhaps you could tell me why a 20 meg downlad of VLC for the Mac -- 10 megs of which I'll never use -- is better than a 2 meg download on Ubuntu.
Because my internet connection is really fast, and my disk is plenty big, so when I upgrade my laptop from 32 bit PPC to 64 bit Intel, VLC is copied over from the old machine and still works without me having to open any package manager.
Linux package management actually handles dependencies -- and has an actual uninstaller.
OS X package management removes the need for either.
So tell me again what you don't like about the Linux package managers. Tell me why a self-contained folder is so much simpler and easier to use than a list of software to install and a button that says "go".
Actually I really like Linux package managers. They provide the ability to manage upgrades from all third party software, make checksums easier, and facilitate software discovery. They also lend themselves well to a signing/verification system for determining application trust. I also like Apple style self contained folder make the packages portable and flexible. They remove the need for installer packages and you can use the same on on multiple machines, even if it is stored on your iPod. Installing some version of photoshop, then uninstalling it, and then installing a new version years later will result in all your preferences still being available if you so desire. Also, finding resources like images and sounds within the packages is much easier.
So here's my question for you: why isn't a good package manager that manages Apple's OpenStep style packages better than either individually? I'd love to see either Apple or a major Linux distro put both of these together and bring all the advantages to desktop users.
Uninstalling could still be improved though. Sometimes I don't mind that the preferences and like stick around. But when I get rid of an app that I know I'm never going to use again, I don't really want any of its files left cluttering up ~/Library or the/Library or.. well. anywhere else. Unfortunately, getting rid of those files is mainly a manual effort. Not a terribly fun one, either. Windows uninstallers, at least, get most of those files it scatters to winds during installation.
Actually, leaving those files has some nice advantages. Often when a user uninstalls an application, they install a newer version, or an alternative. In both cases it is useful for that app to have access to the preferences of the previous program. Also, when running applications off of removable media or the network, every time the disk is unplugged it is uninstalled but it is nice to have the preferences for that user and machine stick around.
I agree it would be nice to have a clean uninstall option that would remove these, but realistically the only reason to do so I can think of is for reasons of privacy. I do think that and a few other functions that a proper package manager could provide. I'd like to see the OpenStep style packages expanded to include repository and/or download locations for the file and for updates as well as an optional directory for source and build instructions so the user or OS can build custom binaries. It would be nice to gain the software discovery and update functionality of a package manager and combine it with the advantages of Apple's existing package style.
But really there is almost no software at this point an "average" user needs to run under Windows where there is not a mac version or equivilent.
This is also true, but also missing the point. The average person may not want or need to perform any given function that is unavailable on the Mac, but a lot of them are likely to want to do at least one of them. 2D CAD users, GPS enthusiasts, skydivers looking for altitude calculation tools, realtors looking for online property lookups, gamers who want to play some particular game with friends, etc. are all going to run into some niche software need that is not met. This is mostly just because of market share, but it is a real issue.
Also, anyone with a Mac knows where to find Mac wares. Just because you've never looked doesn't mean that it doesn't exist.
Really? My mom has a Mac, but I'll bet she doesn't know how to find "warez" or even what that means. People who download cracked copies of software from the internet are not a significant part of the general computer using populace. Most people that borrow, or copy games illegally use copies that are owned by a friend. Most of those friends (statistically speaking) are going to have PC versions.
All macs sold now are Intel macs, and thus all macs sold now support parallels...
For most users, that is not a viable option. Sorry, but most users can't install an OS at all, don't know that they can run Windows on a mac, and the cost barrier of one copy of parallels and one copy of Windows is too high.
But really there is almost no software at this point an "average" user needs to run under Windows where there is not a mac version or equivilent.
This is a logical misapplication. There is no given user who is the "average" user. The question is how many users want to do something and can't because they have a mac and software is unavailable to them either because there is no mac software that does the same thing or because they can't acquire that software (when they can acquire the Windows version).
The average person will never need to run a 2D CAD program. The average person will not want to play Half-Life 2. A significant number of people will want to do either the former, latter, or will want to do some other task that will not be easy for them because they are on the mac platform.
Application packages: Unfortunately not as easy or thorough as you'd like. Don't handle shared libraries. However, I can IM someone the name of a package, and they can install it on their own package manager -- which is more secure, handles real dependencies, and provides an uninstall which can be more than simply deleting files.
Shared libraries are a solved problem with dynamic linking. IM'ing the name of a package is no good if it is commercial software, as hardly any is in repositories. It certainly doesn't help when the company no longer distributes the software at all (this is not theoretical, it has already happened to me once). It is often more convenient to drag a file from one computer to another on a local network than to re-download it. You did not address access to programs on portable media. As for uninstalling, I think it is more important for this to be easy for novices (drag to trash) than for it to provide advanced options. Both would be better yet. You might have the mistaken impression that I'm arguing against package managers. I'm not. I think OS X should integrate a package manager to allow for a single channel for updates for all software, and to handle application uninstalling where you want to delete all traces including your preference files (although this is a rare need and not hard now). I'd be a happy camper if either Linux or OS X would integrate both sets of functionality.
