well, it was an Irish woman out of London, but search for Anne Murphy. not that she wanted to blow up the plane, she was just (unknowingly) a bomb carrier.
Because, *obviously*, the same criteria should be used to identify potential threats everywhere, no matter what the situation. Never should the actual nuances of the potential perpetrators, victims and methodologies be taken into account.
It is, until your adversaries figure this out, at which point finding a young swedish radical blonde convert and buying her a round the world ticket with a credit card becomes incredibly easy.
Even then, you'd still have more success ignoring such a demographic, because statistics says it would be so uncommon.
If one Swedish girl takes down a plane, but ten such incidents are stopped by targeting Muslim youths, you still come out ahead. Harsh, but true.
Indeed, Christians have a very long history of atrocities, conquest, genocide, repression,
and the list goes on.
No argument from me there. Of course, when we look at data a little more relevant to modern society, Christians, on average, are not a particularly major threat.
Indeed! Many would include USA.
Myself included.
(Hint: I'm not American.)
This profiling reminds me more of racism than anything.
There is a vast gulf of difference between isolating certain people because of the colour of their skin, and isolating certain people because of their culture. No mattter how much "the world is a simple black and white picture" people like to argue otherwise.
I believe that profile-driven searches are flawed. The flaw is that the attacker can always avoid the profile you're trying to detect. For example, if I profile for young Muslim men with turbans the attacker can simply pick disaffected white middle-class women.
"Profiling" - for security reasons - is generally not performed on anything close to the level of simplicity most people criticising it think it is.
"Profiling" - performed properly - helps by directing scarce resources where they are most likely to produce a positive result.
Suggesting that a young single woman flying out of Sweden with a round-the-world ticket is equally as likely to hijack/destroy the plane as a group of young single "middle eastern" men with one-way tickets flyng out of Saudi Arabia, is letting your idealistic bleeding-heart-leftist-stupidity get in the way of common sense.
Not to mention, *everyone* "profiles", every day, all the time. It's impossible (not to mention stupid) not to. Some people just can't admit to it.
Those words transcend race, religion and colour. We should not judge because a man reads the Koran any more than we should judge because he is Black. Muslims are not terrorists.
Selecting *solely* because of skin colour I can certainly agree with, because it is both a) out of a person's control and b) utterly irrelevant to how a person behaves.
However, the same cannot be said of religion.
How a person behaves is strongly influenced by their culture. Unfortunately, in many parts of the world, intolerant and short-sighted religious beliefs are a significant contributor to culture.
Skin colour (more accurately, ethnicity) is not. However, there are many areas in the world where ethnicity and culture are strongly correlated. To ignore this - or, even worse, actively deny it - is folly.
I must be the only one here who prefers the "nipple" to a trackpad. I tend to have more control with it than the trackpad, and it's quicker to get to than a trackpad, as it's in the middle of the keyboard. I barely have to shift my hand to be able to control it.
Aside from the fact that we're talking about hundreds of years of deliberate human selection, not thousands of years of evolution, no.
I think you'll find selective breeding has been going on a hell of a lot more than "hundreds of years".
I find it amazing you see no difference between seective breeding and splicing together genes in ways that could never occur in nature and how the latter may produce dangers the former does not if for no other reason than the timescales involved.
oth. Either they're doing something so bad that we should declare war and start bombing, or they're behaving well enough that we should talk to them and try to benefit from negotiations and agreements.
This "they're either with us or against us" mentality is both short-sighted and unsustainable.
Speaking as the guy who has to keep the critical factory applications running for one of the largest vendors of Microsoft operating systems on planet Earth, I wish like heck that there were a version of "Windows" that didn't include so much unnecessary cruft. Yes, unnecessary. Why do I need an html rendering component on an application server? Or a database server? Or a middleware server?
If the code isn't running (and unless it's being used - in which case by definition it's necessary) then it isn't being used and reprsents nothing more than additional used disk space.
And, as I've said elsewhere, if a few tens or hundreds of megabytes worth of diskspace is of any great importance to you, then you have bigger problems to worry about.
What it means is that every time there's another security patch for that unused Piece Of S...oftware on my servers, I have to shut down operations and apply patches. And then when one of those servers fails to boot after patching (happens almost every time) we have to bring the backups online or reinstall the OS, or...
If you cannot handle a planned outage on a single machine for patching without loss of service to your clients, then your architecture and/or processes are broken.
(Arguably, the same could be said about unplanned outages, but since complete redundancy is not only very difficult but also extrememly expensive, IMHO that attitude is unreasonable.)
If you're applying patches to production systems without testing them in a staging environment first, then you deserve the problems you get.
That applies no matter what your OS is - and if you've never seen a Linux patch break everything, then you can't have been doing it for very long.
If there were fewer unused bits on the box, this would be much easier.
If the bits are unused, why are you patching them ?
Like it is on the Linux, Solaris, SCO (eek!), Non-Stop, and assorted other server OSes I have running in my data centers. (None of which has any kind of GUI installed - much less a pile of unremovable desktop apps) [...]
There's my two cents. Don't spend it all in one place.
I'd be happy to bet you several thousand dollars that your linux and unix machines have code installed on them that has never been executed - and in normal usage of the machine, never will.
Nobody is disputing that EVERYBODY wants more features included, more capabilities out of the box. That has nothing to do with software architecture.
Funny, because there's thousands (if not millions) of posts on Slashdot trying to conflate the two topics (including most of the ones here I've replied to, including yours).
