-NOT MAKING ORDINARY USERS ADMINISTRATORS! (usually do to laziness because some lame app written for win95 don't work and the 'IT guy' doesn't know how to change a reg permission).
Ok. In almost all cases, not necessary for spyware.
"CEO Smithley? Yes, this is CFO Barker. Well, I was just working on my Excel numbers for our shareholder presentation, and my machine rebooted when I went out for a cup of coffee and I lost all my work. IT says something about "security holes", and how they won't stop doing this. Can we just get rid of that new CIO? He's been a pain in the ass since he got here."
-Using a "bare minimum to do what I need to do" model for security access
Sounds great. Not real practical except in the presence of competent security admins to define "what needs to be done". Not a lot of those floating around.
-Firewalls that block certain *outgoing* access as well as incoming
Useless, because of the "IE hole". IE essentially has to be allowed free access, and it's easy for applications to request IE to send data over the network. There are a ton of vectors to use.
-Disabling, not installing, etc. software and services that are unneccesary. (again, frequent IT ignorance here. Idiots who don't know anything about software installation other than to select
And you've got everything locked down and then something comes along that needs to use Active Directory. Uh, huh.
-Some modicum of Blocking/Blacklisting/etc. access to sites/services that are known to be nothing but viruses, spyware, etc.
Not a reliable blocking mechanism, and probably done by many companies.
-Education, education, education. e.g. "No Ms. Jacobs, you should not click yes to the Bonzi Buddy installer." or "No, Mr. Harris, you should not type your local network password into that website's Java popup window just because it is asking for it.
I agree that this can be done with some things, but training is expensive, and things that are obvious to someone with years of experience in the computer industry may not be to Joe User.
-A well thought, clearly-defined acceptable use policy that is enforced - including termination for serious violations
Yeah, firing a leading salesman because he clicked "OK" in a Bonzi Buddy dialog is going to go over *real* well with upper management.
There are a couple issues here.
(a) Microsoft has made many extremely poor decisions WRT remote control over the local computer. Outlook hands email off to a full-blown HTML renderer, MSIE allows to be communicated with in many ways, is tied tightly into the OS, allows popups, has been used to push ActiveX and the like. Windows runs a number of network services out of box (and Microsoft treats the solution to the exposure of their poorly-designed-from-a-security-standpoint set of on-by-default Windows networking stuff as IP-based firewalling). Many folks are stuck with this (barring something extreme like switching to Linux, which is frequently not an option). A quick change to some policy will not fix these problems.
(b) Spyware vendors are smart and computer systems are complex. I won't bet on the ability of Joe User to avoid being gulled by SpywareCo programmer Mike Assmunch.
(c) Windows does not provide good tools for analyzing what programs are doing. Linux does not provide good easy-to-use tools.
(d) Personal computer OSes (Windows, classic Mac OS) are designed around easy configuration and administration by users rather than operating like a kiosk.
(e) Users value features and performance over security (which is really hard to see and measure, anyway...most people that "sell security" in a way that can be understood by the end user are selling the illusion of security -- personal firewall vendors, Verisign in general, etc)
It would be a royal pain in the ass to write out an idiot-understandable version of what is being sent out over the network for each release of the software.
It's quite possible for someone to break out Ethereal and see exactly what is currently going over the network, if they don't need a dumbed down description (Imagine trying to explain the security implications of an IP address -- "sorta a user identifier but not really, depending upon your situation" to a grandmother). Also, given the degree to which software uses the network today, I think that such warnings would simply become overwhelming amounts of crap that are ignored (like product safety warnings are in the US today).
Now, I *would* be more than happy to see an organization get together to certify software, and businesses require that software packages they purchase be certified (and that Windows/Linux/whatnot be able to be set to disallow use of uncertified software). That doesn't make everyone everywhere miserable -- just lets the folks that want such protection have it.
Finally, pop-ups in particular are a technical problem relating mostly to Microsoft software. There is little technical reason to provide popup functionality in webpages. When I browse the web, I don't see popups, because I block unrequested windows -- it's a feature that should never, ever have been added to web browsers. It was added without a thought for security -- allowing a remote system to make my local machine unusable is simply stupid.
Seems to me that simply notifying the customer of the exact actions of the software and making removal of the software a normal process would be sufficient.
This can be done more reliably by the OS with technical means than by legal pressure on the program author.
Now, Windows may well lack said functionality, but that's life, and something to convince Microsoft to repair.
The last thing we need is more legal crap affecting software development and slowing everything down.
