Sorry - but the rest of your post seemed to contradict the "if they want to": if the VERY_LARGE_VALUE_OF_N choose the bleeding edge release then its probably being mispromoted (e.g. the KDE debacle a while back when several distros pushed a major KDE upgrade when it was only intended for developers).
I think what we're both proposing is basically what many distros (e.g. Debian) already do. You can certainly choose your distro accordingly (e.g. CentOS vs. Fedora). This is really about Ubuntu not doing it this way.
Basically, c.f. http://www.debian.org/ and http://www.ubuntu.com/ - Debian headlines the stable release now 21+ months old, Ubuntu always headlines the latest 6-month release (which I', sure the maintainers regard as "stable" but do tend to introduce major infrastucture changes, unlike the Debian point releases).
The best way I can think of doing this is by letting all the VERY_LARGE_VALUE_OF_N use the bleeding-edge less-tested versions.
That works well if you're a large near-monopoly with such a stranglehold on the market that, however much you piss off your customers, they'll keep coming back.
Seriously, though, plan B is to have a bleeding-edge "test" distribution alongside your stable version aimed at geeks who, though smaller in numbers, are likely to provide more helpful feedback than "I was using my USB and it said error". Which is what Ubuntu do, except that they put in on their front page and promote it to everybody, with the LTS version buried a few links down (and not as well supplied with non-critical backports as it could be).
Unless one wants to use a new version of an application that relies on a new version of a language interpreter (e.g. PHP, Python, Perl) or a new version of a library (e.g. SDL_ttf, which added hinting and kerning controls in 2.0.10 but Ubuntu 10.10 is stuck on 2.0.9).
Well, I'd probably grab the latest PHP tarball and build it just the way I liked (which I tend to do anyway) - although doing that on a production server without a lot of testing would be batshit insane. Its an equally silly suggestion for a non-power-user who can't fix it if they break it. So lets say I wanted a well-tested build from the official repository. Trouble is, if my Ubuntu is more than 6 months old, I'll probably find that the latest version of whatever isn't on offer - I'll have to upgrade to Timorous Tapir, an even longer and riskier process.
The question you really need to ask is "why does that sort of situation arise so often in Linux, and less often in Windows or Mac OS?"
Mainstream software shouldn't rely (as opposed to be able to take advantage of) on library or interpreter less than 18 months old. Distro maintainers could statically link it; bundle a private copy of the interpreter; or just offer a version compiled --without-font-hinting.
Commercial software producers are used to this - if you were starting on a bit of SOHO software for Windows today, you'd still have to think hard before making it Windows 7 (or OSX 10.6) only, because even today, that's going to lose you sales. Understandably, if you're programming for fun rather than money you'll want to use all the latest toys - but you can't then wonder why the rest of the world prefers operating systems with 5-year lifecycles.
Exactly what I would do - or Debian stable. But the point is not about what I or anybody with a slashdot account would do: its about what a non-power-user who's OK installing packaged applications but would quail at the thought of installing an OS and ask "what?" at the suggestion of compiling a tarball.
The problem is Ubuntu is very much the flagship Linux distro at the moment, the one someone is going to recommend to your gran, or the one Buy More is going to install on its new green PC. Yet it has this break-neck 6-month release cycle and now wants to go even faster. When these end users find that they can't run Firefox n+1 as seen on TV without getting the local geek in to upgrade their entire OS and finding that their WiFi card doesn't work any more, they'll blame Linux.
Most people want updates to the Kernel/Gnome/base libraries/etc...
Er, no: most people wouldn't know what the Kernel/Gnome/base libraries/etc. were if one fell into their soup.
and they want to choose when to upgrade so they can do it when they have time to address issues.
...and under the current system they may be forced to do that when all they really want to do is run the newer version of thingyOffice with the improved.docx filter.
Regularly testing rolling packages together seems like a way to let people just apply security updates until they're ready to "upgrade" to the latest rolling package set.
That depends on the "grain size" of the package sets - currently the "grain size" is the whole OS. Now, package that into smaller chunks so that people can (e.g.) install the latest mail client without having to also install the new WiFi driver system and we're talking.
In the Window world, Nontechs using windows don't care what version of MS Office they have because they're extremely interoperable. The difference between a company owning Word 2003 and Word 2007 is
Bad example - Word 2007 has a totally different user interface to 2003 and, by default, saves in a totally different file format.
Even a non-techie would be smart enough to realise that it doesn't make sense to have 2003 running on your desktop and 2007 on your laptop. Many relative non techies (not the total lusers) are just about smart enough to install the appropriate version of Word.
Now imagine that the only way to upgrade was to install Windows 7, and the only way to downgrade was to downgrade to XP... except your desktop doesn't have win7 drivers and your laptop doesnt have XP drivers...
Here's a better idea - go for more stability, not less. If Linux is maturing as a desktop OS then there shouldn't be a need for 6 monthly, let alone daily, updates.
Here's a better idea:
1. Drop the 6 month release cycle and make LTS the default option. Then people can install an OS with a sensible lifetime.
