The same reason OSX bogs down the G4 is why you should be running linux on the G4.
Plus there have been a barrage of OSX exploits, give it some time to mature. Just because the barrage was mostly ignored in light of the RAGING TORRENT of windows exploits, worms, and viruses, is no reason to forget it.
Best is pretty relative, I find linux gives a better combination of flexiblity and ease of use, OSX has no combination, it trades flexibility for ease of use.
OS X is the best unix desktop (atm) for joe sixpack, or a light user (tech writer, EE etc). For the more hardcore users linux is still the answer. The poster however, sounds like he'd fall into the OS X category.
"The GPL is vague enough with respect to 'a work based on the Program' that I wouldn't bet my business on it. Given Stallman's admission that the LGPL was designed to ensnare developers, so that a later conversion from LGPL to GPL would force them to GPL their own code, I wouldn't put anything past him. If he thought it would help his 'Free Software' cult, he'd be happy to sue Microsoft over GCC in Interix."
Cool, I'll start burning cd's with gcc on them and toss on whatever software I want open sourced next. You have some pretty odd ideas of what constitutes a derivative work. Being distributed on the same media as another program is REALLY stretching it. It was legally debatable whether there was even a need for the LGPL to begin with and that involves the code from a gpl'd library being physically used to provide functionality in your program!
" If and when a court rules that UNIX is no longer a registered trademark of The Open Group, you'll have a point. 'UNIX' is arguably a 'famous' mark as well, like 'Xerox', for example, which adds further protection under American law. Note that all of this applies to commercial use, not common use, e.g. Xerox could sue Canon if the latter called its copiers 'xerox machines', but obviously can't sue someone who uses 'xerox' to mean 'copy' (which apparently some people do)."
In the case of trademark dillution courts recognize WHAT HAS ALREADY HAPPENED. So the trademark having no value is a condition which predates the courts getting ahold of it. It happens to be the case now. MOST trademark owners recognize this as well and don't throw their money away going to court over it.
"Always an easy escape route for someone who's wrong, and afraid to admit it."
No it's recognizing your dealing with someone who is arguing simply for the sake of arguing and getting a rise out of people. Which is just what your doing.
"Whether the companies I named sell hardware is irrelevant. They also sell operating systems to run that hardware. And they all sell UNIX - which is the only reason they're not pushing Linux harder - they don't want to cut off their own proprietary UNIX variants (which are doomed anyway by Linux). So their competition is very relevant. Besides, it doesn't matter WHY they're pushing Linux - to Microsoft, it's still competition."
Perhaps you missed it, so I'll say it again. These companies offer server solutions, ibm and sun have desktop solutions but those solutions come with windows installed on them. They are competiting with MS yes, but MS doesn't have a monopoly in the area they are competing in. MS is a LONG way from having anything vaguely resembling a monopoly on the server.
"However, there is nothing stopping someone from designing a completely and totally reverse-engineered Access database, regardless of MS patents. The data format does not have to be identical - merely the functionality"
That WOULD violate MS patents.
"migration tool to allow extracting Access data to the new database (and there should be the same report generation and migration ease."
Again the PROCESS of reading an access database file is patented. With patents it doesn't matter if you have a clean room implementation.
Obviously you don't understand how severe the software patent issue really is and why software patents are so evil.
And don't forget what I said about yamaha dealers either. Not everything is a matter of coding an alternative. There is no such thing as an alternative for things like automotive, motorcycle, and farm equipment parts. These packages get handed down by the vendors to the dealers and the dealers run what their handed, period. Because what their handed has all the parts, prices etc in it
"If I had $50 billion"
Actually even $50 billion isn't enough to pay for the manpower that goes into open source, wouldn't even pay it for a week. That is why MS will lose in the end. Neither they nor any other corporation can afford to compete with open source.
The death of the write once sell millions of copies software houses is inevitable, it's just a matter of time (which employ a very tiny fraction of programmers no matter what they want you to think). And the way open source works, it can keep going for decades without winning the fight if need be, it's like the energizer bunny, just keeps going and going and going.