If someone wants to set up a certification system for software, and folks want to only purchase certified software, that's real easy to do and provides the political pressure that you want without being a pain in the ass to software developers everywhere.
Re:How fast is swapping really these days?
on
Tuning Linux VM swapping
·
· Score: 2, Interesting
And I'd like to point out that valgrind has patent-related legal problems which has (frusteratingly) kept Red Hat from including it in their distribution.
I've heard before that "you can NOT turn swap off on Windows 2000", but to hell with it, I think I'll try it when I get home tonight. I've got 756 MB of RAM, if the system crashes when I "hit the wall" so what, I don't think I will hit the wall. Any comments? Will Win2000 let me turn them all to zero min/max size? Anyone tried that before and know what the real actual implications are?
I've tried this on NT 4. The system gets very pissy about booting, and a lot of stuff doesn't start.
It *is* an interesting case that many programs are unlikely to be useful in predicting future disk accesses, but that the OS cannot use such data.
Take updatedb on my system. When updatedb runs, masses of cached directory data chew up memory on my system. This happens once a day. Now, I'm very unlikely to actually use that data -- in general, updatedb is a poor predictor. However, Linux can't figure out that updatedb shouldn't be trusted for prediction.
It might be an interesting project for an enterprising CS student with an interest in Linux to try producing a "learning" VM system -- logging which program is responsible for each page getting cached in memory, and then determining whether that page is actually used before being replaced. Store such a profile on disk ("/var/vmlog/bin/ls" or something), and you have a bright adaptive VM system.
Such a system doesn't even have to log all the data, only a randomly selected small percentage, so that it slowly gets smarter.
Such data would also provide valuable tuning data for folks who might want to tweak the VM subsystem (or, in the case of the end user, determine whether to buy more memory...)
If you *really* want to get elaborate, you could even learn to use different eviction algorithms with different programs...
The only time management ever stuck with a process was the medical company that, by law, required governmental oversight that demanded process. And you don't want to know how much we skirted process anyway. (Most of the times we built the product first, then wrote the "planning" documentation second.)
I dunno. I think that suing DaimlerChrysler for legal issues about software that they weren't using is really much more reasonable than some of the earlier SCO escapades, like declaring that they were going to go after the US government for Linux use. The balls they had in announcing the "half off discount for people registering before a specific date" was pretty good. I really think that inspiring the Wall Street Journal to have a front-page article with the title (might not be exact, as this is from memory) "This is One Battle you'll want Goliath to Win" backing Linux was a pretty impressive achievement. I mean, the WSJ is pretty laid back. Provoking the introduction of a major Windows virus that DDoSed their website happens pretty rarely -- I suspect that more people sue DC with zero chance of winning on silly lawsuits that this. I believe that this puts SCO in with the esteemed ranks of the White House and Microsoft.
Other good SCO escapades:
* Darl convinced that Linux peole were tracking him down and that his life was in danger. (actually, this was the most plausible claim on here).
* SCO's second-biggest investor trying to get rid of the management team.
* Darl claiming that PJ was an IBM shill.
* SCO using their customer list as a source of people to sue.
* Darl accidentally mixing up copyright, trademark, and patent law in interviews (and deliberately in press releases) in such blatent fashion that a typical Slashdotter could see through his mistakes.
* Managing to unseat Microsoft as #1 on the geek shitlist. I really didn't think that Microsoft would lose their throne for years to come.
* SCO managing to acquire the description "litigious bastards". Long after SCO is a memory in history books, there will still be vast archives of content containing the phrase "litigious bastards" and SCO will be inextricably tied to the phrase.
* Providing the best evil conspiracy with Microsoft *ever*. ESR's Halloween memos owe a great debt to SCO for enlivening the library.
* Getting a major female Linux character onto the visible scene -- the tireless and talented Pamela Jones. Linux got a spokesperson that (a) isn't a hairy idealist that comes off as a cross between a cult leader and a comunist, (b) doesn't write open letters to Sun's CEO saying "Red Hat roxors you because their share price is higher, nyah nyah nyah!", (c) isn't a geeky engineer who lets slip ever-so-quotable things like "I think the best way to deal with patents is by ignoring them" and "I'd support a law that ensured that geeks get laid", and (d)...well, damn, I can't think of anything to whack Bruce Perens with, but I'm sure he has some sort of skeletons in his closet.:-)
* SCO making a legal claim that necessarily makes them guilty of mass copyright infringment of Linux IP.
There were a couple of things that I *wanted* to see, but that SCO never did:
* Suing McDonalds for using Linux.