You can ship a box with one kernel and 1000 applications and it will have a lot of features, and one of those features will be you can use feature 23 and not crash feature 156.
As long as the programs implementing features 23 and 156 are completely independent of each other, that's true. The increase in modularity and shared code in modern systems, however, is making such a statement less and less likely to be true.
Another will be that the developer can update feature 455 without affecting feature 11.
Unless, of course, feature 11 depends on feature 455. For example, if feature 455 was in a low level system library which feature 11 was using in an undocumented fashion.
Or you can ship a box with one huge software application on it, and it will be a disasterous, unmanageable mess in short order because every time you change one little feature, you have destroyed the whole application and replaced it with a brand-new application.
Which doesn't describe Windows (and never has). Your point ?
Apple's all-in-one focus with the Mac led them to a Mac with one huge single software application running on it, and that was the main criticism of Mac OS 9.
Ignoring that MacOS was never a "huge single software application", I'm pretty sure the main criticisms of Mac OS Classic were it's lack of pre-emptive multitasking, memory protection, pervasive multithreading and SMP support.
However, they fixed this over five years ago.
Indeed. Only ~7 years after Microsoft. Bravo for Apple's market-leading technology !
Now the Mac is a modern system of interconnections that feels like one thing but is actually many modular components, including enough parts to be UNIX-compatible.
WOW. Just like Windows NT was back in 1993. Bravo again !
That's the work that Gartner is recommending that Microsoft get started on sometime soon.
Amazing how credible Gartner is considered when they're criticising Microsoft and how incompetent they are whenever they're not.
If my "rather large contingent", you mean "extremely small group of fringe militants", I might agree with your statement.
And by "extremely small group of fringe militants", you mean "elected government", right ?
When the level of outrage from Muslim countries at the activities of muslim extremists reaches the same levels of outrage that emanate from the western world against Israel defending itself, and the actions of Muslim countries' rulers leans more towards discouraging, rather than encouraging - or conding - terrorism, I'll start to think people like you might have a point.
It's rather late to be fearing that. Farmers and gardeners have been messing with genetics and releasing the results of that into the wild for over a thousand years - it's called "selective cross-breeding". Just about everything food product you buy in the supermarket was created by science, not chance, and has been for as long as you've been alive.
And you don't see any difference in process or potential impact between getting to these end products through slightly differing iterations over thousands of years of evolution and a significant, practically instantaneous, once-off modification ?
Notable among the list of those nations is Israel.
Seems to me that Israel has a hell of a lot more reason to fear its neighbours than any of its neighbours have to fear Israel.
To the best of my knowledge, Israel has never expressed any desire to destroy any of the nations surrounding it (and it certainly has the means to do so, if it was so inclined). One cannot say the same for the countries around it.
It's worth noting that Israel could have caused a hell of a lot more damage in Lebanon then they did, had that been their actual goal.
To my mind, the only possible outcome is for the US, and by extension, its allies, is to move toward acccepting the eventuality that Iran will in due time have nuclear weapons and nothing anyone says or does is going to change that.
The US and its allies would be far, far more inclined to "get over it" if Iran didn't keep telling them all that they want to kill them.
Terrorists come out of all ethnicities and creeds.
Can't say I've seen any Swiss terrorists hit the headlines yet...
Terrorism relies on a fossilization of the mind, and a sociopathic dissociation from other people. It's got nothing inherent to do with Islam. And it's certainly stupid to say that terrorists are "often" muslims, as opposed to "often" being anything else.
Not nearly as stupid as suggesting all cultures are equally likely to produce terrorists.
Currently, "terrorists" are vastly more likely to be (or claim to be) Islamic than any other culture. This has not been true in the past and probably will not be true in the future. But it most certainly is true *now*.
Like it or not, there is a rather large contigent of the Islamic world that wants to destroy a rather large proportion of the western world.
Yes, his argument was stupid because he picked an X, Y, and Z that it just makes sense to include.
No, it's a stupid argument because everyone has a different opinion as to what should be included by default. For example, the vast majority of non-technical end users have little need for a command line shell, but the vast majority of technical users would consider it essential.
The point he didn't make: Since most computers are connected to the Internet today, you really only have to include the bare minimum amount of software to get the user online. After that, they can install the features they want.
But why should he have to ? Why should the user have to spend more time (and typically money) acquiring more functionality that can (and should) be included by default.
This is before even getting into less obvious points, like how the "bundled" software will typically be more integrated and seamless and how Microsoft would be able to compete with OS X and Linux if they weren't allowed to have feature parity in their product.
Package managers even make this easy.
Assuming they know about the package you want and have the version you need. The perpetual beta state of a large proportion of OSS software can make this very much a hit and miss affair. Not to mention it can often be quite a challenge finding a piece of software if you don't know whatever in-joke the developer used when they named it.
Same distro != same process. I think the point here is that MS really does "integrate" far too much, increasing the fragility of the system.
Microsoft doesn't "integrate" any more than the alternatives do - sometimes less - and often only add additional functionality in response to their competitors doing it first (eg: WMP, IE).
No, they got in trouble because they started to componentise Windows (a good software practice) while still trying to sell it as a monolithic blob (an evil business practice).
Ignoring your utterly unsupportable assertion this is an "evil business practice", just because a piece of software is modular, does not mean the vendor must sell individual bits of it for the end user to glue together at their leisure. You are trying to conflate the two completely different fields of software development and product sales.