Alas my friend, I can't shake the feeling we are debating semantics anyway. If we end the thread disagreeing, neither of us leaves better off. If we end the thread agreeing either way, the result is the same, neither of us is ahead in any way.
apt is definately the way to go, generally I install from the dag repository on a fedora system, which loads you up with 3 or 4 very large compatible repositories.
Sun, HP, and IBM sell hardware for starters. And the hardware they sell running linux is in the server market where Microsoft has no monopoly. It's not really relevant in this discussion.
"The bottom line is that Microsoft has no state support for, regulatory or other legal basis for its "monopoly". Therefore it is not a true monopoly."
Well not a Constitutional monopoly anyway. As per the constitution a government granted 3rd party monopoly is illegal in and of itself without any abuse whatsoever.
"Any monopoly must be a coercive monopoly, unless it is a "natural" monopoly."
There has to be a punch behind it's push, this is not milk or some other commodity, this is software and it is designed for incompatiblity. Almost every example of linux breaking these barriers through reverse engineering predates issues regarding software patents as well. All of the new "technologies" Microsoft puts out are patent encumbered and CANNOT be reverse engineered for interaction. Without interaction and a slow transition it is NOT viable to replace windows. Again the best example off the top of my head is an access database, there are no shortage of companies that use them for everything and would crumble if they didn't have the data in them.
The software itself is replaceable, the data however is not and is locked in a propietary format which is patent encumbered so it cannot be legally reverse engineered. Some of the new things coming in terms of trusted computing and such WILL be further protected by law under the DMCA. Microsoft is using it's monopoly to bully vendors into selling windows only motherboards... and succeeding. Workarounds will be DMCA violations and therefore illegal.
"Microsoft has a TEMPORARY natural monopoly position because of the factors you cite and others. It is highly unlikely to last."
At least we are past quasi-monopoly;) Temporary is a given, everything is temporary and history has taught us that all empires crumble. The question is how long will it take on it's own, because each day they exist is a day in which technology is not progressing at anywhere near the rate it should be.
"Incidentally, with a few exceptions like the compiler, the open-source code in SFU is BSD-licensed, not the GPL-licensed."
RTFA, it specifically states that Microsoft is replacing the gpl'd code in SFU with SCO Licensed code. And uses the excuse that commercial code can't be shipped alongside gpl'd code as the reason it's not included with windows (which is ridiculous of course).
"In other words, you can't deny UNIX is a registered trademark of The Open Group, so you have to resort to childish insults. Not really surprising for a Linux zealot, I'm afraid."
lol so we really have reached a time when being suggested to be a lawyer is considered an insult.
Yes there is a registered trademark, yes it's meaningless. Yes there is a generic term unix that has been in use as a generic term to refer to any unix-like system for over a decade.
Yes by law that dillutes and invalidates the trademark whether a court has confirmed it yet or not.
If Unix were merely a trademark, then you'd have no point. If Unix is the generic term then you have no case.
So what exactly is it your trying to prove in all this? NM, I've been baited by a troll, it's time to accept it.
Although if you want to play dvd's, divx, or any other type of worthwhile media aside from mp3's you'll have to hunt down 3rd party sources.
The good thing is that, unlike fedora where 3rd party sources containing MASSIVE repositories of rpms are all over the place confusing you. With SUSE there are like two, and they have less packages than the cd's do not more. In fact, if you actually manage to find a 3rd party repository for yast that has something in it besides the base distribution, let me know.
Are you a lawyer? Perhaps a judge? Or some other form of blind idiot?
Walk into a linux conference (take your pick) or random lug meetings and start asking them if they work with linux. If 9 out of 10 answers is yes (and refering to linux) then the trademark is IN REALITY dilluted, whatever a court rules.
Last I checked the trademark was something pretty much anybody could buy a stake in, as opposed to fitting the generic term, which ACTUALLY REQUIRES a unix system.
BTW are you sure SFU has been certified and not just the code MS licensed to replace the gpl'd stuff in it? There is a big difference and the Open group will tell you as much.
Mozilla passes it with the appropriate security level, indicated it's unfiltered unchecked data coming from the web. They are doing exactly what they are supposed to.
And the whole reason the browser is passing it is because it's NOT a known uri type (who would expect there to be a shell uri, what kind of idiot comes up with the brilliant idea for a shell uri to begin with?).