* Claiming that their Linux patches were "trade secrets" (they already tried every other form of IP claim without success).
Ok, so it would never happen, but hey, they'd save a**loads of cash on lawyers, get a ton of free publicity, and even if they did loose, there's always the inevitable appeal.
DaimlerChrysler almost certainly has in-house lawyers.
Really, though...this is probably the most entertaining thing that's happened to DC's lawyers in a long time. It'd do little more than deprive them of lunchtime conversation.
So, let me get this straight. So far this week, we've seen an article revisiting cold fusion as being possible now, and one that somebody's found Atlantis.
If I don't see one about Sasquatch being located coming up (and that he uses Linux) before the end of the week, I'm going to be terribly disappointed.
Because people were wearing them decoratively in the game, and hard-core Christians can be incredibly mule-headed about forcing their views on other people when it comes to casual use of religious symbols. What if the bad guy likes to wear clothing with a cross on it, wears a full-black outfit (not uncommon for a Japanese game) and looks like a priest? C'mon, this is the nation that had Christian fundamentalists banning Harry Potter from school libraries because it contained "witchcraft".
I wonder how feasible it would be to make this switchable. We currently have "Kid mode" switches in many games. I wonder how feasible it would be to have a "Christian mode" to disable the display of content that Christians might be offended by.
<rant>Man, I wish that people could practice their religions in peace, and not force their religious views on other people (like stuff like this and banning of abortions).</rant>
Re:GRUB is not equivalent to LILO
on
Gentoo Linux Musings
·
· Score: 2, Insightful
GRUB is complex and has different syntax than linux because it's not designed to only be a linux bootloader. It's designed to be able to boot multiple OSs (*NIX, Windows, OS/2, Be, whatever) in a consistant and OS-syntax-agnostic way. It's actually much more like Open Firmware (except that it's soft instead of firm)
I'm aware of this.
Now, if you are just booting linux, then the only real advantage it has is a boot splash screen, and that probably isn't worth the extra hassle. But, you don't have to use it! Last time I checked (1.4rc?), Gentoo supported LILO as well - instructions for it should be in the install doc right after the ones for GRUB.
Being able to specify arbitrary kernel locations has been handy. The problem is that the above guy was getting yelled at for being turned off by Gentoo because he didn't get grub the first time through, which is ridiculous. Maybe he should have used lilo -- beats me.
Finally, complaining that you have to install GRUB yourself with Gentoo when with RedHat it "Just Works" is a non-issue, because if you're using Gentoo, it's because you want to have control over everything, and want to understand how it all works. I personally like it, and also think that having to install manually is worth it for the ease of maintainance ("emerge foo"), but if you don't want to deal with the gory details of your system, just use something else instead (I recommend MacOSX - it's great!)
Oh, that's absurd and you know it. Having a configuration utility set things up intelligently initially does not preclude you from going through and understanding something. It just means that you have a working system while you learn things. I have used Linux heavily for five years, including as an administrator. If I had waited the at least two that it took to get a really good handle on things to have a usable, fully-set-up system, I'd be a grouchy old codger. Furthermore, there are some things that I use occasionally that I have *zero* interest in understanding. I've had to specifically use sendmail as an MTA before on a single system, but I have no interest in ever learning the entire sendmail syntax.
There are times when you must give up power if you want ease-of-use. RedHat/Gentoo is not one of them, though. Linux is Linux. I have my custom emacs and sawfish environments, my custom print filter, and a number of servers that do all sorts of neat things. Just having a GUI config utility available is no requirement to use it. The idea that people should use Gentoo if they want to know what they're doing is absurd. There are many excellent reasons to use Gentoo -- you may like its package management system, may dislike SuSE's focus on KDE or Red Hat's refusal to include useful software it doesn't consider Free enough (valgrind, a JVM, XFree86, etc). You may even just like the name. Using Gentoo because you want "control" , however, is like using Slackware because you want "control" -- it's just plain nonsensical.
A major in something generally does not lock you in to doing that all your life. If you go to a typical office and start asking people what they majored in, you'll get a surprisingly wide range of responses. Many people have degrees that have little to do with their actual job.
Understanding the precise mechanism of benchmarking disk I/O throughput in the 2.6 kernel is not important to an analyst. Understanding what disk I/O throughput *is* (something that improves performance of database and fileserver applications), whether it affects desireability, and the likelihood that it will be improved to a demanded level in the next six months is important.