They could have easily componentised Windows and made it possible to uninstall Internet Explorer, and no one would care.
Except for all the software developers who can no longer make use of the IE component, since there is no longer any guarantee it is present, or that a user-provided alternative has the same level of functionality. Thus removing one of the biggest advantages of having reusable components in the first place.
Not to mention, Microsoft themselves would also no longer be able to make wide use of the functionality provided by IE (or any other "user removable component") for the same reason. Thus, again, removing one of the main reasons for a modular software architecture.
(What I find particularly sickening is the people who criticise Windows for its architecture, but then praise Linux, OS X and others *for that exact same architecture*.)
You pretty much can arbitrarily remove packages that you installed, and reverse dependencies will clear the rest of it out.
Only under the supervision of the package manager. Just start pulling out random and arbitrary bits (as you want to do with Windows) and the system will break.
Alternatively, you might try and remove something relatively benign, but find the cascade of reverse dependencies leaves you without a X server or window manager (and *that* is assuming one of those packages hasn't got some other weird interdependency with something "essential" to the OS that the package manager won't let you remove).
What's funny, in a sad kind of way, is the sheer volume of people who can criticise Microsoft for spidery interdepencies in Windows, while completely ignoring an
I'm sorry, but I'm finding it very hard to take you seriously. I'm finding it very hard to believe you have any experience on this.
Why ? Because I have different opinions to yours ?
One of the benefits of compiling from source is that it neatly gets around most library versioning problems because it compiles against the versions of the libraries you have.
Assuming the versions of the libraries you have are compatible with the source code. *Frequently* in the OSS world, this can be a problem.
It's binary packages that have problems with library versioning.
Indeed. Largely because of the OSS world's disinterest in binary compatibility. The closed-source world does not suffer anywhere nearly as badly with library verioning problems.
Also, where are the "quirks" in "./configure && make"?
That would be when you need a specific version - or newer - of certain tools (like, say, gcc), when you need a specific version - or newer - of certain libraries, when the developer has only ever bothered testing their software on Linux, so it won't compile on FreeBSD or Solaris, when you're missing a needed library, etc, etc.
Compiling software on Linux is frequently a minefield unless you have the same distribution as the developer, or unless they've expended a significant amount of effort making sure their software works on other distros.
Yes, there may be an occasional build error or missing dependency, but since the distribution builders have to package the dependencies anyway, they're not very common.
They're quite common, particularly when you're dealing with newish software and don't already have a huge number of libraries installed and up to date.
Yeah, it's not like I would want to store more of MY data on my hard drive.
Your hard disk is so full that a few tens of MB of disk space makes a difference to whether or not you can store your data ? I find that extremely difficult to believe.
Not to mention Linux distributions are the true kings of wasting disk space on worthless pap. Praising Linux while criticising Windows for wasting disk space is, at best, hypocrisy.
LOL, no it's not. Look at Sourceforge one day. Or get a job. 99.9% of software is in Java, C++, VB, or C. Hardly state of the art high level languages.
You're using *VB* - the quintessential "programming for non-programmers" language - to try and refute my point that computing has become vastly more accessible in the last decade ? Or Java and C++ to refute my argument that developers aren't using high level languages to reduce implementation times and ease maintenance.
Good shot, mate, I think you must have taken your whole lower leg off with that one.
And if you think most software today is designed in a way to trade off size for ease of implementation, you give software designers WAY too much credit.
Find me a software project on similar scales to Linux, Office, OS X, Windows - hell, even something relatively simple like a modern game - written solely in assembler and you might just be able to start making a point.
LOL. I'm half tempted to use that as my signature. You've obviously never worked as a software developer.
Just which developers who can commit changes to the Linux kernel are you calling "idiots" ?
Well, lets say I find a Windows box, install Opera as the default browser, start up Windows Explorer, and type in "google.com" into the address bar. A logical expectation would be that the website would open in my default web browser, or that it would be an unrecognized filename. Except it doesn't do either of those. It transforms the Windows Explorer into Internet Explorer.
The behaviour is perfectly logical and, in fact, an excellent example of Windows's modular design. Explorer loads the IE component and uses it to access the webpage. As does any other application (including many non-Microsoft ones) written to do so.
...WHA? Yes, there is a concept of a kernel module.
Yes, but kernel modules work by loading into the same address space as the running kernel. Modular drivers (as in NT and OS X) work by passing messages via well defined interfaces.
This is why even a trivial change to the Linux kernel requires recompiling the whole thing (including all the modules) whereas hardware drivers on NT and OS X frequently continue working without modification even through multiple.1-level changes.
I guess that's why I have to download source code to the nVidia driver, patch my kernel, and recompile to make my video card work. Oh wait -- I don't. The nVidia driver gets compiled separately, and produces a module that is loaded on demand. If I don't want to start a GUI, I don't have to load the nVidia module.
Ironically, you've proved his point. You *do* have to download and recompile your nvidia driver every time you get a new kernel, because of how Linux kernel modules work. You do *not* need to do this with Windows or OS X.
On Windows, if I don't want to start a GUI, I'm SOL. Hell, if I don't want IE, I'm SOL.
This is completely and utterly irrelevant to a discussion about kernel architecture.
No. OS X is a hybrid, at least until they decide to nuke the microkernel parts to get a performance gain. Windows is just a monolithic kernel with enough well-defined interfaces that you can easily ship binary drivers that don't screw up the system. Want to prove it's monolithic? Any kind of driver, if it's poorly written, can crash your whole system -- just like on Linux.