This is windows, remember that most uri types aren't documented. Since we are only talking about unknown datatypes, it's a safe bet the browser will never know how to treat them. Which is why it does what it's supposed to do and passes it to the OS clearly labeled as hazardous nuclear waste.
Rarely useful?! At least half the media types you load are able to function due to this feature? (assuming you use Mozilla).
There is a security scheme in place in windows for this type of content for just this purpose. Mozilla handles this the way a windows application is supposed to do it. If the security scheme is broken it's 100% a microsoft issue.
The browser checks to see if it knows what to do with it, generally if it does, it blindly passes it to another application (plugin or whathaveyou), if not it blindly passes it to the OS which may or may not have a handler for that type.
First there shouldn't even be a shell uri in the OS! Second, there is a vulnerability IN THE SHELL URI which escalates the priv level to that of the user.
If Mozilla passed the data along and said, here ya go it's good stuff, completely trusted. That would be one thing, but mozilla passes it along and says I have no clue what this is or where it's coming from and have no reason to believe it safe in any fashion. You have any ideas?
If it's the RIGHT data, then windows tells itself it was the current user and not some untrusted guy off the web who gave it that data. The bug is in windows!
Hell the entire scheme or uri handling in windows is fscked up. There shouldn't be any uris which cause local execution!
"PS: Let us when Wine runs MSSQL with any sort of reasonable performance. Oh, sorry, that won't happen."
Why would anyone want to do that? I mean seriously, I can't think of an SQL database that doesn't womp MSSQL in almost every respect... I also can't think of any significant sql offerings that don't run natively on linux?
"Ask the Wine guys and they'll admit that Win32 on Unix is not optimal"
Ask the wine guys and they'll admit that win32 on nt or dos isn't optimal either. It's win32 that isn't optimal not the underlying system.
Your right to a degree, I've seen the windows code (good thing I don't work on wine eh?) and it's not properly modular, there are pieces which spill over where they should be totally abstract.
Basically it's spaghetti code. It's worth noting though, that about half the win32 api runs FASTER on wine than on the MS implementation. Probably due to better implementation as much as anything, but worth thinking about.
As for kernel modules, win32 just isn't fast, it depends on scheduler behavior like that in windows, robbing everything else of performance to speed up gui performance (which is one of the reasons everything but the gui is faster in *nix).
The win32 code still isn't as stable, this is evidenced with something we've all seen. Start an app that does alot of processing, move the window around rapidly.
In windows the window won't keep up with your cursor, in fact the app will usually freeze, sometimes even taking the gui with it (usually explorer reloads to try to prevent you from ever knowing it crashed but your system tray reveals the lie with missing icons).
In linux the window will keep up with your cursor but may leave a trail, even if the window doesn't keep up where you let go of the cursor is where it will be eventually. The display will catch up with it as soon as it can but neither the app or the gui will crash.
Hopefully with the new scheduler tools we've seen recently we'll no longer have to make the trade (overall anyway). You'll pick a scheduler at install time.
But here is something some don't know that will result in the linux gui blowing windows away. x.org is going OpenGL for xwindows. The gui and rendering work will be offloaded to the video card which is fully optimized for opengl (they all are thanks to the beloved quake benchmarks).
Although the method is different, the overall effect will be the same as the mac, the gui will be faster because the videocard is faster at graphics calculations. The cpu won't be bogged down by a pretty gui (unless you use a video card without hardware 3d acceleration but who does?).
Anyway, I've strayed a bit from wine performance (although this will help some aspects of it).
"Windows machine up to date is by no means a chore"
Downloading everything from windows update won't have you fully up to date. There are numerous outstanding critical vulnerabilities.
"My *free* (as in beer) virus checker updates and runs itself on a daily schedule"
Too bad there is no single virus checker which catches all the viruses. When doing a virus removal we typically run several, and usually find it was a good idea.
Because the viruses checkers generally aren't happy with one another, that means uninstalling and updating several scanners everytime you suspect a virus (about once a week for the typical windows user).
And last but certainly not least, pretty much every virus nowdays disabled the virus scanners. Since the viruses come out before the updates a fast spreading virus will hit the user and disable his virus scanner before he ever knows about it. If the virus exhibits no other obvious and visible effect the user, grown complacent by his virus scanner he ignores because it does everything on autopilot, generally won't know until the payload has already hit.