2. Don't push any updates unless they are critical security vulnerabilities.
3. Offer optional upgrades to the major application packages, drivers etc. as they become available and where possible, and keep interdependencies to a minimum - i.e. compile them against the original distro + any vital security patches, not the latest everything (statically link them if you have to - RAM is cheap now).
The problem with the current system comes for the less technical users who want to (or are sensibly advised to) stick with the packages in the official repositories. Currently, you may find that the only "official" way to get the latest office software is to upgrade your whole fricking operating system. Its like having to take the back axle off your car in order to replace the radio.
Remember this is Linux - if we/.ers want to compile our own kernel, install the latest Firefox beta from a source tarball, reformat the drive as ext6 or scour the interwebs for a suitable.deb of the very latest LibreOffice then there's nothing stopping us. Or, we can switch to a more bleeding edge distro. However, that might work for us, but it won't work for others - and even I don't want to install a new kernel just to run the latest word processor unless it really, really needs it.
The problem is particularly bad with Ubuntu: it can't be "the Linux for the rest of us" and bleeding edge, because "the rest of us" don't want to be obliged to upgrade our whole OS every 6 months just to get the latest OpenOffice.
...its understandable with commercial software where the company depends on brining in the upgrade fees, but why should Free Software care?
...because they are lazy, selfish and react to any attempt at coercion with the feline equivalent of "fuck you, where's my dinner?"
In other words, they mimic human nature perfectly.
Sure, dogs can learn to help people cross the road and "sniff out drugs" (i.e. bark at the foreign-looking guy with the rucksack when the TSA man tugs your lead in a particular way) but they're basically subservient and frankly a bit pathetic. How can you respect that?
If you've ever watched a massive Rottweiller being intimidated by a kitten then you need to ask yourself, what's happining here? Is the kitten demonstrating its superiority, or, is it demonstrating that it is too stupid to realise the dog could bite it in half, while the dog has the smarts to know that killing the cute kitty is likely to result in a one-way trip to the animal shelter?
You do know who the main developer for CUPS is, don't you? The same configuration should work on OS X.
Yup, I know OS X uses CUPS under the hood, but the trick as described involves configuring AVAHI (the FOSS implementation of the protocol used by Apple's Bonjour) not CUPS. If you can figure out how to add a service to Bonjour in a similar way then it sounds as if it should work on OS X. I guess this is what Printopia and Fingerprint do.
(Stereotypical click-n-drool Mac users look away now)
Now, it just so happens that my printer is connected to my old PPC Mac Mini running Debian because (a) that's my "always on" system and (b) the Mac OS USB driver for my cheapo Samsung laser is b0rked.
I stumbled upon this page and, Lo!, now I can haz airprint from my iPad without even having to switch on my main Mac. This looks as if it should work from your typical hackable NAS boxen, provided it can run CUPS and AVAHI...
I love Mac OS on the desktop, but, seriously, on the server side, Linux is cookin' with gas.
If, however, you want something that Just Works on your Mac, though, Printopia is worth a look.
I thought the whole point of ARM was it was super low power for mobile devices?
The ARM didn't start off that way: the first ARM chips (back when the "A" stood for "Acorn" not "Advanced") were designed as tasty desktop/workstation processors for systems like this which could seriously put the wind up the x86s (for x in 2,3,4) of the day. There was even an add-on "accelerator" card for PCs (PDF file). The shift of emphasis to embedded/mobile processors was because Wintel had the desktop market sewn up.
while I'm sure cutting down power usage in the server room would not be a BAD thing,
For anybody with a significant sized server farm, power consumption is a very BIG thing - first you have to pay for all that energy, then you have to pay again for all the air conditioning to get rid of the resulting heat. Also, less heat means you can pack more processors into a smaller space.
considering how much software, both for Windows AND Linux, that isn't for ARM based CPUs
Possibly true for Windows, but no so much for Linux, where most software is distributed as source and there's a long tradition of supporting multiple processor architecture. Debian supports ARM, for example. Heck, how many ARM-based NAS boxes are already out there running Samba and LAMP stacks? If you have Samba, Apache, PHP/Perl/Python and MySQL or PostgreSQL then that's already a pretty big range of server apps.
..but it turns out it's just another overhyped Apple marketing press release.
The words "overhyped" and "press release" were brought to you in the same sentence by the department of redundancy department.
AFAIK the "press release" was just a banner ad on the front page of the iTunes Music Store. It was the tech blogs who pumped it up to "Apple about to launch cold-fusion powered holographic iPod with direct-to-brain playback" fever pitch.
Oh, and I'd asssumed that the "Tomorrow is just another day..." tagline was a play on some Beatles lyrics or other (not that much into them myself) just as the image was a play on a Beatles album cover.
Mind you - how freakin' much for 40 year-old recordings!!!?
The question is how successful/important the "run anywhere" aspect of Java is.
How successful/important is 415 million mobile phones in 3 months?
But iPhone (Obj C compiled to ARM native code, or interpreted ECMAScript) and Android (Java, yes, but compiled to Dalvik bytecode, not JVM) are also doing pretty well without Java's "run anywhere" trick.