Your "proof" is broken (and a driver can just as easily crash OS X - and they do).
Both NT and OS X are so-called "hybrid" kernels. Both were designed with a microkernel architecture and both have subsequently moved away from its strict requirements by running various things in kernel space rather than user space to get better performance.
Both, also, are far closer to a microkernel than Linux is, far more modular and have well defined and stable ABIs.
Windows is screwed up because Microsoft integrates tons of stuff in userland that have no business being integrated.
I challenge you to list some of these things that have been "integrated" on Windows, but not on other platforms.
The difference between Windows and Linux is that Windows tries to integrate stupid shit like web browsers and IM clients into the core of the OS (mshtml is integrated into the shell; the shell is closely tied to the GUI; the GUI is in the kernel).
The GUI is not in the kernel. Your example is broken.
Not to mention it applies equally to both Linux and OS X.
Although saying "IE is in the NT kernel" isn't quite accurate, it's a hell of a lot closer to the truth than saying "Firefox is in the Linux kernel" is!
The fact that there are decent numbers of people with no interest in those features suggests it might be more productive to make them into optional modules, no?
Explain how the cost incurred in doing this will be outweighed by the benefits.
The "benefits" in this case, being appeasing a handful of geeks who like to complain about irrelevant semantics.
You seem confused about how a typical Linux distro works.
Well, I'm not.
Distro maintainers make a valiant effort to take all the disparate, inconsistent bits of OSS software out there and glue them together hoping to end up with a product that looks like an integrated, consistent, interoperable platform. Unfortunately they usually end up with something that looks more like a patchwork quilt. I mention Ubuntu because they've probably done the best job thus far.
When a Linux distro includes a piece of software, it's just like if you had gone to the software's website and downloaded it.
Actually, it's not, largely because the distro maintainer has been nice enough to put in the hard yards for you to work around the compiling quirks and library versioning labyrinth endemic to the open source world, but also because they make at least some effort to try and integrate that software with the platform as a whole. Frequently this is little more than automatically adding a menu item, but often it goes as far as widget themes and pre-specified keyboard configurations.
Window's "extras" are built in. You have to install IE, WMP, the firewall, etc. and they can't be removed. It's quite a big difference.
No, it's a semantic difference irrelevant to pretty much everyone except Slashbot whiners.
The fact that a user can't "uninstall" something like WMP is irrelevant. If they don't want to use it, they don't have to.
I really hate the "hardware is cheap" argument in favor of bloated programs and lazy programming.
Not as much as I hate people who think I (and everyone else) should have to expend effort acquiring software that should be included by default and paying for broken development processes that favour saving dirt cheap computing resources over massively expensive human resources.
First, hardware is NOT cheap for a lot of people.
Yes, it is. If you have a computer capable of running Windows XP or Vista, then you have enough hard disk space to spare a few tens of megabytes.
Outside of gaming and high end computing, for some time now, it has typically been software that has expanded to utilise more advanced hardware, not hardware that has had to improve to cope with contemporary software.
It wouldn't be so bad if program size was being traded off for cleaner, more intuitive design.
Typically, it is.
Or if the developers were using higher-level languages and decreasing development times.
Typically, they are.
Or if the new software had some revolutionary new features.
"Revolutionary new features" are very few and far between across the entire computing world. Evolution of existing functionality is what 99% of software development produces.
But none of that is happening.
Yes, it is. The massive explosion in accessibility to computing resources over the last ten - hell, twenty - years is testament to that fact. I've little doubt this pisses you off, since all of those "unworthy" people who know stuff all about computers are actually able to make use of them, but that's the way it is.
The fact of the matter is, most programmers are idiots. It's not that they're trying to make bloated, bad software, they just don't know any better. Or they're too stupid to tell their boss they need more time to do it right. Or they're too rushed to do it the right way.
I doubt any programmer with any sort of real influence over software projects like Windows, OS X or Linux could be accurately described as an "idiot".
If there's one benefit of recent off-shoring, it's that "developer time is more expensive than hardware" will no longer be an acceptable excuse for inefficient, bloated code.
"Bloat" is simply a term used by geeks to describe features they have no interest in.
Nobody claims IE is a part of the kernel.
Yes they do. Frequently. I'm willing to bet there'll be at leas
I did, but his argument is stupid. It essentially boils down to "if Microsoft only included the things *I* think are necessary in Windows, it wouldn't be bloated. But since they include functionality X, Y and Z that I don't have any personal interest in, it's bloated".
He was saying that Windows is screwed up because Microsoft is (more or less) trying to integrate everything into the same huge process.
So is everyone else selling to the same market Windows is. Why ? Because that's what the customers in that market want.
This is in contrast to Linux, where all kinds of stuff are included on the disk but are separate programs using well-defined interfaces.
"Bloated" Linux distros like Ubuntu that are catering to the same market as Windows are *exactly* the same.
Just because Microsoft don't pander to the miniscule percentage of their customers who want to do what some Linux users do, does not mean Windows is not componentised. Hell, Microsoft got in trouble (with IE) precisely because the went down the path of componentising Windows.
Likewise, just because distro maintainers and OSS developers put mountains of effort into reducing the impact of dependency hell, doesn't mean you can just add and remove arbitrary parts of a Linux system without breaking things.
In other words, there's nothing wrong with shipping a kernel and a firewall on the same disk, but the firewall shouldn't be in the kernel!