Upgrade -- I'm not sure why this was so difficult for you; it took me maybe an hour. I did not use FireWire, I put the old hard drive in the new box temporarily, but I could've easily done it over FireWire, or Ethernet, or whatever. Copy over/home and/etc, and since it was Gentoo, I can copy over/var/lib/portage/world. Then perform a fresh Gentoo install. Needs a little manual babysitting, but by the time it's done -- again, less than an hour -- I've got everything I had before, only it can be on a new arch. Case in point, I recently migrated from an x86 box on a single hard drive to an amd64 box on a RAID in this way.
Yeah, because my mom can do that. Upgrading a linux box to a new machine generally requires I manually babysit moving chunks of structure over, and do a fresh install and then fix all the settings on that fresh install that did not get copied over because they exist somewhere else. It is doable for the average Linux user, but not for the average computer user. I think the fact that current Linux users are mostly power users blinds many developers to why most Linux users are power users. A whole lot of the functionality (like this) requires you to have in depth knowledge of the system, unlike plugging in a cable and clicking the upgrade from another computer option.
Not sure what "getting in your way" means, but I can do encryption too, either on an individual account or for the entire machine.
I've tried using account level encryption on two different Linux distros but abandoned both when things stopped working properly. I ended up simply encrypting a few files I cared about. This is pretty anecdotal though, so if you say it works fine for you I'm willing to concede this one.
for instance, OpenOffice can save to PDF, and that's enough, if I ever wanted PDF in the first place.
PDFs are a great archival format. They are exacting and standard and there are multiple open and closed implementations on every platform. I don't think I've ever run across a machine that could not read them. I use PDFs for communications with customers and coworkers. I create them from my browser to archive all the online receipts and orders I make. If you only create documents in OpenOffice I can see why this would not matter much to you, but if you use other products like layout applications, vector graphics suites, etc. it is nice to always have the option without any work on the part of the developers. If applications can print, and convert postscript to PDF, why not make it easy to do
The "Migration", obviously, is usually over before you're half done with your Bagel. So what do you do now? Cheat on your employer's time, just finishing off that Bagel, or finish off that Bagel while griping about your work overload (you are a SysAdmin, right?).
Interestingly, you're mistaken in two ways. First, the upgrade actually takes quite a while when you have 60 gig of applications and data. It is slower than I would expect even. Second, I'm on salary for my primary job. So long as my work gets done, my boss won't complain if I'm playing video games or watching old movies. Not that I slack off a lot, I want those options to be worth something.
You know, this touches on one of my pet peeves. The answer to "Do you want to insert action here" in English is "Yes" or "No." The answer is not "OK" or "Continue."
The point I was making was not a grammar issue. Effective communication often runs afoul of propriety. The point was that buttons should always be labelled with actions, so that the user knows they are choosing a given action. If a message says, "The application has a memory error (OK)(Cancel)" what do those buttons do? Does "Cancel" cancel the error or the program or something else? The message needs to be rewritten with actions, "The application has a memory error. (Stop the application)(Try to keep running the application anyway)"
Also, if there's' no option, why do I have to click on something?
Agreed, this is a fairly common and well known UI snafu. It just trains users to ignore messages that are important.
OpenStep style application packages - application install and uninstall easily, application on removable media save preferences on local machines so you can move one installed app back and forth between machines, resources are easy to find so extracting an image or sound is easy, fat binaries are cake, I can IM an application to someone on a different chip architecture even and it just works without having to find installers
Upgrade via firewire - I now run both Windows and Linux in VMs on top of OS X, mostly because of this feature. Plug in a cable to my old machine and click a button. All my files, applications, user accounts, settings, certificates, everything migrates seamlessly while I go for coffee and a bagel. Migrating Windows or Linux to a new machine takes significant time, sometimes weeks, and nontrivial effort, (or did until I installed them in VMs. Now full installs of those OS's come with me as well.)
System services - applications and plug-ins can easily share functionality across all applications. I only have to train one spelling checker and it works in my mail, web browser, word processor, terminals, pro layout app, photoshop, chat client, etc. The same goes for grammar checking, language translations, a pile of scripts, statistics on text like word count, automated bibliography entries, dictionary and thesaurus lookups, online lookups at numerous references, etc. Writing the same functionality over and over again for each app is outdated. Easily re-implementable libraries like Kparts on KDE only work if the programmer knows beforehand about the library, so no one uses them. Services on OS X requires no work on the part of a given app developer. The maintainers of subethaedit have never heard of omnidictionary. The developers of omnidictionary probably never considered subethaedit users. But in subethaedit I can still easily perform online dictionary lookups at a dozen different dictionaries with a single key combo. Losing this would be disasterous to my everyday workflow.