"Adaware (were I to pay for it) will similarly update and run itself, cleaning my machine of any malware that I inadvertently install (I don't, but I digress)."
99% of malware on your pc comes from vulnerabilities in IE. If your not getting much, it just means your going to websites that give you much. It's always odd to me when people seem to have this crazy idea it comes from software you install (that happens what, once or twice a year?). Also malware scanners are even worse than virusescanners, ad-ware doesn't even begin to catch all the malware on a windows pc. I've NEVER scanned a pc for malware using adware and not found more when I ran other scanners. There are about 5 we run on every pc through the shop, if we find nothing it means the pc wasn't on the web.
None of that is the real maintenance cost of windows though. All that is fine and dandy but there are several other critical problems. Bitrot, for one, over time applications installed will install thier own preferred versions of various system dlls along with their own dlls. When you install another application it might replace 3 of the 10 the last application changed. When you windows update it might replace a couple more. The next app replaces 1 of those. So on and so forth.
The net result is a complete hodgpodge of crap. Pretty soon numerous applications have problems because not all the libraries they use are 100% mesh with the behavior they need. This result in unusual application behavior and strange things happening, sometimes in rarely used functions and sometimes in commonly used ones. Usually this is resolved by reinstalling that application, only to be replaced by problems the next time around.
If the bitrot has gotten too bad the tech will get called. He will do one of two things, perfom an over the top installation (the method varies between windows version, usually an "upgrade" to the same version) replacing all the dlls to their shipped version or he'll just dump and reinstall.
Note your beloved windows update generally will still consider the updates to be installed, even though insecure versions may have been put back in place.
Of course this occurs with any windows system over time, it's most prevalent with those who install lots of applications (thus coupled with spyware, since their webbers, and viruses). Given enough time however, EVERY windows system suffers from bitrot before becoming obsolete. Even servers.
The next biggest reason for windows failure and bitrot is that windows assumes the hardware works, and that the other pieces of itself work!
"SFU is not shipped with Windows because SFU currently contains open-source software, such as the GNU C compiler, which cannot be distributed with commercial software. Zions confirmed that Microsoft is working to replace all open-source code in SFU with commercially licensed alternatives. Last year it licensed Unix software from SCO."
Since when can't open source software be shipped side by side with commercial software?
What Microsoft really means is that they don't want the fact that they use gpl'd software becoming very public.
"but oddly focuses on male homosexual relationships"
Probably because the largest audience for games at the moment is straight men and straight men are turned on by bi-sexual, lesbian, and straight women? Most straight men would not however be comfortable playing a game with a gay male theme.
Most women I've known on the other hand, just want there to be pics of naked studs as well as bimbos, and wouldn't be uncomfortable regardless?
Of course their answer might be different in a forum asking their opinion publically where they feel they "represent" the female view, as opposed to what they'll tell you one on one eating pizza and talking about life.
"Instead of having 10 people constantly re-inventing the wheel, you can have 1 person doing a job, and freeing up the other 9 people to do other innovative things."
If you have a monopoly, and colleges turning out 500 people qualified for a job which the monopoly has streamlined to needing 1 person to fill.
Maybe half of those people will find other similar jobs, the other half will be adding to the unemployment rates.
The one who does get the job, will be paid less (after all, they don't have to compete with other employers, your clamoring for the job instead of their clammoring for the employee), have less benefits, etc.
"Sure, with 10 peopld doing the exact same job, over and over, you might end up with one finding a truly innovative way of doing something, but you might also just end up with 10 more or less equal "wheels"."
Right, so worse case nothing gets better, best case something does. And if one does, there is a push for other 9 to come up with something better than that one did, it's called competition.
With one person doing it and no competition there is no push or incentive and nothing gets better period. If the one finds a better way or something innovative, the company isn't likely to be interested because they don't have to keep up with the joneses.
"Think, for example, about cell-phone service. In many places in the world, there is only 1 provider. This means the consumer doesn't have to worry about roaming to other networks, or whether provider X has better coverage in their area."