The Linux firewall *is* in the kernel. I think you need to try and come up with a better example that I can refute.
Show me proof that the major PC vendors are voluntarily making available equal quantities of AMD and Intel-based PCs (with no sneaky, inside deals with Intel to lock out AMD) so that their customers can have the choice of which one to purchase. Then, I'll reconsider my current stance that Intel is evil.
That's a pretty fucking stupid way of deciding whether or not *intel* is evil...
You've obviously already made your conclusion and have no interest in changing your mind. Why bother pretending otherwise with ridiculous "I could be convinced otherwise if [insert stupidity here] happens" statements ? Just be honest and waste less time in your life.
IMHO, if there's anything Intel has done that you'd consider "evil", then I think you need to step outside of your mum's basement and get a bit more perspective on current events and modern history.
Slashdot Mirror
Because, *obviously*, the same criteria should be used to identify potential threats everywhere, no matter what the situation. Never should the actual nuances of the potential perpetrators, victims and methodologies be taken into account.
</SARCASM>
Even then, you'd still have more success ignoring such a demographic, because statistics says it would be so uncommon.
If one Swedish girl takes down a plane, but ten such incidents are stopped by targeting Muslim youths, you still come out ahead. Harsh, but true.
No argument from me there. Of course, when we look at data a little more relevant to modern society, Christians, on average, are not a particularly major threat.
Indeed! Many would include USA.
Myself included.
(Hint: I'm not American.)
This profiling reminds me more of racism than anything.
There is a vast gulf of difference between isolating certain people because of the colour of their skin, and isolating certain people because of their culture. No mattter how much "the world is a simple black and white picture" people like to argue otherwise.
"Profiling" - for security reasons - is generally not performed on anything close to the level of simplicity most people criticising it think it is.
"Profiling" - performed properly - helps by directing scarce resources where they are most likely to produce a positive result.
Suggesting that a young single woman flying out of Sweden with a round-the-world ticket is equally as likely to hijack/destroy the plane as a group of young single "middle eastern" men with one-way tickets flyng out of Saudi Arabia, is letting your idealistic bleeding-heart-leftist-stupidity get in the way of common sense.
Not to mention, *everyone* "profiles", every day, all the time. It's impossible (not to mention stupid) not to. Some people just can't admit to it.
Those words transcend race, religion and colour. We should not judge because a man reads the Koran any more than we should judge because he is Black. Muslims are not terrorists.
Selecting *solely* because of skin colour I can certainly agree with, because it is both a) out of a person's control and b) utterly irrelevant to how a person behaves.
However, the same cannot be said of religion.
How a person behaves is strongly influenced by their culture. Unfortunately, in many parts of the world, intolerant and short-sighted religious beliefs are a significant contributor to culture.
Skin colour (more accurately, ethnicity) is not. However, there are many areas in the world where ethnicity and culture are strongly correlated. To ignore this - or, even worse, actively deny it - is folly.
I hate them both pretty much equally.
Bring back the trackball !
I think you'll find selective breeding has been going on a hell of a lot more than "hundreds of years".
I find it amazing you see no difference between seective breeding and splicing together genes in ways that could never occur in nature and how the latter may produce dangers the former does not if for no other reason than the timescales involved.
It's quite easy to kill animals without making them "suffer".
Unless, of course, you consider "dying" and "suffering" to be synonymous. In which case I have to come back to the killing plants thing again.
Do you take use appropriate drugs when you get sick or injured ? Does the suffering of the bacteria bother you ? Just what do you mean by "suffer" ?
This "they're either with us or against us" mentality is both short-sighted and unsustainable.
If the code isn't running (and unless it's being used - in which case by definition it's necessary) then it isn't being used and reprsents nothing more than additional used disk space.
And, as I've said elsewhere, if a few tens or hundreds of megabytes worth of diskspace is of any great importance to you, then you have bigger problems to worry about.
What it means is that every time there's another security patch for that unused Piece Of S...oftware on my servers, I have to shut down operations and apply patches. And then when one of those servers fails to boot after patching (happens almost every time) we have to bring the backups online or reinstall the OS, or...
If you cannot handle a planned outage on a single machine for patching without loss of service to your clients, then your architecture and/or processes are broken.
(Arguably, the same could be said about unplanned outages, but since complete redundancy is not only very difficult but also extrememly expensive, IMHO that attitude is unreasonable.)
If you're applying patches to production systems without testing them in a staging environment first, then you deserve the problems you get.
That applies no matter what your OS is - and if you've never seen a Linux patch break everything, then you can't have been doing it for very long.
If there were fewer unused bits on the box, this would be much easier.
If the bits are unused, why are you patching them ?
Like it is on the Linux, Solaris, SCO (eek!), Non-Stop, and assorted other server OSes I have running in my data centers. (None of which has any kind of GUI installed - much less a pile of unremovable desktop apps) [...]
There's my two cents. Don't spend it all in one place.
I'd be happy to bet you several thousand dollars that your linux and unix machines have code installed on them that has never been executed - and in normal usage of the machine, never will.
Funny, because there's thousands (if not millions) of posts on Slashdot trying to conflate the two topics (including most of the ones here I've replied to, including yours).
You can ship a box with one kernel and 1000 applications and it will have a lot of features, and one of those features will be you can use feature 23 and not crash feature 156.
As long as the programs implementing features 23 and 156 are completely independent of each other, that's true. The increase in modularity and shared code in modern systems, however, is making such a statement less and less likely to be true.