Would you go to a bonsai tree gardener to determine whether bonsai tree fertilizer would be a good two year investment, or to a market analyst specializing in the field?
She was talking about the preinstalled server market.
Apache (which is rarely running on Windows, and is most commonly running on Linux) already dominates the webserver market, squashing IIS.
It has managed to do this with little commercial support and few preinstalled Linux boxes being sold.
It's a pretty good bet that Linux has 10% of the *new* server market right now (i.e. there are a lot of people setting up Linux boxes). The only question is whether preinstalled sales will begin to reflect actual Linux usage.
I know that the chance of this is probably awfully low, but you aren't Quinn The Eskimo of Macintosh developer fame, are you?
The whole "develop Linux apps for a living" seemed a bit unlikely, since I thought that Quinn was at Apple these days, but the name is unusual enough that I figured I'd ask.
Darn, you chose The Sims and digital cameras, two areas where Linux does fine...and someone beat me to calling you out.:-)
Yes, Linux has a weak game collection (even if you were unfortunate in choosing something that runs fine, with commercial support, in Linux), but that money goes to where the people are. If Linux starts creeping into the business and schools, game ports will follow (hopefully with better effort toward binary compatibility).
NASA operates a serious marketing engine now (not that I can blame them, since they keep getting their money taken away if they don't) -- all those "beautiful pictures of cosmic objects" are usually a bunch of radiation grabbed from somewhere up in the X-ray range that are then rammed through a mapping program, enhanced and composited, and finally fed to Photoshop or whatever image manipulation program until they look really pretty. The glowing swaths of orange breaking into seas of purple fog and stuff like that, the sort of thing that sci-fi authors sometimes get romantic about, is little more than a tribute to the artistic ability of the folks at NASA. Real astronomy isn't all that exciting or pretty, even if it is scientifically significant and an impressive body of work.
It's hard to get people to donate money for worthy causes like the EFF without engaging in scare tactics. People are pretty lethargic.
Mass media is a wonderful example of marketing -- the newspaper is trying to make news exciting so that they can sell subscripts, hold close enough to the truth so that readers will continue to be able to treat the paper as unvarnished truth, and the newspaper's sources are providing information in such a manner as to promote their own agenda.
Research, IT, organizations in companies market themselves to the head honchos for budget. Each individual person is trying to sell themselves to move up on the ladder.
Everything involves marketing these days (well, I guess it always did, but mass media improved the payoff of spending effort on good marketing). It's kinda depressing, but it's kinda hard to motivate people without taking advantage of quirks of the psyche.
I doubt that Microsoft will *die* if Linux gets big.
That doesn't mean that they will retain their current strength in desktop operating system dominance, though.
Really, though...I wonder what that would look like?
MS would probably do a port of Explorer and Office and friends over to X11. These would remain closed. They might implement hardware compatibility via porting their current libraries (which they'd presumably not open source). VB, everything moves.
You can still make a pretty proprietary platform on top of Linux. If Linux becomes a checkbox, it'll get filled, even by Microsoft.
And you know what? After all the hype about how great expee is, and how great OS X is, I still find that linux suits me just fine - although I will admit that if there were no linux, I'd probably be using OS X now.
Pretty well sums up my own opinion as well.
It's not that OS X is bad. It has some nice things in it. I just don't see it replacing either Linux or Windows. Linux generally does the things that a hacker works on at least as well as OS X. OS X has better mass-market appeal due to Apple's famous ease-of-use (though I don't think that the user interface people at Apple today come close to measuring up to the brilliant people who worked on the classic Mac OS design two decades ago). That would seem to make OS X a Windows-killer. The problem is that Apple made a conscious and measured decision to sell to a limited market. They intend to only sell to people that buy Mac hardware, rather than to try to compete on price with the PC market. That eliminates them as a market-dominating competitor, even if it means that they can live a healthy life selling a premium product.
Plus, I kinda think that Apple would make for a lousy Establishment. I mean: "Think The Same: Use Apple"? No, just wouldn't work well at all.
Slashdot Mirror
-NOT MAKING ORDINARY USERS ADMINISTRATORS! (usually do to laziness because some lame app written for win95 don't work and the 'IT guy' doesn't know how to change a reg permission).
Ok. In almost all cases, not necessary for spyware.
-Centralized, automatic, forced software upgrades.
"CEO Smithley? Yes, this is CFO Barker. Well, I was just working on my Excel numbers for our shareholder presentation, and my machine rebooted when I went out for a cup of coffee and I lost all my work. IT says something about "security holes", and how they won't stop doing this. Can we just get rid of that new CIO? He's been a pain in the ass since he got here."