In this area there are 4. Because none of them can provide decent coverage on their own they solve the problem by sharing towers and because they have to compete with landlines, radio, etc they provide a large area with no roaming charges despite using eachother's towers.
When there is competition, these things get worked out. When it comes down to it the entire cell industry has to compete against other solutions to make Cell phones viable. Competition is good not bad.
"Even better, is that in many places, monopolies are regulated by the government and the government can tell the monopoly "Hey, you guys don't have to worry about competition, so in order to keep your sweet deal, you have to do X, Y, and Z". In the cell-phone example, that can mean forcing them to put up towers to provide coverage for more people, or keeping the rates under control, or many other things."
Government is slow, government regulation is the reason the phone company drags feet and most lines are outdated. Government regulation is the biggest reason for the last mile problem. A monopoly which is government regulated pretty much does nothing unless the government forces them to do it. In fact, if they try to do something to benefit the consumer, it's twice as expensive and takes several times longer because of the government red tape.
Look how long it was after the fact before the anti-trust case against microsoft even happened, look how long the trial took, then look how long it took for it to end. It's not just the legal system that is this way, it's everything done by government. There are checks, upon checks, upon reviews and at the end of it all the finish product is STILL screwed up because none of the people doing the checks and reviews actually care.
"Sometimes monopolies reduce prices, but often they don't. When there is fierce competition, companies spend a lot of money on advertising. Who ends up paying for the advertising? The consumers. Under a monopoly, there's no need for advertising, so those costs aren't passed on. A bigger problem is dealing with all the paperwork. Hospitals in the US have to have enormous paper-pushing departments to deal with the hundreds of insurance companies, health-care plans, and other annoyances resulting from competition. In countries with government-provided healthcare, this process is competely streamlined, and the costs of providing health care are greatly reduced."
Actually KDE takes more like 8-10hrs and as much again if want gnome also (which most do) on an Athlon XP 2600 with 512mb DDR.
I was exaggerating a tad. But your average computer in use today is MUCH slower than my athlon, in fact it's more like p3'ish to t-bird.
Don't get me wrong, I love gentoo. Although there are definately holes in the documentation, what is there is an example how documentation should be done. And portage is a dream in terms of doing what it does and doing it well. A gentoo system is certainly a hell of alot easier than a LFS system.
3hrs for open office compared with 30min (assuming it has to be downloaded as well, probably less) for the rpm, it's a hell of a difference.
The worst thing though, is after spending all the time to compile my gentoo system completely from scratch, along with all the applications I wanted on it. And then to configure everything (most of the confs are in the raw form they are when you download the tarball, which is almost never usable out of the box).
Anyway, after spending the time to get everything installed the way I like it, a full weekend of nonstop compiling and then tweaking and app installing over the next week. The system lasted a couple months, rock solid and very fast (no matter what anyone says about minimal gains compiling from source). My motherboard then died, when I ordered a replacement board they sent me the wrong one... instead they sent me what was the top of the line p4 board at the time.
By this point I was starving for computer and decided to give a p4 a shot (not even thinking about my fully athlon optimized OS). Naturally it wouldn't even run the bootloader.
Somewhere in all that I figured out, I uninstall and install applications daily. I change hardware very very frequently. As much as I loved the performance and pride that came with a gentoo system, I just didn't have time for one.
"Corporate users have lock-in because they don't want to re-train."
Corporate users are a tiny pretty much insignificant factor in the overall scheme of things. They represent the biggest sales in a single shot but overall I'd be suprised if they amounted to double digits of the market.
A far far larger chunk is small to mid-sized businesses. Yup the 99% of businesses that move less than a million annually. Also schools.
Schools have to run a platform which supports the software they have classes for (for instance I haven't heard of a highschool without an autocad class).
Small business is more the concern here. For instance, no matter what dell preinstalls on it's computers, any shop that sells or services yamaha motorcycles will be buying a windows pc. If they currently use dell, next year they might be buying local or using an HP.
As for retraining, the numbers are bigger for a corporation but the impact is much bigger for a small shop. Retraining in corporate america can be done a piece or a few teams at a time, at no point griding them to a halt.
In a small business they can't afford a training lab, and they can't afford to have all their people off the floor. Everything has to be switched at once and everybody needs to be able to use it at once. Not having this condition met doesn't just mean a loss of profit, it means they go bankrupt.
