If I forget so much as a comma, a semicolon, or a close bracket, there's a fair chance that my entire Javascript file will be thrown away. I actually strongly dislike how tolerant the Firefox parser is, as it means that this will occasionally happen only on certain other browsers, and not on Firefox.
This way we have a low barrier for entry for newbies who want to make their own page.
Newbies aren't editing raw HTML, they're using WYSIWYGs. And no WYSIWYG should be putting out invalid markup.
If newbies were editing raw HTML, then they absolutely should be seeing XML parse errors. Let them develop good habits early.
There's also user-generated and third-party content that you can't always control.
HTMLtidy. Or parse it and pull out just what you want.
That parsing part, by the way, is again made easier by actual XHTML, rather than tag soup. Especially if what you want is kept in some sort of microformat.
Finally, there are legacy pages and legacy tools that don't always output good HTML, which can't always be fixed or replaced.
If we're talking about legacy pages, those would be served as HTML, not XHTML. I see no reason they should have any bearing on which I choose to develop with for the future.
I guess what those people fear is: someone takes the source and makes a near-exact replica of a program, but with some malicious function hidden there.
This can be done for any software, anywhere. It's considerably easier to install some trojan, or crack videogame DRM, than to significantly modify and improve a program to which you don't have the source.
Security through obscurity is usually neither.
Oh, and this is quite probably illegal, certainly in the case of Firefox. If you want to make an altered version of Firefox, you can't call it Firefox without Mozilla's blessing.
Everything my package manager does, including security updates, requires my consent before any changes are made to my computer. The only consent the kill switch requires is Google's.
My understanding is that you, and everyone else, are assuming this based on a license agreement -- in which case, remember the kind of license that comes with most open source software?
NO WARRANTY, not even a guarantee of fitness for a purpose. Meaning that if they did have such a kill switch, they could use it with impunity -- just like Google.
Does anyone know what this kill switch actually looks like in operation? Are you absolutely sure it doesn't require consent in practice, whether or not the license requires your consent?
A nit: It turns out you don't know what AJAX is. It's Asynchronous Javascript and XML. Javascript is significantly different than Java.
so instead of just having http links that one can click, there are javascript callbacks that retrieve updated page content without reloading the whole page.
It's actually not all or nothing. Progressive enhancement means that you can treat it as either a hyperlink or a button to trigger JavaScript.
Example: When you reply to my comment, if you've got tabbed browsing turned on, you should be able to middle-click on the "Reply to This" button and open it in a new tab. Or you can simply click on it, and it will expand inside this page, which I find much quicker and more useful.
It's also not just hyperlinks. Example: I recently discovered Gmail's keyboard shortcuts. (Press ? in Gmail for a cheat sheet.) Even if these were real hyperlinks, I'm guessing you're still going to need Javascript to bind those hotkeys.
it breaks functionality in cases where you intend to get completely new pages.
It sucks when that happens, but that's not the AJAX itself. That's when people decide to not provide both options -- or use clever tricks like hash-links. Take Gmail -- your current location is appended to the URL, so if you want to open something in a new tab, it will reload all of Gmail -- and then go to exactly where you wanted.
Hunh? Since when is it a good idea for anyone other than the owner of a piece of hardware to decide, without the right of the owner to override, to uninstall software?
Mostly because it's via the marketplace. From TFA:
Google intends to have its Android Market be the central repository for the vast majority of mobile app distribution. Their oversight will provide users a reservoir of safe, trusted apps under the promise that they have been checked for quality, much like the promise of the App Store....
Sounds very much like what I get from the Ubuntu repositories.
Think about it -- every repository for every distro, or even every sufficiently-privileged package manager, is a kill switch for your computer. When a repository has (very occasionally) accidentally delivered a package with some sort of malware attached, that package was immediately rolled back -- effectively killing the malware. There's no reason a critical update couldn't do anything it wanted to my system -- after all, I have absolutely no warranty to fall back on.