-Using a "bare minimum to do what I need to do" model for security access
Sounds great. Not real practical except in the presence of competent security admins to define "what needs to be done". Not a lot of those floating around.
-Firewalls that block certain *outgoing* access as well as incoming
Useless, because of the "IE hole". IE essentially has to be allowed free access, and it's easy for applications to request IE to send data over the network. There are a ton of vectors to use.
-Disabling, not installing, etc. software and services that are unneccesary. (again, frequent IT ignorance here. Idiots who don't know anything about software installation other than to select
And you've got everything locked down and then something comes along that needs to use Active Directory. Uh, huh.
-Some modicum of Blocking/Blacklisting/etc. access to sites/services that are known to be nothing but viruses, spyware, etc.
Not a reliable blocking mechanism, and probably done by many companies.
-Education, education, education. e.g. "No Ms. Jacobs, you should not click yes to the Bonzi Buddy installer." or "No, Mr. Harris, you should not type your local network password into that website's Java popup window just because it is asking for it.
I agree that this can be done with some things, but training is expensive, and things that are obvious to someone with years of experience in the computer industry may not be to Joe User.
-A well thought, clearly-defined acceptable use policy that is enforced - including termination for serious violations
Yeah, firing a leading salesman because he clicked "OK" in a Bonzi Buddy dialog is going to go over *real* well with upper management.
There are a couple issues here.
(a) Microsoft has made many extremely poor decisions WRT remote control over the local computer. Outlook hands email off to a full-blown HTML renderer, MSIE allows to be communicated with in many ways, is tied tightly into the OS, allows popups, has been used to push ActiveX and the like. Windows runs a number of network services out of box (and Microsoft treats the solution to the exposure of their poorly-designed-from-a-security-standpoint set of on-by-default Windows networking stuff as IP-based firewalling). Many folks are stuck with this (barring something extreme like switching to Linux, which is frequently not an option). A quick change to some policy will not fix these problems.
(b) Spyware vendors are smart and computer systems are complex. I won't bet on the ability of Joe User to avoid being gulled by SpywareCo programmer Mike Assmunch.
(c) Windows does not provide good tools for analyzing what programs are doing. Linux does not provide good easy-to-use tools.
(d) Personal computer OSes (Windows, classic Mac OS) are designed around easy configuration and administration by users rather than operating like a kiosk.
(e) Users value features and performance over security (which is really hard to see and measure, anyway...most people that "sell security" in a way that can be understood by the end user are selling the illusion of security -- personal firewall vendors, Verisign in general, etc)
Uh...I write software.
It sometimes uses the network.
It would be a royal pain in the ass to write out an idiot-understandable version of what is being sent out over the network for each release of the software.
It's quite possible for someone to break out Ethereal and see exactly what is currently going over the network, if they don't need a dumbed down description (Imagine trying to explain the security implications of an IP address -- "sorta a user identifier but not really, depending upon your situation" to a grandmother). Also, given the degree to which software uses the network today, I think that such warnings would simply become overwhelming amounts of crap that are ignored (like product safety warnings are in the US today).
Now, I *would* be more than happy to see an organization get together to certify software, and businesses require that software packages they purchase be certified (and that Windows/Linux/whatnot be able to be set to disallow use of uncertified software). That doesn't make everyone everywhere miserable -- just lets the folks that want such protection have it.
Finally, pop-ups in particular are a technical problem relating mostly to Microsoft software. There is little technical reason to provide popup functionality in webpages. When I browse the web, I don't see popups, because I block unrequested windows -- it's a feature that should never, ever have been added to web browsers. It was added without a thought for security -- allowing a remote system to make my local machine unusable is simply stupid.
The FTC is *spot on* on this issue.
Seems to me that simply notifying the customer of the exact actions of the software and making removal of the software a normal process would be sufficient.
This can be done more reliably by the OS with technical means than by legal pressure on the program author.
Now, Windows may well lack said functionality, but that's life, and something to convince Microsoft to repair.
The last thing we need is more legal crap affecting software development and slowing everything down.
If someone wants to set up a certification system for software, and folks want to only purchase certified software, that's real easy to do and provides the political pressure that you want without being a pain in the ass to software developers everywhere.
And I'd like to point out that valgrind has patent-related legal problems which has (frusteratingly) kept Red Hat from including it in their distribution.