Save to PDF from every application.
Expose for easily finding and switching to one of my two dozen or more of windows.
user account encryption that both works and does not ever get in my way.
Ubiquitous application of zero-conf for local discovery of chat, music streaming, filesharing, collaboration tools, etc.
Automater - I know I didn't think I'd use it either, but this is the fastest way I've ever found to do things like add a watermark to every page in a PDF when I don't have the source file. It is also the only scripting some nontechnical coworkers have ever managed to use. Before it became available they would sit an rename 500 files by hand, rather than spending 60 seconds writing a script.
I'm sure there are more items I'm forgetting and again I want to stress that OS X is not ahead in all areas and can really benefit from improvements. It is just that some of these things have been on OS X for quite a while and most Linux developers I talk to don't even recognize the value in them. A lot of them are things that you can work around on Linux, or hack something that works in one instance, but until they are available to average and novice users, they are just ignored anyway. I'd love to see Linux catch up to OS X on the desktop, I just don't anticipate it happening anytime soon. I don't think Linux developers are willing to make some of the hard choices needed or will be willing to accept complexity on the server for the sake of making Linux nice on the desktop.
Major innovation will be on Linux and OS X. It will take a month or two for Linux to absorb new OS X features...
I wish. I don't see linux absorbing major new features from OS X in a month, nor many of them even in years. I like Linux. I run Linux, even on the desktop. Linux is ahead of OS X in a number of ways. But Linux is still missing a whole lot of features that lay the groundwork for what makes OS X my main desktop. Most of what I see Linux taking from OS X is minor eye candy and UI ideas.
...and OS X will generally absorb some of the good ideas from Linux in the next release of their OS...
Yeah, you hit the nail on the head here. OS X adds a few Linux/UNIX features with every release, while ignoring yet other features. Yay! traceD and virtual desktops. Boo! no ubiquitous update manager for all applications.
Yep, the games and apps people use are definitely not available on any version of OS X.
You are correct in your facts, even when you just consider games, but I think you are looking at the wrong information. Whether or not the most popular software is needed/available on OS X is not as important as if the average person wants to run software or perform a function which they cannot. There are many applications that don't have a port and while individually they may not have a lot of market share, together they account for a lot of people being stopped from doing something. Also, software piracy/lending plays a big part. A whole lot of people use games and applications they borrowed or copied from someone else, and even if there is a Mac version for sale, that does not mean there is a Mac version accessible to them.
Slashdot Mirror
Cisco clearly owns the trademarked name, and in the phone category. And they've owned it sine long before the iPod, a name registered only in 2001. And they're using it. Sooner or later, probably sooner, Apple will have to pay money to Cisco. Lots of money.
I'm not sure about that. Apple owns the trademark in other countries. In the US, Linksys trademarked iPhone, but the product they sold died and they did nothing with the trademark for many, many years. It wasn't until the hype about Apple releasing a phone came out, with everyone calling it an "iPhone" that Cisco (who acquired Linksys) quickly rushed a product with that name to market. In fact, if you did a Google search a week after Cisco announced their product, 8 out of the top 10 results returned were talking about an unannounced product from Apple, one was cisco's marketing page and one was a page talking about how Cisco was trying to confuse customers by releasing something called an iPhone. If you don't defend a trademark and if the public perception of your trademark is that it is from another product, you lose that trademark. This is not like a copyright. The point of trademark law is to stop companies from tricking consumers into thinking a product is from one company when it is from another. Most people assume anything called "iPhone" is made by Apple, even before Apple announced it. This is no legal slam dunk for Cisco.
I think you might have misunderstood my original message. I wasn't suggesting charging for a "lite" version of the product. Rather, you'd maintain your free product with bug/security fixes and maybe the occasional small update - but put most of your innovative work into developing a "for pay" version of the same product.
I understood your message, I just disagreed with it. I've seen that business plan work when marketed properly, but that is not the case here. This is an established product with a user base. They are using it because it is free. Moving all innovation to a non-free version is unlikely to do anything other than steadily move your user base to Thunderbird and Evolution. A few might upgrade and become paying customers, but I doubt it would be enough to support the development.
And people who just want a free program to use can still do so with your original product.
Yeah, I've worked on just such a product. The non-free version falls behind the innovation curve very quickly. Thunderbird is not spending most of their time working on a pay version and it shows. If the client is to survive it needs to harness the user base and others interested in an alternative for a given niche, not drive them elsewhere. I just don't think your plan is workable for a product with an established user base of cheapskates.
Not at that price tag, they're not. That's with a two year agreement with Cingular. That's way, way beyond most peoples' price points, and with "only" 4 or 8 gigs of storage, it's roughly useless for the people who WOULD use it.
That has always been Apple's strategy. Skim the cream first. A lot of people I know paid that much or more for a PDA phone. This was really close to what I was guessing their prices would be, but a lot nicer looking than I anticipated.