My money would be on ECMAscript and what is innaccurately but conveniently referred to as "HTML5" as the future for mobile apps - but not as an alternative to the more heavyweight uses of Java.
Not quite. Java is compiled to bytecode for the JVM, Go is compiled to native machine code. The distinction between "compiled" and "interpreted" is not what it once was. The question is how successful/important the "run anywhere" aspect of Java is.
Mind you, the elephant in the room has nothing to do with paradigms and whether the language supports exceptions: the deal breaker is the huge ecosystem (and accompanying jargon thicket) of APIs and frameworks around Java. Commercial users with a lot of existing expertise in this will not be falling over themselves to re-train and re-tool.
The answer is, of course, dependent the maximum time for which you can stonewall your Pointy Haired Boss before he orders you to tell him the root password or else.
GP was talking about "brute-force over the password hash".
I.e. the rule refers to situations where someone had grabbed your/etc/passwd or equivalent (fairly easy on old-style Unix systems where it was world readble) and just has to find the original passwords that give rise to the encrypted values in that file.
What they're talking about here is the use of Steam as a copy-protection system, which is completely integrated into a game, including the retail "disk" version.
Given a choice between an old-fangled, non-DRM'd disc from the high street or Steam DRM I'd go for the first - unfortunately, that doesn't seem to be what is on offer. We worry about what value our games will have if the DRM servers go down, but, last time I tried to revisit an old disc-based game, my modern computer wouldn't read the silly fsked-about with copy protected CD anyway, so forget that.
So, if DRM is unavoidable, I'd rather have a well-thought-out system like Steam (or iTunes App Store) with a big user-base and solid industry support than a mass of different half-baked product activation and online play systems. As well as being more likely to disappear overnight, the latter tend to be far more restrictive in terms of (e.g.) letting you install the software on all the computers you own. Not to mention the ones that install rootkits and spyware.
I can see the stores' frustration, but... high-street sales of digital media are going the way of the dodo, canals, blacksmiths and photo processors so they'd be better off diversifying or planning their exit strategies than whingeing.
(Of course, "Linux" in this context is not the kernel, but "Linux+GNU+X.Org+Gnome/KDE/*+...)
Method 1: invent a time machine, go back to that day in the early 1980s when the Digital Research CEO was out flying his plane when the guys from IBM called, and get the contract for supplying the OS for the original IBM PC. That's the only way you're going to re-create Microsoft's business model.
Method 2: look instead at what has worked for Apple since the second coming of Jobs:
Bundle your Linux distro with some cool hardware.
Don't sell competetively-priced priced mini-towers: these only make money if you throw them together from whatever commodity parts are available cheaply at the time (which you can't always do if you need linux-supported components), load them with adware and do a hard-sell on extended warranties. Make small-form-factor desktops, laptops and workstations which will sell at a premium.
Polish your Linux distro so that it makes sense to end users (e.g. call your web browser "Web Browser" not "PlasmaWolf", don't have an obscure division between "Settings" and "Administration", don't have two slightly different package managers...)
Establish an App Store-like repository with minimum quality criteria (e.g. Apps should be stand-alone packages which do something useful to end-users, there should be an accurate description of what they do, they should come with online help - and that doesn't mean a man page - and, after you install them, you should get a message saying what to do next. If you think existing distros do this, you're part of the problem (although they are improving). Of course, you don't have to stop people going to the standard repositories or installing tarballs if they want.
Make sure that the system can play MP3s and common video formats out of the box. Since you're selling hardware, you can pay royalties if necessary. If you're a GPL purist, give up now.
Pay people to provide user support - its easy to get techie advice on linux online, less easy for newbies to get hand-holding (its not just the handful of flame-happy assholes, even a helpful response luke "It works for me: send me an [incomprehensible to non-techies] and I'll see if I can reproduce it." is not much use for a newbie).
Spend lots of money promoting it, and ensure that people can actually buy it (looking at you Dell, HP...)
Just because the OS is free, it doesn't mean that bundling Linux is a zero-cost option - it needs serious investment and support. The one thing the FOSS community is not interested in doing for free is providing "customer is king" support. Of course, the big PC box shifters aren't good at that either: that should be an opportunity.
Of course, (post-iMac comeback) Apple had the advantage that Macs had a (deserved or otherwise) reputation for being reliable and easy to use. Linux has the opposite customer perception. So perhaps the answer is Don't call it Linux - at least, not in large letters on the front (see: Android and all those embedded systems).
Asus got about 3/10 with the original EEEPC, but they didn't invest enough time in customising their distro and ensuring that there was a stream of interesting apps coming into their repository. Then they essentialy lost interest and started pushing Windows instead.
Luckily there are some very good Linux packages out there that support Mac very nicely for a lot of this, Netatalk, OpenLDAP (Gee where did Mac get theirs), NFS, etc...
Bingo! The Mac's unique selling points over other platforms are:
Nice desktop GUI sitting on top of Unix-y goodness.
Ability to run "industry standard" desktop applications from MS, Adobe etc.