Which means I guess we'll all have to wait and see if this applies -- or is ever used -- for software other than malware, and/or software distributed through channels other than the Marketplace.
That's the real difference -- we're all speculating about how this might work. Apple already has banned apps for no discernable reason whatsoever ("I Am Rich").
Normally when you click and hold on an object to drag it, you may reach the edge of your usable space before you've completed your drag.
I would say that this phenomenon (called "pedaling") is far more annoying than whether or not it works while dragging. Turn up your sensitivity!
Then again, given how low the sensitivity is on a default infrared Mighty Mouse and a 24" monitor, it's a legitimate concern. And I usually don't adjust the sensitivity on touchpads, which makes it that much worse.
Users who had nothing to do with the development of the page do not deserve it.
Again, you're assuming users will ever see this page. Why would they?
It's kind of like avoiding ODF because someone could leave off a closing tag. I suppose technically they could, but it's no longer something we have to hand edit.
It's also a bit like suggesting that Javascript parsers should be able to detect (and compensate for) missing close brackets.
He spends his free time lifting weights and you're a nerd. He's going to win that battle.
Unless I'm a nerd with a gun.
Good points otherwise... I'm not sure who'd be the best candidate for building a new government out of total anarchy, but I"m guessing brute physical strength won't matter.
That's the essential difference between software and politics:
And ultimately, who do appoint as our "constitution-kernel" manager to approve any constitution-kernel amendment-patches?
In software, it's fine to have a benevolent dictator of the official Linux kernel tree. That's because nothing stops you from forking this tree -- in fact, it's encouraged. And if you need people, those people can be from anywhere.
In the real world, "forking" is known as "seceding", and it involves very real property (and property disputes), and if you want to go join the new country, you have to physically go there.
Viruses - THis is not a OS problem, its a user problem.... Malware - Again not specific to Windows.
Granted.
But I'll take an OS any day that works with me on this problem, not against me. Yes, Windows is trying to improve with Vista -- too little, too late. Too many programs that won't scale to limited user accounts, no unified system-wide updates.
Prove conclusively, once and for all that X percentage of crashes are because of MS code, Vs. X percentage of crashes on Linux.
Doesn't matter. If the crashes were because of a driver (that is, the manufacturer's fault), and Linux has a more solid driver for that (but it's in-kernel, and therefore unofficial), then Linux wins that argument.
Drivers - Add all the drivers to the kernel? So the manufacturers of devices have to wait till the kernel maintainer decides on his own sweet time when to integrate patches. AND THEN wait till picks them up downstream.
This assumes several things:
- That all drivers must be in the kernel source. There are drivers which are separately maintained, just like on Windows.
- That all drivers are even in-kernel. The bulk of printer drivers, for example, are PPDs, used by CUPS -- entirely userspace.
- That the kernel maintainer is the one integrating. Nope, that's up to you. The kernel maintainer just decides if your patch makes it into the kernel. Nothing stopping you from maintaining it as a separate module, or letting distros merge it into their own kernel forks.
- That the kernel maintainer is a different person than the manufacturer.
- That the manufacturer is even involved in the process, beyond publishing specs.
- That any of this is remotely an issue for OS X. Apple pretty much gets to pick and choose what hardware will be supported, and how.
All the software in the world at a single spot. i.e. Google for applications.
Or search your package manager.
Who addresses commercial software? Who handles payments for this? Who will handle updates?
If you're on Ubuntu, the answer to all of the above is "Canonical". I can buy Parallels directly from Canonical, add an official repository, and install it (and get updates) through the same channels I install all my other software.
Or you could ask Dell how they handle the Fluendo codecs. I'm betting it's the same mechanism, though.
Do users want to download Multi GB Games/Applications?
I don't know that I've seen a multi-GB downloadable game or application for Linux. Sure, the whole distro is huge, but individual apps aren't, even when you factor in needed libraries.