It's a hybrid bastardization of several products that turned out to be a poor idea.
I have a cheap cell phone because I've never found an expensive one I liked. They all had crappy interfaces and were pretty indifferent at being phones. I don't own an iPod or any portable mp3 player because I never thought I would use it that much. I do own a GPS, but I rarely use it because I only have it on me when I'm hiking. I own a cheap digital camera I rarely use because I only have it with me on vacations. I figured Apple would be coming out with a phone, but I did not expect I would want one. But this is it. I'm willing to spend my money if I find a quality product that I think is worth it. My pocket knife cost $200, but it is not going to snap or fold while I'm using it. In general hybrid devices are not good at multiple things because the integration and interfaces suck. In principal they are great for items you carry with you because you actually are willing to carry them. The scissors on my old swiss army knife were not as good as a regular, full sized pair, but they got used because I had them with me. The same principal holds here. I'll buy a phone+PDA+mp3 player+GPS+camera because I will use those features if I have them with me all the time. I'll pay for them, if someone puts them together correctly, and it looks like someone finally has.
I predict Apple will not be able to keep these in stock and they will be the most popular an imitated device in years.
While DRM is intended to prevent or reduce casual copyright infringement...
I disagree with this. In my opinion DRM is intended to prevent lawful use of copyrighted material and motivate people to buy multiple copies of the same work by intentionally breaking interoperability with other devices. That is to say, content producers would like their customers to buy one copy for their home CD player another copy for the tape player in the car and another copy for their portable player. The industry is used to income from people periodically re-buying their favorite media in the new format or to replace the copy they have broken. They are terrified of the idea that a person could buy one copy and use it forever, handing it down to their children.
Media companies claim that they are trying to stop illegal copyright infringement, but they also claim accidentally posting a song on a file sharing network costs them hundreds of thousands of dollars in lost revenue, and if not for file sharing networks 90% of the gross national income would be spent on music. Why anyone would believe such obvious liars is beyond my understanding.
A cursory glance at the article would reveal that the spooks also work with Apple and that Novel also works with "somebody" in the govt.
It's not surprising that Apple would be partnering with the NSA. They briefly announced then removed all mention of a framework in Leopard that implements the mandatory access controls the NSA developed for SELinux. I have no doubt that they would be a valuable resource in auditing such an implementation.
You can have that now on linux, if you stick to KDE. (I imagine you can also have it if you stick to gnome) I have my mail client, web browser, IM program, word processor etc. with working spellcheck using the same dictionaries and, internally, the same component.
The problem is this only works for KDE applications that know about the component beforehand (called Kparts) and specifically include it. As a result, sure a few integrate spell checking, but there is no way to tailor this for other functions the user may want. On OS X these are drop in services (literally you drop them in the Services directory) and all programs that use the native APIs can use services that apply to those APIs. That means even if the creator of my IM program did not think I'd use a grammar checker, if I install one I still can use it in my IM client. The same goes for a dictionary/thesaurus lookup, online reference lookups, language translations, bibliography auto-formatting, etc. Application developers cannot know all the functions I will and won't use, nor even what functions will be available. Maybe some commercial software I buy includes excellent translation of English to and from Japanese. The creator of Kopete can't assume most people will have this software, and can't build it into Kopete. On the other hand, on OS X, Apple may know nothing about this commercial translation app, but if the app provides that translation as a service, iChat can use it anyway.
Kparts is better than nothing, but it is, in my opinion, insufficient. Building on the assumption that developers will know better than users what components each user will want, need, and have available is a very poor assumption.
I, personally, don't want spellchecking in my OS, unless its a subsystem/service that can be disabled. I almost never spellcheck (not much of what I write is important enough to justify it) and I don't want to incur any operating system overhead. I just want it to make sure my applications don't crash!
That's the point of implementing these at the OS level. One chunk of code running per function rather than on per function per application. Globally disabling and installing services is easy and if you don't use it in an app, it doesn't cost you anything. So maybe you couldn't care less about spellchecking. How about an MD5 checksum service? I have one of those installed. How about removing Windows style line endings? I have that too. How about looking up the highlighted function in the online docs? Application designers don't know everything you're going to do. By making the system modular everyone can install those services that they want/use and disable those they don't.
That much is true. However, commercial software does tend to package itself in such a way as to be usable anyway.
I haven't found that to be the case, much of the time. It certainly doesn't get the benefit of auto updates through the package manager, not the benefits of portability.
It might be nice to have the package manager be able to repackage things for easy transport, and include the dependencies there, but on a live system, I don't need that cruft.
Yeah power users often prefer a little saved disk, or RAM, especially on the server. That is why I doubt Linux (which face it is a great server but not ready for general purpose, novice desktop use) will get these usability features. Being able to uninstall by dragging to the trash and move to portable media or another computer with a simple drag operation, or update to a new machine on a new chip with one click are all features that benefit less expert users more than hardcore geeks. For those who scoff when they hear the mac "just works" this is what they're talking about.