Cool-looking designer hardware.
...none of which are a particular advantage on a box sitting in a rack in the basement.
With the trend towards "cloud" based computing rather than old-style fileservers, any old commodity x86 server with a LAMP stack can do the job.
To get a foothold on the corporate desktop, Mac (desktop) has to play nicely with Linux or Windows servers, so saying "if you want x to work get an Xserve" won't cut it. Where OS X server is used, then customers will increasingly want it running on blades or virtual machines alongside other OSs.
If you want a compute/render farm then the future is going to be GPUs rather than CPUs - that will mean a Mac Pro stuffed with graphics cards, not a 1U box.
I really did like Time Machine for ease of use
...and almost every review you see of Time Machine will say that its perfect for domestic backups by people who wouldn't otherwise bother, but no substitute for proper backup software (as would be used by someone in the market for an XServe).
it looks like Steve is getting a little, OK, a lot more arrogant, and driving quite a few of us away with it.
Or, applying Occam's Razor, it might just be that Apple (a business, with shareholders) isn't shifting enough XServes to justify the cost of producing the next revision. The existance of a few people who do use/like XServe isn't a refutation of this.
Also, why the paranoia?
First, they came for Flash, and I said nothing. Next, they came for Java, and I was silent. Then, they opened the App Store, and I bought a copy of Pages. Yesterday, they came for my bash shell. So I said "Ah, Fuck it" and installed Linux.
...doesn't really cut it as a tale of poetic justice and the need for eternal vigilance. If Steve stops making general purpose computers, then its his loss, not mine. Apple isn't a monopolist.
Why do so many articles ignore the fact that there is more to Flash than video?
Probably because, for most users, the deal-breaker is not being able to access video content on news and, er, other types of site. There are plenty of casual games and other apps in the App Store - but not many non-Flash sources of video.
Being able to access those sites or not is a pretty big deal if your out and about and need to look up information on a nearby business.
Perhaps the pundits who write these articles would avoid those sites on principle anyway. Plus, what are the chances of those sites working with a touch interface?
Or, this method of playing Flash video doesn't breach Apple's App Store rules.
Apple doesn't care about letting people watch videos on iOS (which, despite the frothings of the iTunes conspiracy theorists, happily plays non-DRMd videos in several standard formats, somewhat restricted because the CPU doesn't have the horsepower to handle software-only codecs and relies on the available hardware decoders).
What Apple does care about is limiting the potential number of security holes in Safari by not having third party scripting/VM plugins like Flash Player.
Flash is a pretty bizzarre way of doing "just" video - requiring a proprietary VM, which then runs a (possibly custom) player applet provided by the video site which, in turn, streams and displays the video. (If the Flash VM had the cojones to download and run custom codecs then it might make sense, but it relies on the VM having the correct, native codec.) People used it because it got the job done and was less annoying than RealPlayer.
Of course, Flash has other uses and is actually great for applets and casual games - although how much existing Flash content can cope with a touch interface is questionable.
Actually, given that CSS generally fails in its goal in separating layout from content
I'm OK with the idea of having some formatting "hints" in with the text - the idea that you can, and should, always separate layout from content is just too idealistic. The problem with HTML/CSS is that you just can't achieve many common formatting tasks unless the divs in the content are "just so".
The worst part is - doing it the "bad" way, with tables
Don't lose sleep - just put <table class="layout"> so so anyone tweaking the CSS can target the correct class of tables, and then make sure that your text still makes sense if you just ignore the table tags.
Then again, I am just hapring on my pet peeve: no way to, in a reasonable manner, center a box of indefinite size on the page.
The number of column-inches dedicated to the various tricks and tips necessary for the ubiqutous "banner/menu column/resizable body" page layout also adds to the impression that none of the CSS designers ever tried to use it to lay out a typical web page. Add to that the inability to define constants such as "LEFT_PADDING=2em", "HIGHLIGHT_BACKGROUND=0x99ccff" so they can be easily changed (how much processing power could that take?) the past lack of support for invaluable attributes such as display=inline-block and the way that most implementations just don't seem to care about the "page-break" attributes.
Also, what CSS really lacks is some sort of layout manager akin to the ones used for Java forms.
However, the sad truth is that although CSS really needs a redo-from-start, it seems to be the only game in town (does any browser usefully support xsl/s=xslt/fo? Even so, that's a bit rocket science for page layout).
I'm vaguely surprised (but glad) that Adobe haven't filled the vacuum with PDF-based solutions that allow robust control over page layout. I suppose they have, to the extent that many online journals rely on PDF downloads rather than trying to present the material in HTML.
Not do they cover CSS or other standards that have nothing to do with HTML5
Interesting use of the phrase "nothing to do with..." there. I guess they don't fall under the remit of the HTML5 standards committee, but they have a hell of a lot to do with whether or not a real-world webapp is going to work properly in your browser. If IE9 doesn't have a good CSS implementation then its not going to be much use (which is tough, because "good" and "CSS" don't belong in the same sentence).
but are somehow lumped under HTML5 by the likes of Apple, Google, and Microsoft.