Note: That was "downloadable". Most games for Linux come in a small-ish demo form through the repositories. If you buy the full game (for Windows, presumably), you can copy the game files off the disc, type in the CD key, and you've now got the full game -- but the patches still come through the package manager. Honestly, the binaries are small enough.
Who pays for the massive bandwidth?
Again, Canonical.
Are you not aware of how existing package managers work, for existing apps?
What if you're not connected online
Then, presumably, you get a disc which has the files on it. Granted, no one's built a disc that is specifically a compilation of all the demo versions of various games -- but it could be done.
I'll wait for some real responses now...
I actually like the UI.
And I like the fact that if I didn't like the UI, there are dozens (hundreds?) of window managers, all compatible enough that I can run any Linux app on them -- or write my own.
I like the fact that I can have both a rock-solid OS (and one which doesn't nag me all the time) and better desktop effects than Vi
Oddly enough, this does not seem to apply to middle-click. I tend to use middle-click to zoom, so that would work.
It's such a Lloydian design -- beautiful, awesome theory, gaping obvious holes in practice.
The scroll ball, instead of a scroll wheel -- looks cool, good idea to scroll in two dimensions -- same exact problems (down to cleaning it) as a ball mouse. Would it have killed them to put an infrared sensor on it?
And this -- touch is good. The iPhone is actually a really good idea. It's aesthetically appealing. It might even be cheaper to manufacture, and more reliable -- just guessing, could go entirely the other way there.
But not being able to right-click without sticking my left finger up like a sore... finger... that sucks.
I won't even get started on their keyboard, since I actually use that -- so much more to complain about there.
Many free as well as for $$$ Mac programs allow for automatic checking for updates.
And many don't.
On Ubuntu, almost every piece of software can be auto-updated through my package manager. (I say "almost", because I do develop software, and I do have a few programs installed from source. But that's by choice -- these programs are also available through the package manager.)
Explain to me, so I understand: Installing software is, essentially, copying files from somewhere else to your computer, and getting them set up.
Why does it matter whether the files come from a CD or from the Internet?
Maybe for someone who develops software or a power user such things might be important.
Which is the fucking point.
Unless you're forgetting something, this thread was about me, and why I might want to use Linux instead of OS X, even on a Powerbook. I think that pretty clearly answers that -- I can do things, that I want to do, that are more difficult (or impossible) on OS X.
To most people, computers like cars have a utility function.
Me too. I don't know much about cars -- I don't even change my own oil. (I don't currently own a car, but I didn't change my own oil when I did.)
But I would be very hesitant to buy a car which had the hood welded shut.
Downloading a bunch of software is not what Joe or Jane user want to have to do, before the new computer works with their new printer or scanner. They want to pop the disk that came with the device in, click "INSTALL" and then print or scan.
They'd rather pop in a CD? Really?
Never mind that much more hardware is supported by Linux, out of the box, than Windows. Very likely, you wouldn't have to download anything -- your distro would already have it preinstalled.
That is why Linux is for geeks and nerds but is in fact a great OS. Joe and Jane don't care about source code or any other code. They expect their computer, like a TV or toaster, to "just work" when they turn it on.
Downright flamebait.
You see, Linux does this, far better than Windows.
It's when you start adding things, quite unlike a TV or a toaster, that it gets complicated. For Linux, I plug in a printer and click "add printer", if it doesn't install itself automatically.
For Windows, I plug in a printer, pop in the driver CD, click "install", click "next" a bunch of times, then, more often than not, check the Internet for updates anyway.
Joe and Jane user need to be able to call an 800 number of the manufacturer or bug someone at the store they bought their shiny new computer.
I've got one: Dell.
They cannot go to a Linux store like they can to an Apple store
No, but they can go to a LUG.
Maybe some geeks enjoy being the support dept. for their relatives and friends.
My point is that people don't think about the browser, their focus is on the site they're at.