Now, as for "my mom could do that", she certainly could do at least the part about copying a list of packages. She might have trouble copying config files over, but ideally, she won't have touched them much anyway, so she won't have to. And she can drag her home directory to an iPod, so that's not a big deal.
My mom can barely drag things to the trash. She doesn't have an iPod and if you said "config file" or "package list" to her she's politely nod then tell you to please write out a list of mouse clicks exactly what she should do, which would be too long and confusing and she would never actually even try it.
A significant amount of the software on my Mac cannot be uninstalled by dragging to trash. For instance: VPN software (menu + kernel extensions), Insomnia (kernel extension), a virtual desktop implementation, and a few other things.
Yeah, but that is all software that installs with an installer anyway, and hence cannot benefit from OS X style packages. You're probably in the 1% of users who would ever install such things. Normal people don't use VPNs or kernel extensions. Only power users do, and while they would benefit from a package manager (which I think Apple should add) they are not hindered significantly by OS X style packages for the rest of the software.
Regarding encryption: I don't use it, but I imagine it works fine -- there's at least a couple of solutions which use FUSE, which makes them just simple enough to seem bulletproof to me.
Others have said it worked fine for them, but it never has for me, and I tried twice. Installation was not as easy as clicking an "encrypt my account" and things broke. I'm willing to write this off as possibly bad distro choices or rotten luck.
It creates a PS file, not a PDF, but I imagine that's only a short step away. So, not there yet, but close, probably a script away.
That sort of highlights my point though. I imagine a lot of users make PDFs this way and don't even think about it. What is easy for them is rocket science for the average user. It might as well require them to write the code themselves if they must use the CLI. They are that reluctant to mess with it. I've found a lot of Linux tasks that are like this. They just aren't quite usable because no one has bothered.
That's starting to get scary. What kind of security is there?
Very little I imagine. You can add a password and limit who has access, but the text going back and forth is probably there for anyone on the LAN to sniff. For work and collaborate coding (it rules for pair programming, a text editor with two cursors going at once) we run it over a VPN.
Most of the software I run on my Mac is open source, and Mac support can be very strange. I have a list of maybe 20-30 apps I'd have to check -- some of them have their own autoupdate, most of them don't, and almost all of them are a royal pain when it's actually t
Oh come on go there!...lame argument from someone who can't possibly be a web developer...and if by some chance he is a web developer, he must not be very good at it. Once you get better at your job you might be able to afford the right tools for your job.
There are a lot of people who end up doing some casual Web development as a part of their job or hobby. Personally, I do Web development as a minor part of my job and I test with IE7 in a VM. That does not mean everyone who makes Web pages is going to have the same setup all the time. A lot of people who work remotely, for example, may not have access to a Windows machine when using a laptop, and this sort of a solution could save them a lot of debugging time.
Then I'd bet you'd find that IE hovers around 80%-85% of all browsers.
IE 7 has about 55% of the market according to the best numbers I've seen. It was autoloaded on pretty much all IE6 machines except Win2K and corporate machines that blocked it.
What is so compelling about IE7 that I'd want to go through any effort at all? I'm using Firefox 2.0something, it meets my needs. If I were to jump through hoops to install this on my linux box, what would that get for me?
Assuming you, like half of the people here, end up doing some Web development at some point, you get the ability to test those pages in IE7, which has about 50% of the market right now. Being able to do that without having to buy a copy of Windows is a pretty big deal to a lot of people.
To overcome this, I think the real answer is to invest some money in advertising a newer, more advanced version of the software product.
That sounds like a terrible business plan. After it has already been free as in beer and when your client base is built up of those users, you need to run with that theme. Keep it free as in beer and open source the project. Then profit in the following ways:
Please note, while I think this is a sound way to run a business, I'm not sure it will succeed in this case because it is very, very late to the game. Thunderbird, eVolution, and several others have been free and open for a long time and I'm not sure the feature set of Pegasus is strong enough to compete with them. They have the mind share right now and you really have to be better, at least in some ecosystem, as well as open and free if you want to be widely deployed. Still, it is a better shot than selling a "lite" version of a product everyone is used to having for free.
Now Thunderbird offers in-place spell-check and other features which were considered very advanced just a few years ago.
I can't help but express my disappointment that this is still an issue. Spellchecking should not be implemented on a per-application level. It should be implemented at an OS level and offered as a service to all applications (along with other such services). I mean sure it's nice to have spellchecking in your mail client and your word processor, but what about your chat client? What about in vi? What about in your Web browser and calendar and graphics program. More importantly, why should you have to train dictionaries for all of these programs separately? I already taught my layout program that MPLS isn't a misspelling, why should I have to do it again and again? And what about my grammar checker? Should I wait another four years until they add that to Thunderbird? What about online dictionary lookups, and thesaurus, and language translations, and bibliography references? All these things I can do today in most programs on Mac OS X, which is great, but it is high time Linux and Windows caught up. That would be evolution in action.