Probably because saying "HTML5 + DOM3 + EcmaScript5 + CSS3 + H.264 + SVG +..." every time kinda ruins the flow of your press release, and nobody has yet come up with a better umbrella term (Web 3.0? Ugh. Rich Internet Application Archictecture? Yikes, no!!!)
Slashdot Mirror
You misquoted my " if they want to." as just "."
Sorry - but the rest of your post seemed to contradict the "if they want to": if the VERY_LARGE_VALUE_OF_N choose the bleeding edge release then its probably being mispromoted (e.g. the KDE debacle a while back when several distros pushed a major KDE upgrade when it was only intended for developers).
I think what we're both proposing is basically what many distros (e.g. Debian) already do. You can certainly choose your distro accordingly (e.g. CentOS vs. Fedora). This is really about Ubuntu not doing it this way.
Basically, c.f. http://www.debian.org/ and http://www.ubuntu.com/ - Debian headlines the stable release now 21+ months old, Ubuntu always headlines the latest 6-month release (which I', sure the maintainers regard as "stable" but do tend to introduce major infrastucture changes, unlike the Debian point releases).
The best way I can think of doing this is by letting all the VERY_LARGE_VALUE_OF_N use the bleeding-edge less-tested versions.
That works well if you're a large near-monopoly with such a stranglehold on the market that, however much you piss off your customers, they'll keep coming back.
Seriously, though, plan B is to have a bleeding-edge "test" distribution alongside your stable version aimed at geeks who, though smaller in numbers, are likely to provide more helpful feedback than "I was using my USB and it said error". Which is what Ubuntu do, except that they put in on their front page and promote it to everybody, with the LTS version buried a few links down (and not as well supplied with non-critical backports as it could be).
Unless one wants to use a new version of an application that relies on a new version of a language interpreter (e.g. PHP, Python, Perl) or a new version of a library (e.g. SDL_ttf, which added hinting and kerning controls in 2.0.10 but Ubuntu 10.10 is stuck on 2.0.9).
Well, I'd probably grab the latest PHP tarball and build it just the way I liked (which I tend to do anyway) - although doing that on a production server without a lot of testing would be batshit insane. Its an equally silly suggestion for a non-power-user who can't fix it if they break it. So lets say I wanted a well-tested build from the official repository. Trouble is, if my Ubuntu is more than 6 months old, I'll probably find that the latest version of whatever isn't on offer - I'll have to upgrade to Timorous Tapir, an even longer and riskier process.
The question you really need to ask is "why does that sort of situation arise so often in Linux, and less often in Windows or Mac OS?"
Mainstream software shouldn't rely (as opposed to be able to take advantage of) on library or interpreter less than 18 months old. Distro maintainers could statically link it; bundle a private copy of the interpreter; or just offer a version compiled --without-font-hinting.
Commercial software producers are used to this - if you were starting on a bit of SOHO software for Windows today, you'd still have to think hard before making it Windows 7 (or OSX 10.6) only, because even today, that's going to lose you sales. Understandably, if you're programming for fun rather than money you'll want to use all the latest toys - but you can't then wonder why the rest of the world prefers operating systems with 5-year lifecycles.
Use Red Hat/CentOS if that's what you want.
Exactly what I would do - or Debian stable. But the point is not about what I or anybody with a slashdot account would do: its about what a non-power-user who's OK installing packaged applications but would quail at the thought of installing an OS and ask "what?" at the suggestion of compiling a tarball.
The problem is Ubuntu is very much the flagship Linux distro at the moment, the one someone is going to recommend to your gran, or the one Buy More is going to install on its new green PC. Yet it has this break-neck 6-month release cycle and now wants to go even faster. When these end users find that they can't run Firefox n+1 as seen on TV without getting the local geek in to upgrade their entire OS and finding that their WiFi card doesn't work any more, they'll blame Linux.
Most people want updates to the Kernel/Gnome/base libraries/etc...
Er, no: most people wouldn't know what the Kernel/Gnome/base libraries/etc. were if one fell into their soup.
and they want to choose when to upgrade so they can do it when they have time to address issues.
...and under the current system they may be forced to do that when all they really want to do is run the newer version of thingyOffice with the improved .docx filter.
Regularly testing rolling packages together seems like a way to let people just apply security updates until they're ready to "upgrade" to the latest rolling package set.
That depends on the "grain size" of the package sets - currently the "grain size" is the whole OS. Now, package that into smaller chunks so that people can (e.g.) install the latest mail client without having to also install the new WiFi driver system and we're talking.
In the Window world, Nontechs using windows don't care what version of MS Office they have because they're extremely interoperable. The difference between a company owning Word 2003 and Word 2007 is
Bad example - Word 2007 has a totally different user interface to 2003 and, by default, saves in a totally different file format.
Even a non-techie would be smart enough to realise that it doesn't make sense to have 2003 running on your desktop and 2007 on your laptop. Many relative non techies (not the total lusers) are just about smart enough to install the appropriate version of Word.