And my point is that most people do realize computers "break", and if Google tells them their computer is "broken" somehow, they'll believe it.
Because IE is part of the background, they won't think to blame IE.
Even if there was a big page from Google saying so?
IT would take longer because people have invested a lot of time in their MySpace pages and so will accept a little more pain to avoid switching but eventually more and more people would move to Facebook and MySpace would disappear.
Do you honestly believe that these people who have so much invested in Myspace -- so much so that they'll download a Myspace-specific IM client -- wouldn't install a good browser in order to use Myspace?
Your assumption is that people would simply tolerate Myspace not working, and not use it, until they'd gotten everyone switched over to Facebook. I sincerely doubt that. It seems far more likely that they'd install Firefox at least once before doing this -- and, since the web browser is so transparent to them, they'd forget that there was ever a problem.
It wasn't going anywhere. Look at the XHTML2 drafts.
XHTML and HTML5 aren't mutually exclusive.
As Opera's study confirms, 95+% out there is tag soup
No, Opera's study confirms that it's not standards-compliant. That has nothing to do with whether it's valid XML.
and this is not going to change in the coming 5 years.
And how is this an excuse for letting your own tags slide?
Supporting XHTML isn't worth it when there's barely any page out there that is actually served as XHTML (or even validates properly as such), even now, after all these years.
Supporting XHTML, and XML namespaces, even if you cheat on it with the occasional script tag, still gives you microformats. And believe it or not, browsers are not the only things trying to crawl your crappy HTML.
No user should be faced with a cryptic XML error when they want to access a page.
Nor should users be faced with a cryptic 404 error, but shit happens.
And if you put a page out there as XHTML, and left off a closing tag, you absolutely deserve for your users to see a cryptic XML error -- because it means you never bothered to check it yourself.
The web browser is a viewer, not a development tool.
Spoken like someone who has never done actual web development. Have you heard of Firebug?
-digital audio (I don't know anything about that, so I can't comment, it seems like a big deal.)
This doesn't seem like a case where bandwidth would be an issue -- more that all your existing stuff no longer works, and you need to buy some sort of adapter.
-firewire target disk mode
That's cool, but doesn't seem like _quite_ what you describe...
(huge deal for those of us supporting friends and family, even bigger for those of us who have to deploy tens of laptops at the same time. We use firewire drives to slap images on them. If you've never done this you probably don't understand the huge time saving.)
What I don't understand here is why you can't do the same via some other tech. My own preference would be gigabit ethernet (faster than FireWire anyway) and a Linux livecd. You could also build a bootable USB hard drive with an image on it, or boot from the network and do it all that way.
-firewire devices (I've invested in a few FW hard drives because of their power through bus capability, portability and speed, now they're all useless for data storage, time machine, etc.)
That's the part that really sucks -- although the last FireWire disk I saw had a USB cable as well.
I've successfully "switched" over a dozen friends and family to macs, knowing that in a pinch I could boot into FW TDM and recover their data
In a pinch, I'll boot a Linux livecd. This goes for PCs, too.
It's amazing how many people don't realize that this is possible, or that the OS is distinct from the hardware. Just heard of someone getting his hard drive replaced because he got some spyware, which rendered Windows unbootable.
Slashdot Mirror
JavaScript has no set structure, while HTML does.
How so?
If I forget so much as a comma, a semicolon, or a close bracket, there's a fair chance that my entire Javascript file will be thrown away. I actually strongly dislike how tolerant the Firefox parser is, as it means that this will occasionally happen only on certain other browsers, and not on Firefox.
This way we have a low barrier for entry for newbies who want to make their own page.
Newbies aren't editing raw HTML, they're using WYSIWYGs. And no WYSIWYG should be putting out invalid markup.
If newbies were editing raw HTML, then they absolutely should be seeing XML parse errors. Let them develop good habits early.
There's also user-generated and third-party content that you can't always control.
HTMLtidy. Or parse it and pull out just what you want.