I've heard before that "you can NOT turn swap off on Windows 2000", but to hell with it, I think I'll try it when I get home tonight. I've got 756 MB of RAM, if the system crashes when I "hit the wall" so what, I don't think I will hit the wall. Any comments? Will Win2000 let me turn them all to zero min/max size? Anyone tried that before and know what the real actual implications are?
I've tried this on NT 4. The system gets very pissy about booting, and a lot of stuff doesn't start.
It *is* an interesting case that many programs are unlikely to be useful in predicting future disk accesses, but that the OS cannot use such data.
Take updatedb on my system. When updatedb runs, masses of cached directory data chew up memory on my system. This happens once a day. Now, I'm very unlikely to actually use that data -- in general, updatedb is a poor predictor. However, Linux can't figure out that updatedb shouldn't be trusted for prediction.
It might be an interesting project for an enterprising CS student with an interest in Linux to try producing a "learning" VM system -- logging which program is responsible for each page getting cached in memory, and then determining whether that page is actually used before being replaced. Store such a profile on disk ("/var/vmlog/bin/ls" or something), and you have a bright adaptive VM system.
Such a system doesn't even have to log all the data, only a randomly selected small percentage, so that it slowly gets smarter.
Such data would also provide valuable tuning data for folks who might want to tweak the VM subsystem (or, in the case of the end user, determine whether to buy more memory...)
If you *really* want to get elaborate, you could even learn to use different eviction algorithms with different programs...
2.6 is nicer for desktop use than 2.4.
It reduces starvation in the name of throughput and application startup time.
I was watching a friend play Tropico last night. His dialog, condensed, sounded something like this.
"I'm going to be Fidel Castro!"
"Hey, Russia just gave me a bunch of money! I can do anything!"
"Hey, that guy's trying to run against me for President. I'll just suppress him and his followers and rig the elections."
"Hey, the people are rising against me!"
So we have a 100% rate of questionable voting when it comes to father-son Presidencies?
The only time management ever stuck with a process was the medical company that, by law, required governmental oversight that demanded process. And you don't want to know how much we skirted process anyway. (Most of the times we built the product first, then wrote the "planning" documentation second.)
Oh joy.
I dunno. I think that suing DaimlerChrysler for legal issues about software that they weren't using is really much more reasonable than some of the earlier SCO escapades, like declaring that they were going to go after the US government for Linux use. The balls they had in announcing the "half off discount for people registering before a specific date" was pretty good. I really think that inspiring the Wall Street Journal to have a front-page article with the title (might not be exact, as this is from memory) "This is One Battle you'll want Goliath to Win" backing Linux was a pretty impressive achievement. I mean, the WSJ is pretty laid back. Provoking the introduction of a major Windows virus that DDoSed their website happens pretty rarely -- I suspect that more people sue DC with zero chance of winning on silly lawsuits that this. I believe that this puts SCO in with the esteemed ranks of the White House and Microsoft.
:-)
Other good SCO escapades:
* Darl convinced that Linux peole were tracking him down and that his life was in danger. (actually, this was the most plausible claim on here).
* SCO's second-biggest investor trying to get rid of the management team.
* Darl claiming that PJ was an IBM shill.
* SCO using their customer list as a source of people to sue.
* Darl accidentally mixing up copyright, trademark, and patent law in interviews (and deliberately in press releases) in such blatent fashion that a typical Slashdotter could see through his mistakes.
* Managing to unseat Microsoft as #1 on the geek shitlist. I really didn't think that Microsoft would lose their throne for years to come.
* SCO managing to acquire the description "litigious bastards". Long after SCO is a memory in history books, there will still be vast archives of content containing the phrase "litigious bastards" and SCO will be inextricably tied to the phrase.
* Providing the best evil conspiracy with Microsoft *ever*. ESR's Halloween memos owe a great debt to SCO for enlivening the library.
* Getting a major female Linux character onto the visible scene -- the tireless and talented Pamela Jones. Linux got a spokesperson that (a) isn't a hairy idealist that comes off as a cross between a cult leader and a comunist, (b) doesn't write open letters to Sun's CEO saying "Red Hat roxors you because their share price is higher, nyah nyah nyah!", (c) isn't a geeky engineer who lets slip ever-so-quotable things like "I think the best way to deal with patents is by ignoring them" and "I'd support a law that ensured that geeks get laid", and (d)...well, damn, I can't think of anything to whack Bruce Perens with, but I'm sure he has some sort of skeletons in his closet.
* SCO making a legal claim that necessarily makes them guilty of mass copyright infringment of Linux IP.