Now imagine that the only way to upgrade was to install Windows 7, and the only way to downgrade was to downgrade to XP... except your desktop doesn't have win7 drivers and your laptop doesnt have XP drivers...
Here's a better idea - go for more stability, not less. If Linux is maturing as a desktop OS then there shouldn't be a need for 6 monthly, let alone daily, updates.
Here's a better idea:
1. Drop the 6 month release cycle and make LTS the default option. Then people can install an OS with a sensible lifetime.
2. Don't push any updates unless they are critical security vulnerabilities.
3. Offer optional upgrades to the major application packages, drivers etc. as they become available and where possible, and keep interdependencies to a minimum - i.e. compile them against the original distro + any vital security patches, not the latest everything (statically link them if you have to - RAM is cheap now).
The problem with the current system comes for the less technical users who want to (or are sensibly advised to) stick with the packages in the official repositories. Currently, you may find that the only "official" way to get the latest office software is to upgrade your whole fricking operating system. Its like having to take the back axle off your car in order to replace the radio.
Remember this is Linux - if we /.ers want to compile our own kernel, install the latest Firefox beta from a source tarball, reformat the drive as ext6 or scour the interwebs for a suitable .deb of the very latest LibreOffice then there's nothing stopping us. Or, we can switch to a more bleeding edge distro. However, that might work for us, but it won't work for others - and even I don't want to install a new kernel just to run the latest word processor unless it really, really needs it.
The problem is particularly bad with Ubuntu: it can't be "the Linux for the rest of us" and bleeding edge, because "the rest of us" don't want to be obliged to upgrade our whole OS every 6 months just to get the latest OpenOffice.
...its understandable with commercial software where the company depends on brining in the upgrade fees, but why should Free Software care?
...because they are lazy, selfish and react to any attempt at coercion with the feline equivalent of "fuck you, where's my dinner?"
In other words, they mimic human nature perfectly.
Sure, dogs can learn to help people cross the road and "sniff out drugs" (i.e. bark at the foreign-looking guy with the rucksack when the TSA man tugs your lead in a particular way) but they're basically subservient and frankly a bit pathetic. How can you respect that?
If you've ever watched a massive Rottweiller being intimidated by a kitten then you need to ask yourself, what's happining here? Is the kitten demonstrating its superiority, or, is it demonstrating that it is too stupid to realise the dog could bite it in half, while the dog has the smarts to know that killing the cute kitty is likely to result in a one-way trip to the animal shelter?
You do know who the main developer for CUPS is, don't you? The same configuration should work on OS X.
Yup, I know OS X uses CUPS under the hood, but the trick as described involves configuring AVAHI (the FOSS implementation of the protocol used by Apple's Bonjour) not CUPS. If you can figure out how to add a service to Bonjour in a similar way then it sounds as if it should work on OS X. I guess this is what Printopia and Fingerprint do.
(Stereotypical click-n-drool Mac users look away now)
Now, it just so happens that my printer is connected to my old PPC Mac Mini running Debian because (a) that's my "always on" system and (b) the Mac OS USB driver for my cheapo Samsung laser is b0rked.
I stumbled upon this page and, Lo!, now I can haz airprint from my iPad without even having to switch on my main Mac. This looks as if it should work from your typical hackable NAS boxen, provided it can run CUPS and AVAHI...
I love Mac OS on the desktop, but, seriously, on the server side, Linux is cookin' with gas.
If, however, you want something that Just Works on your Mac, though, Printopia is worth a look.
I thought the whole point of ARM was it was super low power for mobile devices?
The ARM didn't start off that way: the first ARM chips (back when the "A" stood for "Acorn" not "Advanced") were designed as tasty desktop/workstation processors for systems like this which could seriously put the wind up the x86s (for x in 2,3,4) of the day. There was even an add-on "accelerator" card for PCs (PDF file). The shift of emphasis to embedded/mobile processors was because Wintel had the desktop market sewn up.
while I'm sure cutting down power usage in the server room would not be a BAD thing,
For anybody with a significant sized server farm, power consumption is a very BIG thing - first you have to pay for all that energy, then you have to pay again for all the air conditioning to get rid of the resulting heat. Also, less heat means you can pack more processors into a smaller space.
considering how much software, both for Windows AND Linux, that isn't for ARM based CPUs
Possibly true for Windows, but no so much for Linux, where most software is distributed as source and there's a long tradition of supporting multiple processor architecture. Debian supports ARM, for example. Heck, how many ARM-based NAS boxes are already out there running Samba and LAMP stacks? If you have Samba, Apache, PHP/Perl/Python and MySQL or PostgreSQL then that's already a pretty big range of server apps.
"Hip 13044b" is a waste of a perfectly good excuse to name it Eddore.
OK, so its extra-galactic rather than extra-dimensional, but that's the closest we're likely to get, and Doc Smith had colliding galaxies, too.
..but it turns out it's just another overhyped Apple marketing press release.
The words "overhyped" and "press release" were brought to you in the same sentence by the department of redundancy department.