You would be right about "most" users except that the people who really, really need a Windows applicaton are also going to be able to install Parallels - I have seen it happen a lot, even non technical people do not really have trouble with the concept.
Most people don't even know what an OS is. It's not that only really technical people need/want to run Windows apps, it is just that really technical people are the only people that know they can, are capable of installing it, and are willing to shell out the required money. My mom would like to her old, Windows only genealogy program rather than retype all the info, but she sure isn't willing to pay $100 bucks for Parallels, $120 for Windows, and however much it costs to find someone to install it.
CAD users, and a few other speciailzed applications - primarily technical. Anything non-technical is covered for sure, as are most technical areas at this point.
Games are not technical. Lots of random hardware devices that interface with computers (like kids toys) are not technical, realty listing software is not technical. You are, I believe, very much mistaken.
As far as gamers go, I completely disregard that argument since a gamer can either get a console or go with Bootcamp - gamers are those highly technical people that are installing Windows anyway to boost performance (at least the ones that "must" play HL2 are).
Maybe you've been hanging out with hardcore gamers too much, but most gamers are not computer people. They play games their friend brings over. Not all of these are available on the console and even if they are their friend may not own a console version. They are not going to be installing Bootcamp because they don't know it exists and don't know how to install Windows anyway. Expert users will find a way. Average users will not and I almost guarantee they will run into some software or task, even if it is just the IE only Website at a bank or other company they do business with.
When I upgraded my desktop from x86 to amd64 Linux, I had to install a new OS anyway -- and I'm assuming you would, too.
Upgrading OS X from an old machine to a new one consists of plugging a firewire cable between them and clicking the "upgrade from previous computer" option. It handles moving all the user accounts, settings, applications, files, certificates, etc. Sure you have to install a new OS, but installing a new OS does not me re-downloading and registering all my software.
Oh, and I thought I'd mention: I suppose you have unlimited RAM, also? VLC is a small-ish example, but just keep in mind, every library that you're not sharing for this odd kind of convenience is not just costing you disk and bandwidth, it's costing you RAM and execution speed.
RAM is always an issue, but I think you're overstating the case here. Larger downloads do not necessarily mean larger RAM fingerprints. Much of the download size is because it is a FAT binary with 32 and 64 bit PPC and Intel versions of the binary. As for shared Libraries, dynamic linking works just fine as do the included libraries so I'm not sure why it would be an issue.
What, you never uninstall programs?
Sure I do, but I rarely, if ever, want to remove the preference file that went with that application, which is the only thing with regard to uninstalling that is not accomplished by dragging it to the trash.
have apps that install things into the menubar, which won't go away just by throwing away the app.
Umm, what broken apps would those be?
I have apps that install kernel extensions... Basically, anything that comes in a .mpkg file is exactly the kind of thing which should have an uninstaller, but never does.
Oh you mean OS X style packages that don't come in the style of package we were talking about? So your problems with them are an argument against not using them? How interesting. Your arguments are for package managers, not against OS X style packages.
This puts OS X years behind Windows, by the way.
Heh. If being years behind means I never have any problems while being years ahead means I regularly have to run untrusted binaries, which sometimes don't work, and which are horribly inconvenient, I'll stick with years behind.
As for dependencies, I'll refer you back to VLC. Nobody, but nobody, is on such a fast connection that they don't notice the difference between 2 megs and 20 megs. 2 megs is "I want my VLC... Oh cool, it's here." 20 megs is go get some coffee. So, while you may prefer to waste everybody's resources so you can spend a half-second less time thinking, I'll stick with real package management.
20 megs is pretty fast for me and that 20 megs is not all wasted space. It adds functionality, even if it is functionality that only becomes apparent when I move between platforms and machines. As for an order of magnitude difference, that is also highly unusual. Most OS X applications bundles are not significantly different in size.
For the reasons I've stated above. Basically, if you have a good package manager, there isn't much of a reason for the OpenStep style package.
I strongly disagree. Being able to easily move applications between local drives, portable drives, network drives, and machines is a big win in my mind. Being able to treat them just like files really helps novice users and being able to delete by dragging to the trash has no real negative consequences for them, makes the system much more learnable. My mother can't go to add/remove programs and hunt through a list. It is just too complex. She does know how to drag files to the trash. I've seen Windows users drag applications to the trash in an attempt to uninstall them. I think what you really mean is you don't care about the benefits of these packages, so like most Linux dev people you ignore them.
I mean, everyone loves to talk about how they could put an app on their iPod,
Perhaps you could tell me why a 20 meg downlad of VLC for the Mac -- 10 megs of which I'll never use -- is better than a 2 meg download on Ubuntu.
Because my internet connection is really fast, and my disk is plenty big, so when I upgrade my laptop from 32 bit PPC to 64 bit Intel, VLC is copied over from the old machine and still works without me having to open any package manager.