Another will be that the developer can update feature 455 without affecting feature 11.
Unless, of course, feature 11 depends on feature 455. For example, if feature 455 was in a low level system library which feature 11 was using in an undocumented fashion.
Or you can ship a box with one huge software application on it, and it will be a disasterous, unmanageable mess in short order because every time you change one little feature, you have destroyed the whole application and replaced it with a brand-new application.
Which doesn't describe Windows (and never has). Your point ?
Apple's all-in-one focus with the Mac led them to a Mac with one huge single software application running on it, and that was the main criticism of Mac OS 9.
Ignoring that MacOS was never a "huge single software application", I'm pretty sure the main criticisms of Mac OS Classic were it's lack of pre-emptive multitasking, memory protection, pervasive multithreading and SMP support.
However, they fixed this over five years ago.
Indeed. Only ~7 years after Microsoft. Bravo for Apple's market-leading technology !
Now the Mac is a modern system of interconnections that feels like one thing but is actually many modular components, including enough parts to be UNIX-compatible.
WOW. Just like Windows NT was back in 1993. Bravo again !
That's the work that Gartner is recommending that Microsoft get started on sometime soon.
Amazing how credible Gartner is considered when they're criticising Microsoft and how incompetent they are whenever they're not.
NT's video drivers run in kernel space, for performance reasons (just like they do in Linux and OS X).
Video drivers != GUI.
Vista will be the first version of MS Windows with graphics outside the kernel.
NT 3.1, 3.5 and 3.51 all shipped with video drivers in user space.
MS famously put the NT graphics in the kernel in v4.0.
Indeed. "Famously" because they were changing from something different to something everyone else was doing.
And by "extremely small group of fringe militants", you mean "elected government", right ?
When the level of outrage from Muslim countries at the activities of muslim extremists reaches the same levels of outrage that emanate from the western world against Israel defending itself, and the actions of Muslim countries' rulers leans more towards discouraging, rather than encouraging - or conding - terrorism, I'll start to think people like you might have a point.
And you don't see any difference in process or potential impact between getting to these end products through slightly differing iterations over thousands of years of evolution and a significant, practically instantaneous, once-off modification ?
Do you consider the best interests of the US lie in communicating with a secular, civilised democracy, or a genocidal, theocratic dictatorship ?
Seems to me that Israel has a hell of a lot more reason to fear its neighbours than any of its neighbours have to fear Israel.
To the best of my knowledge, Israel has never expressed any desire to destroy any of the nations surrounding it (and it certainly has the means to do so, if it was so inclined). One cannot say the same for the countries around it.
It's worth noting that Israel could have caused a hell of a lot more damage in Lebanon then they did, had that been their actual goal.
To my mind, the only possible outcome is for the US, and by extension, its allies, is to move toward acccepting the eventuality that Iran will in due time have nuclear weapons and nothing anyone says or does is going to change that.
The US and its allies would be far, far more inclined to "get over it" if Iran didn't keep telling them all that they want to kill them.
*Generally* speaking ? What are the exceptions ? Why do they exist ?
Why do you consider eating plants to be any different to eating animals ?
Can you say the same?
No, but I don't have the need to try and make myself feel superior to other people by presenting false moral dilemmas.
Can't say I've seen any Swiss terrorists hit the headlines yet...
Terrorism relies on a fossilization of the mind, and a sociopathic dissociation from other people. It's got nothing inherent to do with Islam. And it's certainly stupid to say that terrorists are "often" muslims, as opposed to "often" being anything else.
Not nearly as stupid as suggesting all cultures are equally likely to produce terrorists.
Currently, "terrorists" are vastly more likely to be (or claim to be) Islamic than any other culture. This has not been true in the past and probably will not be true in the future. But it most certainly is true *now*.
Like it or not, there is a rather large contigent of the Islamic world that wants to destroy a rather large proportion of the western world.
No, it's a stupid argument because everyone has a different opinion as to what should be included by default. For example, the vast majority of non-technical end users have little need for a command line shell, but the vast majority of technical users would consider it essential.
The point he didn't make: Since most computers are connected to the Internet today, you really only have to include the bare minimum amount of software to get the user online. After that, they can install the features they want.
But why should he have to ? Why should the user have to spend more time (and typically money) acquiring more functionality that can (and should) be included by default.
This is before even getting into less obvious points, like how the "bundled" software will typically be more integrated and seamless and how Microsoft would be able to compete with OS X and Linux if they weren't allowed to have feature parity in their product.
Package managers even make this easy.
Assuming they know about the package you want and have the version you need. The perpetual beta state of a large proportion of OSS software can make this very much a hit and miss affair. Not to mention it can often be quite a challenge finding a piece of software if you don't know whatever in-joke the developer used when they named it.
Same distro != same process. I think the point here is that MS really does "integrate" far too much, increasing the fragility of the system.
Microsoft doesn't "integrate" any more than the alternatives do - sometimes less - and often only add additional functionality in response to their competitors doing it first (eg: WMP, IE).
No, they got in trouble because they started to componentise Windows (a good software practice) while still trying to sell it as a monolithic blob (an evil business practice).
Ignoring your utterly unsupportable assertion this is an "evil business practice", just because a piece of software is modular, does not mean the vendor must sell individual bits of it for the end user to glue together at their leisure. You are trying to conflate the two completely different fields of software development and product sales.
They could have easily componentised Windows and made it possible to uninstall Internet Explorer, and no one would care.