AFAIK the "press release" was just a banner ad on the front page of the iTunes Music Store. It was the tech blogs who pumped it up to "Apple about to launch cold-fusion powered holographic iPod with direct-to-brain playback" fever pitch.
Oh, and I'd asssumed that the "Tomorrow is just another day..." tagline was a play on some Beatles lyrics or other (not that much into them myself) just as the image was a play on a Beatles album cover.
Mind you - how freakin' much for 40 year-old recordings!!!?
The question is how successful/important the "run anywhere" aspect of Java is.
How successful/important is 415 million mobile phones in 3 months?
But iPhone (Obj C compiled to ARM native code, or interpreted ECMAScript) and Android (Java, yes, but compiled to Dalvik bytecode, not JVM) are also doing pretty well without Java's "run anywhere" trick.
My money would be on ECMAscript and what is innaccurately but conveniently referred to as "HTML5" as the future for mobile apps - but not as an alternative to the more heavyweight uses of Java.
Java is interpreted, Go is compiled.
Not quite. Java is compiled to bytecode for the JVM, Go is compiled to native machine code. The distinction between "compiled" and "interpreted" is not what it once was. The question is how successful/important the "run anywhere" aspect of Java is.
Mind you, the elephant in the room has nothing to do with paradigms and whether the language supports exceptions: the deal breaker is the huge ecosystem (and accompanying jargon thicket) of APIs and frameworks around Java. Commercial users with a lot of existing expertise in this will not be falling over themselves to re-train and re-tool.
The answer is, of course, dependent the maximum time for which you can stonewall your Pointy Haired Boss before he orders you to tell him the root password or else.
That's about how often.
What's the possible tries per second reference?
GP was talking about "brute-force over the password hash".
I.e. the rule refers to situations where someone had grabbed your /etc/passwd or equivalent (fairly easy on old-style Unix systems where it was world readble) and just has to find the original passwords that give rise to the encrypted values in that file.
What they're talking about here is the use of Steam as a copy-protection system, which is completely integrated into a game, including the retail "disk" version.
Given a choice between an old-fangled, non-DRM'd disc from the high street or Steam DRM I'd go for the first - unfortunately, that doesn't seem to be what is on offer. We worry about what value our games will have if the DRM servers go down, but, last time I tried to revisit an old disc-based game, my modern computer wouldn't read the silly fsked-about with copy protected CD anyway, so forget that.
So, if DRM is unavoidable, I'd rather have a well-thought-out system like Steam (or iTunes App Store) with a big user-base and solid industry support than a mass of different half-baked product activation and online play systems. As well as being more likely to disappear overnight, the latter tend to be far more restrictive in terms of (e.g.) letting you install the software on all the computers you own. Not to mention the ones that install rootkits and spyware.
I can see the stores' frustration, but... high-street sales of digital media are going the way of the dodo, canals, blacksmiths and photo processors so they'd be better off diversifying or planning their exit strategies than whingeing.
regarding the "chance to really get going", Gosu is already used by multi-billion dollar companies around the world already, for a bunch of years.
Sorry - you mistook a facetious joke for serious commentary. You should, however, hand in your geek card for failing to spot the Firefly reference.
The whooshing noise you heard was a battered spaceship swooping over a group of galloping horses...
(Of course, "Linux" in this context is not the kernel, but "Linux+GNU+X.Org+Gnome/KDE/*+...)
Method 1: invent a time machine, go back to that day in the early 1980s when the Digital Research CEO was out flying his plane when the guys from IBM called, and get the contract for supplying the OS for the original IBM PC. That's the only way you're going to re-create Microsoft's business model.
Method 2: look instead at what has worked for Apple since the second coming of Jobs:
Just because the OS is free, it doesn't mean that bundling Linux is a zero-cost option - it needs serious investment and support. The one thing the FOSS community is not interested in doing for free is providing "customer is king" support. Of course, the big PC box shifters aren't good at that either: that should be an opportunity.
Of course, (post-iMac comeback) Apple had the advantage that Macs had a (deserved or otherwise) reputation for being reliable and easy to use. Linux has the opposite customer perception. So perhaps the answer is Don't call it Linux - at least, not in large letters on the front (see: Android and all those embedded systems).
Asus got about 3/10 with the original EEEPC, but they didn't invest enough time in customising their distro and ensuring that there was a stream of interesting apps coming into their repository. Then they essentialy lost interest and started pushing Windows instead.
"But it IS a piece of Gosu!"
Don't worry: the language will probably be cancelled before it has a chance to really get going.
Anyway, that was my first thought too, then I wondered if it was an invitation to Oracle's lawyers...?
Luckily there are some very good Linux packages out there that support Mac very nicely for a lot of this, Netatalk, OpenLDAP (Gee where did Mac get theirs), NFS, etc...
Bingo! The Mac's unique selling points over other platforms are:
...none of which are a particular advantage on a box sitting in a rack in the basement.
With the trend towards "cloud" based computing rather than old-style fileservers, any old commodity x86 server with a LAMP stack can do the job.