There were a couple of things that I *wanted* to see, but that SCO never did:
* Suing McDonalds for using Linux.
* Claiming that their Linux patches were "trade secrets" (they already tried every other form of IP claim without success).
Nevermind how stupid that sounds, but when does it matter that a non-government document is unconstitutional?
It doesn't, but it makes for good press fodder: "The GPL is Unconstitutional!" Really, doesn't that just have a *ring* to it?
Ok, so it would never happen, but hey, they'd save a**loads of cash on lawyers, get a ton of free publicity, and even if they did loose, there's always the inevitable appeal.
DaimlerChrysler almost certainly has in-house lawyers.
Really, though...this is probably the most entertaining thing that's happened to DC's lawyers in a long time. It'd do little more than deprive them of lunchtime conversation.
So, let me get this straight. So far this week, we've seen an article revisiting cold fusion as being possible now, and one that somebody's found Atlantis.
If I don't see one about Sasquatch being located coming up (and that he uses Linux) before the end of the week, I'm going to be terribly disappointed.
Uh, the main page says "Supports Windows, Linux, UNIX" right smack dab in the middle of it.
Of course, it's a tough sell to try to fight emacs and vi...
Because people were wearing them decoratively in the game, and hard-core Christians can be incredibly mule-headed about forcing their views on other people when it comes to casual use of religious symbols. What if the bad guy likes to wear clothing with a cross on it, wears a full-black outfit (not uncommon for a Japanese game) and looks like a priest? C'mon, this is the nation that had Christian fundamentalists banning Harry Potter from school libraries because it contained "witchcraft".
I wonder how feasible it would be to make this switchable. We currently have "Kid mode" switches in many games. I wonder how feasible it would be to have a "Christian mode" to disable the display of content that Christians might be offended by.
<rant>Man, I wish that people could practice their religions in peace, and not force their religious views on other people (like stuff like this and banning of abortions).</rant>
GRUB is complex and has different syntax than linux because it's not designed to only be a linux bootloader. It's designed to be able to boot multiple OSs (*NIX, Windows, OS/2, Be, whatever) in a consistant and OS-syntax-agnostic way. It's actually much more like Open Firmware (except that it's soft instead of firm)
I'm aware of this.
Now, if you are just booting linux, then the only real advantage it has is a boot splash screen, and that probably isn't worth the extra hassle. But, you don't have to use it! Last time I checked (1.4rc?), Gentoo supported LILO as well - instructions for it should be in the install doc right after the ones for GRUB.
Being able to specify arbitrary kernel locations has been handy. The problem is that the above guy was getting yelled at for being turned off by Gentoo because he didn't get grub the first time through, which is ridiculous. Maybe he should have used lilo -- beats me.
Finally, complaining that you have to install GRUB yourself with Gentoo when with RedHat it "Just Works" is a non-issue, because if you're using Gentoo, it's because you want to have control over everything, and want to understand how it all works. I personally like it, and also think that having to install manually is worth it for the ease of maintainance ("emerge foo"), but if you don't want to deal with the gory details of your system, just use something else instead (I recommend MacOSX - it's great!)
Oh, that's absurd and you know it. Having a configuration utility set things up intelligently initially does not preclude you from going through and understanding something. It just means that you have a working system while you learn things. I have used Linux heavily for five years, including as an administrator. If I had waited the at least two that it took to get a really good handle on things to have a usable, fully-set-up system, I'd be a grouchy old codger. Furthermore, there are some things that I use occasionally that I have *zero* interest in understanding. I've had to specifically use sendmail as an MTA before on a single system, but I have no interest in ever learning the entire sendmail syntax.
There are times when you must give up power if you want ease-of-use. RedHat/Gentoo is not one of them, though. Linux is Linux. I have my custom emacs and sawfish environments, my custom print filter, and a number of servers that do all sorts of neat things. Just having a GUI config utility available is no requirement to use it. The idea that people should use Gentoo if they want to know what they're doing is absurd. There are many excellent reasons to use Gentoo -- you may like its package management system, may dislike SuSE's focus on KDE or Red Hat's refusal to include useful software it doesn't consider Free enough (valgrind, a JVM, XFree86, etc). You may even just like the name. Using Gentoo because you want "control" , however, is like using Slackware because you want "control" -- it's just plain nonsensical.
A major in something generally does not lock you in to doing that all your life. If you go to a typical office and start asking people what they majored in, you'll get a surprisingly wide range of responses. Many people have degrees that have little to do with their actual job.