Except for all the software developers who can no longer make use of the IE component, since there is no longer any guarantee it is present, or that a user-provided alternative has the same level of functionality. Thus removing one of the biggest advantages of having reusable components in the first place.
Not to mention, Microsoft themselves would also no longer be able to make wide use of the functionality provided by IE (or any other "user removable component") for the same reason. Thus, again, removing one of the main reasons for a modular software architecture.
(What I find particularly sickening is the people who criticise Windows for its architecture, but then praise Linux, OS X and others *for that exact same architecture*.)
You pretty much can arbitrarily remove packages that you installed, and reverse dependencies will clear the rest of it out.
Only under the supervision of the package manager. Just start pulling out random and arbitrary bits (as you want to do with Windows) and the system will break.
Alternatively, you might try and remove something relatively benign, but find the cascade of reverse dependencies leaves you without a X server or window manager (and *that* is assuming one of those packages hasn't got some other weird interdependency with something "essential" to the OS that the package manager won't let you remove).
What's funny, in a sad kind of way, is the sheer volume of people who can criticise Microsoft for spidery interdepencies in Windows, while completely ignoring an
Why ? Because I have different opinions to yours ?
One of the benefits of compiling from source is that it neatly gets around most library versioning problems because it compiles against the versions of the libraries you have.
Assuming the versions of the libraries you have are compatible with the source code. *Frequently* in the OSS world, this can be a problem.
It's binary packages that have problems with library versioning.
Indeed. Largely because of the OSS world's disinterest in binary compatibility. The closed-source world does not suffer anywhere nearly as badly with library verioning problems.
Also, where are the "quirks" in "./configure && make"?
That would be when you need a specific version - or newer - of certain tools (like, say, gcc), when you need a specific version - or newer - of certain libraries, when the developer has only ever bothered testing their software on Linux, so it won't compile on FreeBSD or Solaris, when you're missing a needed library, etc, etc.
Compiling software on Linux is frequently a minefield unless you have the same distribution as the developer, or unless they've expended a significant amount of effort making sure their software works on other distros.
Yes, there may be an occasional build error or missing dependency, but since the distribution builders have to package the dependencies anyway, they're not very common.
They're quite common, particularly when you're dealing with newish software and don't already have a huge number of libraries installed and up to date.
Yeah, it's not like I would want to store more of MY data on my hard drive.
Your hard disk is so full that a few tens of MB of disk space makes a difference to whether or not you can store your data ? I find that extremely difficult to believe.
Not to mention Linux distributions are the true kings of wasting disk space on worthless pap. Praising Linux while criticising Windows for wasting disk space is, at best, hypocrisy.
LOL, no it's not. Look at Sourceforge one day. Or get a job. 99.9% of software is in Java, C++, VB, or C. Hardly state of the art high level languages.
You're using *VB* - the quintessential "programming for non-programmers" language - to try and refute my point that computing has become vastly more accessible in the last decade ? Or Java and C++ to refute my argument that developers aren't using high level languages to reduce implementation times and ease maintenance.
Good shot, mate, I think you must have taken your whole lower leg off with that one.
And if you think most software today is designed in a way to trade off size for ease of implementation, you give software designers WAY too much credit.
Find me a software project on similar scales to Linux, Office, OS X, Windows - hell, even something relatively simple like a modern game - written solely in assembler and you might just be able to start making a point.
LOL. I'm half tempted to use that as my signature. You've obviously never worked as a software developer.
Just which developers who can commit changes to the Linux kernel are you calling "idiots" ?
Well, lets say I find a Windows box, install Opera as the default browser, start up Windows Explorer, and type in "google.com" into the address bar. A logical expectation would be that the website would open in my default web browser, or that it would be an unrecognized filename. Except it doesn't do either of those. It transforms the Windows Explorer into Internet Explorer.
The behaviour is perfectly logical and, in fact, an excellent example of Windows's modular design. Explorer loads the IE component and uses it to access the webpage. As does any other application (including many non-Microsoft ones) written to do so.
O
Yes, but kernel modules work by loading into the same address space as the running kernel. Modular drivers (as in NT and OS X) work by passing messages via well defined interfaces.
This is why even a trivial change to the Linux kernel requires recompiling the whole thing (including all the modules) whereas hardware drivers on NT and OS X frequently continue working without modification even through multiple .1-level changes.
I guess that's why I have to download source code to the nVidia driver, patch my kernel, and recompile to make my video card work. Oh wait -- I don't. The nVidia driver gets compiled separately, and produces a module that is loaded on demand. If I don't want to start a GUI, I don't have to load the nVidia module.
Ironically, you've proved his point. You *do* have to download and recompile your nvidia driver every time you get a new kernel, because of how Linux kernel modules work. You do *not* need to do this with Windows or OS X.
On Windows, if I don't want to start a GUI, I'm SOL. Hell, if I don't want IE, I'm SOL.
This is completely and utterly irrelevant to a discussion about kernel architecture.
No. OS X is a hybrid, at least until they decide to nuke the microkernel parts to get a performance gain. Windows is just a monolithic kernel with enough well-defined interfaces that you can easily ship binary drivers that don't screw up the system. Want to prove it's monolithic? Any kind of driver, if it's poorly written, can crash your whole system -- just like on Linux.
Your "proof" is broken (and a driver can just as easily crash OS X - and they do).