That parsing part, by the way, is again made easier by actual XHTML, rather than tag soup. Especially if what you want is kept in some sort of microformat.
Finally, there are legacy pages and legacy tools that don't always output good HTML, which can't always be fixed or replaced.
If we're talking about legacy pages, those would be served as HTML, not XHTML. I see no reason they should have any bearing on which I choose to develop with for the future.
Combinations of even one four-letter word, when used creatively, can be quite educational.
"Fornicate that stool"?
Your mom is a [CLASSY LADY]!
a few flights down the stairs takes you into a much more uncertain and chaotic world.
And this is different than proprietary software, how?
Bonzi Buddy wasn't open source. Neither was Gator.
I guess what those people fear is: someone takes the source and makes a near-exact replica of a program, but with some malicious function hidden there.
This can be done for any software, anywhere. It's considerably easier to install some trojan, or crack videogame DRM, than to significantly modify and improve a program to which you don't have the source.
Security through obscurity is usually neither.
Oh, and this is quite probably illegal, certainly in the case of Firefox. If you want to make an altered version of Firefox, you can't call it Firefox without Mozilla's blessing.
Because shit happens?
Shit happens regardless. I'd rather have instant feedback when shit happens.
A scripting language is a totally different beast than a mark-up language.
A parser is a parser is a parser. Why should HTML be less strict than Javascript? Be more specific than "they're different"...
Everything my package manager does, including security updates, requires my consent before any changes are made to my computer. The only consent the kill switch requires is Google's.
My understanding is that you, and everyone else, are assuming this based on a license agreement -- in which case, remember the kind of license that comes with most open source software?
NO WARRANTY, not even a guarantee of fitness for a purpose. Meaning that if they did have such a kill switch, they could use it with impunity -- just like Google.
Does anyone know what this kill switch actually looks like in operation? Are you absolutely sure it doesn't require consent in practice, whether or not the license requires your consent?
A nit: It turns out you don't know what AJAX is. It's Asynchronous Javascript and XML. Javascript is significantly different than Java.
so instead of just having http links that one can click, there are javascript callbacks that retrieve updated page content without reloading the whole page.
It's actually not all or nothing. Progressive enhancement means that you can treat it as either a hyperlink or a button to trigger JavaScript.
Example: When you reply to my comment, if you've got tabbed browsing turned on, you should be able to middle-click on the "Reply to This" button and open it in a new tab. Or you can simply click on it, and it will expand inside this page, which I find much quicker and more useful.
It's also not just hyperlinks. Example: I recently discovered Gmail's keyboard shortcuts. (Press ? in Gmail for a cheat sheet.) Even if these were real hyperlinks, I'm guessing you're still going to need Javascript to bind those hotkeys.
it breaks functionality in cases where you intend to get completely new pages.
It sucks when that happens, but that's not the AJAX itself. That's when people decide to not provide both options -- or use clever tricks like hash-links. Take Gmail -- your current location is appended to the URL, so if you want to open something in a new tab, it will reload all of Gmail -- and then go to exactly where you wanted.
Hunh? Since when is it a good idea for anyone other than the owner of a piece of hardware to decide, without the right of the owner to override, to uninstall software?
Mostly because it's via the marketplace. From TFA:
Google intends to have its Android Market be the central repository for the vast majority of mobile app distribution. Their oversight will provide users a reservoir of safe, trusted apps under the promise that they have been checked for quality, much like the promise of the App Store....
Sounds very much like what I get from the Ubuntu repositories.
Think about it -- every repository for every distro, or even every sufficiently-privileged package manager, is a kill switch for your computer. When a repository has (very occasionally) accidentally delivered a package with some sort of malware attached, that package was immediately rolled back -- effectively killing the malware. There's no reason a critical update couldn't do anything it wanted to my system -- after all, I have absolutely no warranty to fall back on.
Which means I guess we'll all have to wait and see if this applies -- or is ever used -- for software other than malware, and/or software distributed through channels other than the Marketplace.