Linux package management actually handles dependencies -- and has an actual uninstaller.
OS X package management removes the need for either.
So tell me again what you don't like about the Linux package managers. Tell me why a self-contained folder is so much simpler and easier to use than a list of software to install and a button that says "go".
Actually I really like Linux package managers. They provide the ability to manage upgrades from all third party software, make checksums easier, and facilitate software discovery. They also lend themselves well to a signing/verification system for determining application trust. I also like Apple style self contained folder make the packages portable and flexible. They remove the need for installer packages and you can use the same on on multiple machines, even if it is stored on your iPod. Installing some version of photoshop, then uninstalling it, and then installing a new version years later will result in all your preferences still being available if you so desire. Also, finding resources like images and sounds within the packages is much easier.
So here's my question for you: why isn't a good package manager that manages Apple's OpenStep style packages better than either individually? I'd love to see either Apple or a major Linux distro put both of these together and bring all the advantages to desktop users.
Uninstalling could still be improved though. Sometimes I don't mind that the preferences and like stick around. But when I get rid of an app that I know I'm never going to use again, I don't really want any of its files left cluttering up ~/Library or the /Library or.. well. anywhere else. Unfortunately, getting rid of those files is mainly a manual effort. Not a terribly fun one, either. Windows uninstallers, at least, get most of those files it scatters to winds during installation.
Actually, leaving those files has some nice advantages. Often when a user uninstalls an application, they install a newer version, or an alternative. In both cases it is useful for that app to have access to the preferences of the previous program. Also, when running applications off of removable media or the network, every time the disk is unplugged it is uninstalled but it is nice to have the preferences for that user and machine stick around.
I agree it would be nice to have a clean uninstall option that would remove these, but realistically the only reason to do so I can think of is for reasons of privacy. I do think that and a few other functions that a proper package manager could provide. I'd like to see the OpenStep style packages expanded to include repository and/or download locations for the file and for updates as well as an optional directory for source and build instructions so the user or OS can build custom binaries. It would be nice to gain the software discovery and update functionality of a package manager and combine it with the advantages of Apple's existing package style.
But really there is almost no software at this point an "average" user needs to run under Windows where there is not a mac version or equivilent.
This is also true, but also missing the point. The average person may not want or need to perform any given function that is unavailable on the Mac, but a lot of them are likely to want to do at least one of them. 2D CAD users, GPS enthusiasts, skydivers looking for altitude calculation tools, realtors looking for online property lookups, gamers who want to play some particular game with friends, etc. are all going to run into some niche software need that is not met. This is mostly just because of market share, but it is a real issue.
Also, anyone with a Mac knows where to find Mac wares. Just because you've never looked doesn't mean that it doesn't exist.
Really? My mom has a Mac, but I'll bet she doesn't know how to find "warez" or even what that means. People who download cracked copies of software from the internet are not a significant part of the general computer using populace. Most people that borrow, or copy games illegally use copies that are owned by a friend. Most of those friends (statistically speaking) are going to have PC versions.
All macs sold now are Intel macs, and thus all macs sold now support parallels...
For most users, that is not a viable option. Sorry, but most users can't install an OS at all, don't know that they can run Windows on a mac, and the cost barrier of one copy of parallels and one copy of Windows is too high.
But really there is almost no software at this point an "average" user needs to run under Windows where there is not a mac version or equivilent.
This is a logical misapplication. There is no given user who is the "average" user. The question is how many users want to do something and can't because they have a mac and software is unavailable to them either because there is no mac software that does the same thing or because they can't acquire that software (when they can acquire the Windows version).
The average person will never need to run a 2D CAD program. The average person will not want to play Half-Life 2. A significant number of people will want to do either the former, latter, or will want to do some other task that will not be easy for them because they are on the mac platform.
Application packages: Unfortunately not as easy or thorough as you'd like. Don't handle shared libraries. However, I can IM someone the name of a package, and they can install it on their own package manager -- which is more secure, handles real dependencies, and provides an uninstall which can be more than simply deleting files.
Shared libraries are a solved problem with dynamic linking. IM'ing the name of a package is no good if it is commercial software, as hardly any is in repositories. It certainly doesn't help when the company no longer distributes the software at all (this is not theoretical, it has already happened to me once). It is often more convenient to drag a file from one computer to another on a local network than to re-download it. You did not address access to programs on portable media. As for uninstalling, I think it is more important for this to be easy for novices (drag to trash) than for it to provide advanced options. Both would be better yet. You might have the mistaken impression that I'm arguing against package managers. I'm not. I think OS X should integrate a package manager to allow for a single channel for updates for all software, and to handle application uninstalling where you want to delete all traces including your preference files (although this is a rare need and not hard now). I'd be a happy camper if either Linux or OS X would integrate both sets of functionality.