To get a foothold on the corporate desktop, Mac (desktop) has to play nicely with Linux or Windows servers, so saying "if you want x to work get an Xserve" won't cut it. Where OS X server is used, then customers will increasingly want it running on blades or virtual machines alongside other OSs.
If you want a compute/render farm then the future is going to be GPUs rather than CPUs - that will mean a Mac Pro stuffed with graphics cards, not a 1U box.
I really did like Time Machine for ease of use
...and almost every review you see of Time Machine will say that its perfect for domestic backups by people who wouldn't otherwise bother, but no substitute for proper backup software (as would be used by someone in the market for an XServe).
it looks like Steve is getting a little, OK, a lot more arrogant, and driving quite a few of us away with it.
Or, applying Occam's Razor, it might just be that Apple (a business, with shareholders) isn't shifting enough XServes to justify the cost of producing the next revision. The existance of a few people who do use/like XServe isn't a refutation of this.
Also, why the paranoia?
First, they came for Flash, and I said nothing.
Next, they came for Java, and I was silent.
Then, they opened the App Store, and I bought a copy of Pages.
Yesterday, they came for my bash shell.
So I said "Ah, Fuck it" and installed Linux.
...doesn't really cut it as a tale of poetic justice and the need for eternal vigilance. If Steve stops making general purpose computers, then its his loss, not mine. Apple isn't a monopolist.
Why do so many articles ignore the fact that there is more to Flash than video?
Probably because, for most users, the deal-breaker is not being able to access video content on news and, er, other types of site. There are plenty of casual games and other apps in the App Store - but not many non-Flash sources of video.
Being able to access those sites or not is a pretty big deal if your out and about and need to look up information on a nearby business.
Perhaps the pundits who write these articles would avoid those sites on principle anyway. Plus, what are the chances of those sites working with a touch interface?
Or, this method of playing Flash video doesn't breach Apple's App Store rules.
Apple doesn't care about letting people watch videos on iOS (which, despite the frothings of the iTunes conspiracy theorists, happily plays non-DRMd videos in several standard formats, somewhat restricted because the CPU doesn't have the horsepower to handle software-only codecs and relies on the available hardware decoders).
What Apple does care about is limiting the potential number of security holes in Safari by not having third party scripting/VM plugins like Flash Player.
Flash is a pretty bizzarre way of doing "just" video - requiring a proprietary VM, which then runs a (possibly custom) player applet provided by the video site which, in turn, streams and displays the video. (If the Flash VM had the cojones to download and run custom codecs then it might make sense, but it relies on the VM having the correct, native codec.) People used it because it got the job done and was less annoying than RealPlayer.
Of course, Flash has other uses and is actually great for applets and casual games - although how much existing Flash content can cope with a touch interface is questionable.
Actually, given that CSS generally fails in its goal in separating layout from content
I'm OK with the idea of having some formatting "hints" in with the text - the idea that you can, and should, always separate layout from content is just too idealistic. The problem with HTML/CSS is that you just can't achieve many common formatting tasks unless the divs in the content are "just so".
The worst part is - doing it the "bad" way, with tables
Don't lose sleep - just put <table class="layout"> so so anyone tweaking the CSS can target the correct class of tables, and then make sure that your text still makes sense if you just ignore the table tags.
Then again, I am just hapring on my pet peeve: no way to, in a reasonable manner, center a box of indefinite size on the page.
The number of column-inches dedicated to the various tricks and tips necessary for the ubiqutous "banner/menu column/resizable body" page layout also adds to the impression that none of the CSS designers ever tried to use it to lay out a typical web page. Add to that the inability to define constants such as "LEFT_PADDING=2em", "HIGHLIGHT_BACKGROUND=0x99ccff" so they can be easily changed (how much processing power could that take?) the past lack of support for invaluable attributes such as display=inline-block and the way that most implementations just don't seem to care about the "page-break" attributes.
Also, what CSS really lacks is some sort of layout manager akin to the ones used for Java forms.
However, the sad truth is that although CSS really needs a redo-from-start, it seems to be the only game in town (does any browser usefully support xsl/s=xslt/fo? Even so, that's a bit rocket science for page layout).
I'm vaguely surprised (but glad) that Adobe haven't filled the vacuum with PDF-based solutions that allow robust control over page layout. I suppose they have, to the extent that many online journals rely on PDF downloads rather than trying to present the material in HTML.
Not do they cover CSS or other standards that have nothing to do with HTML5
Interesting use of the phrase "nothing to do with..." there. I guess they don't fall under the remit of the HTML5 standards committee, but they have a hell of a lot to do with whether or not a real-world webapp is going to work properly in your browser. If IE9 doesn't have a good CSS implementation then its not going to be much use (which is tough, because "good" and "CSS" don't belong in the same sentence).
but are somehow lumped under HTML5 by the likes of Apple, Google, and Microsoft.
Probably because saying "HTML5 + DOM3 + EcmaScript5 + CSS3 + H.264 + SVG +..." every time kinda ruins the flow of your press release, and nobody has yet come up with a better umbrella term (Web 3.0? Ugh. Rich Internet Application Archictecture? Yikes, no!!!)