That's the real difference -- we're all speculating about how this might work. Apple already has banned apps for no discernable reason whatsoever ("I Am Rich").
The Macbook Pro has an 800 port, which makes the discussion about that vs USB legitimate again.
And even if it's only a 400 port, at least it was leaving you an upgrade path if you ever need that.
Normally when you click and hold on an object to drag it, you may reach the edge of your usable space before you've completed your drag.
I would say that this phenomenon (called "pedaling") is far more annoying than whether or not it works while dragging. Turn up your sensitivity!
Then again, given how low the sensitivity is on a default infrared Mighty Mouse and a 24" monitor, it's a legitimate concern. And I usually don't adjust the sensitivity on touchpads, which makes it that much worse.
Users who had nothing to do with the development of the page do not deserve it.
Again, you're assuming users will ever see this page. Why would they?
It's kind of like avoiding ODF because someone could leave off a closing tag. I suppose technically they could, but it's no longer something we have to hand edit.
It's also a bit like suggesting that Javascript parsers should be able to detect (and compensate for) missing close brackets.
He spends his free time lifting weights and you're a nerd. He's going to win that battle.
Unless I'm a nerd with a gun.
Good points otherwise... I'm not sure who'd be the best candidate for building a new government out of total anarchy, but I"m guessing brute physical strength won't matter.
That's the essential difference between software and politics:
And ultimately, who do appoint as our "constitution-kernel" manager to approve any constitution-kernel amendment-patches?
In software, it's fine to have a benevolent dictator of the official Linux kernel tree. That's because nothing stops you from forking this tree -- in fact, it's encouraged. And if you need people, those people can be from anywhere.
In the real world, "forking" is known as "seceding", and it involves very real property (and property disputes), and if you want to go join the new country, you have to physically go there.
Viruses - THis is not a OS problem, its a user problem.... Malware - Again not specific to Windows.
Granted.
But I'll take an OS any day that works with me on this problem, not against me. Yes, Windows is trying to improve with Vista -- too little, too late. Too many programs that won't scale to limited user accounts, no unified system-wide updates.
Prove conclusively, once and for all that X percentage of crashes are because of MS code, Vs. X percentage of crashes on Linux.
Doesn't matter. If the crashes were because of a driver (that is, the manufacturer's fault), and Linux has a more solid driver for that (but it's in-kernel, and therefore unofficial), then Linux wins that argument.
Drivers - Add all the drivers to the kernel? So the manufacturers of devices have to wait till the kernel maintainer decides on his own sweet time when to integrate patches. AND THEN wait till picks them up downstream.
This assumes several things:
- That all drivers must be in the kernel source. There are drivers which are separately maintained, just like on Windows.
- That all drivers are even in-kernel. The bulk of printer drivers, for example, are PPDs, used by CUPS -- entirely userspace.
- That the kernel maintainer is the one integrating. Nope, that's up to you. The kernel maintainer just decides if your patch makes it into the kernel. Nothing stopping you from maintaining it as a separate module, or letting distros merge it into their own kernel forks.
- That the kernel maintainer is a different person than the manufacturer.
- That the manufacturer is even involved in the process, beyond publishing specs.
- That any of this is remotely an issue for OS X. Apple pretty much gets to pick and choose what hardware will be supported, and how.
All the software in the world at a single spot. i.e. Google for applications.
Or search your package manager.
Who addresses commercial software? Who handles payments for this? Who will handle updates?
If you're on Ubuntu, the answer to all of the above is "Canonical". I can buy Parallels directly from Canonical, add an official repository, and install it (and get updates) through the same channels I install all my other software.
Or you could ask Dell how they handle the Fluendo codecs. I'm betting it's the same mechanism, though.
Do users want to download Multi GB Games/Applications?
I don't know that I've seen a multi-GB downloadable game or application for Linux. Sure, the whole distro is huge, but individual apps aren't, even when you factor in needed libraries.