Both NT and OS X are so-called "hybrid" kernels. Both were designed with a microkernel architecture and both have subsequently moved away from its strict requirements by running various things in kernel space rather than user space to get better performance.
Both, also, are far closer to a microkernel than Linux is, far more modular and have well defined and stable ABIs.
Windows is screwed up because Microsoft integrates tons of stuff in userland that have no business being integrated.
I challenge you to list some of these things that have been "integrated" on Windows, but not on other platforms.
The GUI is not in the kernel. Your example is broken.
Not to mention it applies equally to both Linux and OS X.
Although saying "IE is in the NT kernel" isn't quite accurate, it's a hell of a lot closer to the truth than saying "Firefox is in the Linux kernel" is!
Firefox is not IE's equivalent. KHTML is.
Explain how the cost incurred in doing this will be outweighed by the benefits.
The "benefits" in this case, being appeasing a handful of geeks who like to complain about irrelevant semantics.
Well, I'm not.
Distro maintainers make a valiant effort to take all the disparate, inconsistent bits of OSS software out there and glue them together hoping to end up with a product that looks like an integrated, consistent, interoperable platform. Unfortunately they usually end up with something that looks more like a patchwork quilt. I mention Ubuntu because they've probably done the best job thus far.
When a Linux distro includes a piece of software, it's just like if you had gone to the software's website and downloaded it.
Actually, it's not, largely because the distro maintainer has been nice enough to put in the hard yards for you to work around the compiling quirks and library versioning labyrinth endemic to the open source world, but also because they make at least some effort to try and integrate that software with the platform as a whole. Frequently this is little more than automatically adding a menu item, but often it goes as far as widget themes and pre-specified keyboard configurations.
Window's "extras" are built in. You have to install IE, WMP, the firewall, etc. and they can't be removed. It's quite a big difference.
No, it's a semantic difference irrelevant to pretty much everyone except Slashbot whiners.
The fact that a user can't "uninstall" something like WMP is irrelevant. If they don't want to use it, they don't have to.
I really hate the "hardware is cheap" argument in favor of bloated programs and lazy programming.
Not as much as I hate people who think I (and everyone else) should have to expend effort acquiring software that should be included by default and paying for broken development processes that favour saving dirt cheap computing resources over massively expensive human resources.
First, hardware is NOT cheap for a lot of people.
Yes, it is. If you have a computer capable of running Windows XP or Vista, then you have enough hard disk space to spare a few tens of megabytes.
Outside of gaming and high end computing, for some time now, it has typically been software that has expanded to utilise more advanced hardware, not hardware that has had to improve to cope with contemporary software.
It wouldn't be so bad if program size was being traded off for cleaner, more intuitive design.
Typically, it is.
Or if the developers were using higher-level languages and decreasing development times.
Typically, they are.
Or if the new software had some revolutionary new features.
"Revolutionary new features" are very few and far between across the entire computing world. Evolution of existing functionality is what 99% of software development produces.
But none of that is happening.
Yes, it is. The massive explosion in accessibility to computing resources over the last ten - hell, twenty - years is testament to that fact. I've little doubt this pisses you off, since all of those "unworthy" people who know stuff all about computers are actually able to make use of them, but that's the way it is.
The fact of the matter is, most programmers are idiots. It's not that they're trying to make bloated, bad software, they just don't know any better. Or they're too stupid to tell their boss they need more time to do it right. Or they're too rushed to do it the right way.
I doubt any programmer with any sort of real influence over software projects like Windows, OS X or Linux could be accurately described as an "idiot".
If there's one benefit of recent off-shoring, it's that "developer time is more expensive than hardware" will no longer be an acceptable excuse for inefficient, bloated code.
"Bloat" is simply a term used by geeks to describe features they have no interest in.
Nobody claims IE is a part of the kernel.
Yes they do. Frequently. I'm willing to bet there'll be at leas
I did, but his argument is stupid. It essentially boils down to "if Microsoft only included the things *I* think are necessary in Windows, it wouldn't be bloated. But since they include functionality X, Y and Z that I don't have any personal interest in, it's bloated".
He was saying that Windows is screwed up because Microsoft is (more or less) trying to integrate everything into the same huge process.
So is everyone else selling to the same market Windows is. Why ? Because that's what the customers in that market want.
This is in contrast to Linux, where all kinds of stuff are included on the disk but are separate programs using well-defined interfaces.
"Bloated" Linux distros like Ubuntu that are catering to the same market as Windows are *exactly* the same.
Just because Microsoft don't pander to the miniscule percentage of their customers who want to do what some Linux users do, does not mean Windows is not componentised. Hell, Microsoft got in trouble (with IE) precisely because the went down the path of componentising Windows.
Likewise, just because distro maintainers and OSS developers put mountains of effort into reducing the impact of dependency hell, doesn't mean you can just add and remove arbitrary parts of a Linux system without breaking things.
In other words, there's nothing wrong with shipping a kernel and a firewall on the same disk, but the firewall shouldn't be in the kernel!
The Linux firewall *is* in the kernel. I think you need to try and come up with a better example that I can refute.
That's a pretty fucking stupid way of deciding whether or not *intel* is evil...
You've obviously already made your conclusion and have no interest in changing your mind. Why bother pretending otherwise with ridiculous "I could be convinced otherwise if [insert stupidity here] happens" statements ? Just be honest and waste less time in your life.
IMHO, if there's anything Intel has done that you'd consider "evil", then I think you need to step outside of your mum's basement and get a bit more perspective on current events and modern history.