Understanding the precise mechanism of benchmarking disk I/O throughput in the 2.6 kernel is not important to an analyst. Understanding what disk I/O throughput *is* (something that improves performance of database and fileserver applications), whether it affects desireability, and the likelihood that it will be improved to a demanded level in the next six months is important.
Would you go to a bonsai tree gardener to determine whether bonsai tree fertilizer would be a good two year investment, or to a market analyst specializing in the field?
She was talking about the preinstalled server market.
Apache (which is rarely running on Windows, and is most commonly running on Linux) already dominates the webserver market, squashing IIS.
It has managed to do this with little commercial support and few preinstalled Linux boxes being sold.
It's a pretty good bet that Linux has 10% of the *new* server market right now (i.e. there are a lot of people setting up Linux boxes). The only question is whether preinstalled sales will begin to reflect actual Linux usage.
I know that the chance of this is probably awfully low, but you aren't Quinn The Eskimo of Macintosh developer fame, are you?
The whole "develop Linux apps for a living" seemed a bit unlikely, since I thought that Quinn was at Apple these days, but the name is unusual enough that I figured I'd ask.
1998 called...He wants his headline back.
Good, I've already debunked this claim and merely need to provide a link.
Darn, you chose The Sims and digital cameras, two areas where Linux does fine...and someone beat me to calling you out. :-)
Yes, Linux has a weak game collection (even if you were unfortunate in choosing something that runs fine, with commercial support, in Linux), but that money goes to where the people are. If Linux starts creeping into the business and schools, game ports will follow (hopefully with better effort toward binary compatibility).
Ya?
NASA operates a serious marketing engine now (not that I can blame them, since they keep getting their money taken away if they don't) -- all those "beautiful pictures of cosmic objects" are usually a bunch of radiation grabbed from somewhere up in the X-ray range that are then rammed through a mapping program, enhanced and composited, and finally fed to Photoshop or whatever image manipulation program until they look really pretty. The glowing swaths of orange breaking into seas of purple fog and stuff like that, the sort of thing that sci-fi authors sometimes get romantic about, is little more than a tribute to the artistic ability of the folks at NASA. Real astronomy isn't all that exciting or pretty, even if it is scientifically significant and an impressive body of work.
It's hard to get people to donate money for worthy causes like the EFF without engaging in scare tactics. People are pretty lethargic.
Mass media is a wonderful example of marketing -- the newspaper is trying to make news exciting so that they can sell subscripts, hold close enough to the truth so that readers will continue to be able to treat the paper as unvarnished truth, and the newspaper's sources are providing information in such a manner as to promote their own agenda.
Research, IT, organizations in companies market themselves to the head honchos for budget. Each individual person is trying to sell themselves to move up on the ladder.
Everything involves marketing these days (well, I guess it always did, but mass media improved the payoff of spending effort on good marketing). It's kinda depressing, but it's kinda hard to motivate people without taking advantage of quirks of the psyche.
Could you provide a link? I can't seem to find evidence of deprecated drivers in the xfree86 4.4 hardware compatibility list.
I doubt that Microsoft will *die* if Linux gets big.
That doesn't mean that they will retain their current strength in desktop operating system dominance, though.
Really, though...I wonder what that would look like?
MS would probably do a port of Explorer and Office and friends over to X11. These would remain closed. They might implement hardware compatibility via porting their current libraries (which they'd presumably not open source). VB, everything moves.
You can still make a pretty proprietary platform on top of Linux. If Linux becomes a checkbox, it'll get filled, even by Microsoft.
Yup.
And you know what? After all the hype about how great expee is, and how great OS X is, I still find that linux suits me just fine - although I will admit that if there were no linux, I'd probably be using OS X now.
Pretty well sums up my own opinion as well.
It's not that OS X is bad. It has some nice things in it. I just don't see it replacing either Linux or Windows. Linux generally does the things that a hacker works on at least as well as OS X. OS X has better mass-market appeal due to Apple's famous ease-of-use (though I don't think that the user interface people at Apple today come close to measuring up to the brilliant people who worked on the classic Mac OS design two decades ago). That would seem to make OS X a Windows-killer. The problem is that Apple made a conscious and measured decision to sell to a limited market. They intend to only sell to people that buy Mac hardware, rather than to try to compete on price with the PC market. That eliminates them as a market-dominating competitor, even if it means that they can live a healthy life selling a premium product.
Plus, I kinda think that Apple would make for a lousy Establishment. I mean: "Think The Same: Use Apple"? No, just wouldn't work well at all.