Note: That was "downloadable". Most games for Linux come in a small-ish demo form through the repositories. If you buy the full game (for Windows, presumably), you can copy the game files off the disc, type in the CD key, and you've now got the full game -- but the patches still come through the package manager. Honestly, the binaries are small enough.
Who pays for the massive bandwidth?
Again, Canonical.
Are you not aware of how existing package managers work, for existing apps?
What if you're not connected online
Then, presumably, you get a disc which has the files on it. Granted, no one's built a disc that is specifically a compilation of all the demo versions of various games -- but it could be done.
I'll wait for some real responses now...
I actually like the UI.
And I like the fact that if I didn't like the UI, there are dozens (hundreds?) of window managers, all compatible enough that I can run any Linux app on them -- or write my own.
I like the fact that I can have both a rock-solid OS (and one which doesn't nag me all the time) and better desktop effects than Vi
And you're free to pick one. On KDE, I can also pick one theme, and have it automatically applied to the two major widget sets.
It ends up being quite a lot more uniform, for the apps I actually use, than Windows ever was.
Oddly enough, this does not seem to apply to middle-click. I tend to use middle-click to zoom, so that would work.
It's such a Lloydian design -- beautiful, awesome theory, gaping obvious holes in practice.
The scroll ball, instead of a scroll wheel -- looks cool, good idea to scroll in two dimensions -- same exact problems (down to cleaning it) as a ball mouse. Would it have killed them to put an infrared sensor on it?
And this -- touch is good. The iPhone is actually a really good idea. It's aesthetically appealing. It might even be cheaper to manufacture, and more reliable -- just guessing, could go entirely the other way there.
But not being able to right-click without sticking my left finger up like a sore... finger... that sucks.
I won't even get started on their keyboard, since I actually use that -- so much more to complain about there.
somebody gets blamed, and it isn't the hobby developer who wrote the app.
And why the hell not? That's usually what happens on desktop computers.
Well, except when the user gets blamed. Good old Bonzi Buddy.
Many free as well as for $$$ Mac programs allow for automatic checking for updates.
And many don't.
On Ubuntu, almost every piece of software can be auto-updated through my package manager. (I say "almost", because I do develop software, and I do have a few programs installed from source. But that's by choice -- these programs are also available through the package manager.)
Explain to me, so I understand: Installing software is, essentially, copying files from somewhere else to your computer, and getting them set up.
Why does it matter whether the files come from a CD or from the Internet?
Maybe for someone who develops software or a power user such things might be important.
Which is the fucking point.
Unless you're forgetting something, this thread was about me, and why I might want to use Linux instead of OS X, even on a Powerbook. I think that pretty clearly answers that -- I can do things, that I want to do, that are more difficult (or impossible) on OS X.
To most people, computers like cars have a utility function.
Me too. I don't know much about cars -- I don't even change my own oil. (I don't currently own a car, but I didn't change my own oil when I did.)
But I would be very hesitant to buy a car which had the hood welded shut.
Downloading a bunch of software is not what Joe or Jane user want to have to do, before the new computer works with their new printer or scanner. They want to pop the disk that came with the device in, click "INSTALL" and then print or scan.
They'd rather pop in a CD? Really?
Never mind that much more hardware is supported by Linux, out of the box, than Windows. Very likely, you wouldn't have to download anything -- your distro would already have it preinstalled.
That is why Linux is for geeks and nerds but is in fact a great OS. Joe and Jane don't care about source code or any other code. They expect their computer, like a TV or toaster, to "just work" when they turn it on.
Downright flamebait.
You see, Linux does this, far better than Windows.
It's when you start adding things, quite unlike a TV or a toaster, that it gets complicated. For Linux, I plug in a printer and click "add printer", if it doesn't install itself automatically.
For Windows, I plug in a printer, pop in the driver CD, click "install", click "next" a bunch of times, then, more often than not, check the Internet for updates anyway.