Upgrade -- I'm not sure why this was so difficult for you; it took me maybe an hour. I did not use FireWire, I put the old hard drive in the new box temporarily, but I could've easily done it over FireWire, or Ethernet, or whatever. Copy over /home and /etc, and since it was Gentoo, I can copy over /var/lib/portage/world. Then perform a fresh Gentoo install. Needs a little manual babysitting, but by the time it's done -- again, less than an hour -- I've got everything I had before, only it can be on a new arch. Case in point, I recently migrated from an x86 box on a single hard drive to an amd64 box on a RAID in this way.
Yeah, because my mom can do that. Upgrading a linux box to a new machine generally requires I manually babysit moving chunks of structure over, and do a fresh install and then fix all the settings on that fresh install that did not get copied over because they exist somewhere else. It is doable for the average Linux user, but not for the average computer user. I think the fact that current Linux users are mostly power users blinds many developers to why most Linux users are power users. A whole lot of the functionality (like this) requires you to have in depth knowledge of the system, unlike plugging in a cable and clicking the upgrade from another computer option.
Not sure what "getting in your way" means, but I can do encryption too, either on an individual account or for the entire machine.
I've tried using account level encryption on two different Linux distros but abandoned both when things stopped working properly. I ended up simply encrypting a few files I cared about. This is pretty anecdotal though, so if you say it works fine for you I'm willing to concede this one.
for instance, OpenOffice can save to PDF, and that's enough, if I ever wanted PDF in the first place.
PDFs are a great archival format. They are exacting and standard and there are multiple open and closed implementations on every platform. I don't think I've ever run across a machine that could not read them. I use PDFs for communications with customers and coworkers. I create them from my browser to archive all the online receipts and orders I make. If you only create documents in OpenOffice I can see why this would not matter much to you, but if you use other products like layout applications, vector graphics suites, etc. it is nice to always have the option without any work on the part of the developers. If applications can print, and convert postscript to PDF, why not make it easy to do
The "Migration", obviously, is usually over before you're half done with your Bagel. So what do you do now? Cheat on your employer's time, just finishing off that Bagel, or finish off that Bagel while griping about your work overload (you are a SysAdmin, right?).
Interestingly, you're mistaken in two ways. First, the upgrade actually takes quite a while when you have 60 gig of applications and data. It is slower than I would expect even. Second, I'm on salary for my primary job. So long as my work gets done, my boss won't complain if I'm playing video games or watching old movies. Not that I slack off a lot, I want those options to be worth something.
You know, this touches on one of my pet peeves. The answer to "Do you want to insert action here" in English is "Yes" or "No." The answer is not "OK" or "Continue."
The point I was making was not a grammar issue. Effective communication often runs afoul of propriety. The point was that buttons should always be labelled with actions, so that the user knows they are choosing a given action. If a message says, "The application has a memory error (OK)(Cancel)" what do those buttons do? Does "Cancel" cancel the error or the program or something else? The message needs to be rewritten with actions, "The application has a memory error. (Stop the application)(Try to keep running the application anyway)"
Also, if there's' no option, why do I have to click on something?
Agreed, this is a fairly common and well known UI snafu. It just trains users to ignore messages that are important.
Care to enumerate them?
I can name a few off the top of my head:
I'm sure there are more items I'm forgetting and again I want to stress that OS X is not ahead in all areas and can really benefit from improvements. It is just that some of these things have been on OS X for quite a while and most Linux developers I talk to don't even recognize the value in them. A lot of them are things that you can work around on Linux, or hack something that works in one instance, but until they are available to average and novice users, they are just ignored anyway. I'd love to see Linux catch up to OS X on the desktop, I just don't anticipate it happening anytime soon. I don't think Linux developers are willing to make some of the hard choices needed or will be willing to accept complexity on the server for the sake of making Linux nice on the desktop.
Major innovation will be on Linux and OS X. It will take a month or two for Linux to absorb new OS X features...
I wish. I don't see linux absorbing major new features from OS X in a month, nor many of them even in years. I like Linux. I run Linux, even on the desktop. Linux is ahead of OS X in a number of ways. But Linux is still missing a whole lot of features that lay the groundwork for what makes OS X my main desktop. Most of what I see Linux taking from OS X is minor eye candy and UI ideas.
Yeah, you hit the nail on the head here. OS X adds a few Linux/UNIX features with every release, while ignoring yet other features. Yay! traceD and virtual desktops. Boo! no ubiquitous update manager for all applications.
Yep, the games and apps people use are definitely not available on any version of OS X.
You are correct in your facts, even when you just consider games, but I think you are looking at the wrong information. Whether or not the most popular software is needed/available on OS X is not as important as if the average person wants to run software or perform a function which they cannot. There are many applications that don't have a port and while individually they may not have a lot of market share, together they account for a lot of people being stopped from doing something. Also, software piracy/lending plays a big part. A whole lot of people use games and applications they borrowed or copied from someone else, and even if there is a Mac version for sale, that does not mean there is a Mac version accessible to them.