Joe and Jane user need to be able to call an 800 number of the manufacturer or bug someone at the store they bought their shiny new computer.
I've got one: Dell.
They cannot go to a Linux store like they can to an Apple store
No, but they can go to a LUG.
Maybe some geeks enjoy being the support dept. for their relatives and friends.
Once I set up Ubuntu, I don't have many calls.
Actually, Gmail works perfectly fine without Javascript enabled, they have a basic HTML mode just for that.
Gmail without Javascript is pretty much an entirely different app.
Though I admit, I do use it with Javascript on.
So you admit that it's not always a crutch.
My point is that people don't think about the browser, their focus is on the site they're at.
And my point is that most people do realize computers "break", and if Google tells them their computer is "broken" somehow, they'll believe it.
Because IE is part of the background, they won't think to blame IE.
Even if there was a big page from Google saying so?
IT would take longer because people have invested a lot of time in their MySpace pages and so will accept a little more pain to avoid switching but eventually more and more people would move to Facebook and MySpace would disappear.
Do you honestly believe that these people who have so much invested in Myspace -- so much so that they'll download a Myspace-specific IM client -- wouldn't install a good browser in order to use Myspace?
Your assumption is that people would simply tolerate Myspace not working, and not use it, until they'd gotten everyone switched over to Facebook. I sincerely doubt that. It seems far more likely that they'd install Firefox at least once before doing this -- and, since the web browser is so transparent to them, they'd forget that there was ever a problem.
It wasn't going anywhere. Look at the XHTML2 drafts.
XHTML and HTML5 aren't mutually exclusive.
As Opera's study confirms, 95+% out there is tag soup
No, Opera's study confirms that it's not standards-compliant. That has nothing to do with whether it's valid XML.
and this is not going to change in the coming 5 years.
And how is this an excuse for letting your own tags slide?
Supporting XHTML isn't worth it when there's barely any page out there that is actually served as XHTML (or even validates properly as such), even now, after all these years.
Supporting XHTML, and XML namespaces, even if you cheat on it with the occasional script tag, still gives you microformats. And believe it or not, browsers are not the only things trying to crawl your crappy HTML.
No user should be faced with a cryptic XML error when they want to access a page.
Nor should users be faced with a cryptic 404 error, but shit happens.
And if you put a page out there as XHTML, and left off a closing tag, you absolutely deserve for your users to see a cryptic XML error -- because it means you never bothered to check it yourself.
The web browser is a viewer, not a development tool.
Spoken like someone who has never done actual web development. Have you heard of Firebug?
-digital audio (I don't know anything about that, so I can't comment, it seems like a big deal.)
This doesn't seem like a case where bandwidth would be an issue -- more that all your existing stuff no longer works, and you need to buy some sort of adapter.
-firewire target disk mode
That's cool, but doesn't seem like _quite_ what you describe...
(huge deal for those of us supporting friends and family, even bigger for those of us who have to deploy tens of laptops at the same time. We use firewire drives to slap images on them. If you've never done this you probably don't understand the huge time saving.)
What I don't understand here is why you can't do the same via some other tech. My own preference would be gigabit ethernet (faster than FireWire anyway) and a Linux livecd. You could also build a bootable USB hard drive with an image on it, or boot from the network and do it all that way.
-firewire devices (I've invested in a few FW hard drives because of their power through bus capability, portability and speed, now they're all useless for data storage, time machine, etc.)
That's the part that really sucks -- although the last FireWire disk I saw had a USB cable as well.
I've successfully "switched" over a dozen friends and family to macs, knowing that in a pinch I could boot into FW TDM and recover their data
In a pinch, I'll boot a Linux livecd. This goes for PCs, too.
It's amazing how many people don't realize that this is possible, or that the OS is distinct from the hardware. Just heard of someone getting his hard drive replaced because he got some spyware, which rendered Windows unbootable.