They don't have an IT staff, they hire the local tech shop. They might have one person who knows enough to be able to install acrobat reader if someone walks them through the download over the phone. On a bad day it could kill this months profit if that person is out for retraining.
The people who own and run these shops are the meat of america. They are the ones who actually pay for software, the average home user certainly doesn't. Because the average home user is pulling less than $25k/yr and can barely afford to eat. Joe sixpack doesn't buy MS Office or windows unless he happens to be the joe sixpack who owns joe's pizza corner.
Joe's pizza corner does it's order with a proprietary application which only runs on windows. And Joe's pizza corner can't afford to buy a computer just to find out if that application will work in wine. And guess what, installing that application on another computer to test that is called software piracy and illegal.
You'd say this would be a problem with the software vendor not Microsoft. I'd say this is a problem with the win32 api not being fully and properly documented.
Someone else said something about "if they can upgrade to xp...". Most of them HAVEN'T upgraded to XP. Most are still running win98.
"Dell and HP (and of course IBM) can accelerate that process by providing Linux at low-cost in new machines with more support."
More than anything though, this is the point. You view MS as a quasi-monopoly because there is competition. But there isn't, linux doesn't qualify, it is NOT a competitor who is successfully competing against windows in the marketplace.
Lets give an example. Does anyone question AT&T was a monopoly? I don't think so, no COMPANY could successfully compete in the marketplace against them.
But lets say that jed designs his own phone equipment, jed gets together a group of people who have all the resources and knowledge needed, iron mines, the whole works to build everything from scratch needed to run the phone systems. Jed and buddies all get together and produce the equiptment and run the service all on their own, everyone on the service just has to pay the expenses to maintain the equipment at their own house, no monthly fees, nothing.
Could Jed and his buddy compete with AT&T in it's prime, slowly chipping away at it's marketshare? Certainly they could. Does it make AT&T any less of a monopoly? Of course not, Jed is not AT&T's peer, jed doesn't have to turn a profit. The companies that work with linux to make their profit aren't competing with Microsoft either. Either
"Why do people think that Install Wizards are so easy?"
Because they are time saving. The mac handles installs very poorly for instance. One click is a beautiful dream but not worthwhile in practice. Even intermediate users want to change options alot of the time.
What is poor is your installation class choices. With linux we don't have to care about sounding professional like commercial offerings.
A few sample possibilities:
"I Intimidated, figure all this out for me."
Picking this and clicking next will install using all defaults without asking another question unless it cannot continue (lack of diskspace or some such).
"Beginner"
Basically the same sort of simple questions asked by a windows installer now. Where to install it, do you want a shortcut on the desktop.
"Intermediate"
Configuration choices, both general and specific, anything that doesn't take knowledge of programming to understand.
"Advanced"
All available options, as well as general config schemes (advanced doesn't mean i should have to set EVERY option, just that I want them at my fingertips).
"But some people would prefer to take the easy way out, and require the user to make all the decisions."
The problem with this logic is that linux packages require all the configuration to be done after the fact. Sometimes hours of pointless reading through options in conf files to set one or two of them.
And because there are so many, you might do this several times before remembering what the option you needed to set is without going through most of the conf.
With an install wizard, postfix for instance could ship ready to go out of the box with no additional configuration by asking just 1 question, local subnet. Now you have to find and edit the conf file, if your not particulary familiar with postfix that means reading about 100 sometimes ambiguous options to determine if you need to set them. For most people the end result is that they needed to set ONE thing out of all those options.
Just about every package should be fully functional and configured upon installation, not just chucked on the drive.
And almost all sourcecode comes with those instructions, most don't give much documentation beyond that. And 98% fail EVERY time if you do exactly that and don't use any additional options or provide additional dependencies.
Unfortunately mandrake has chosen to rename their core system packages and libraries in such a fashion that a redhat rpm won't recognize them as dependencies and vice versa.
Do they gain anything by this? NO Did it involve lots of work with absolutely no gain whatsoever to go from standard redhat naming to the Mandrake scheme? Yes.
why? So they could claim they were their own distro and not just a redhat knockoff with their own logos and some graphical configuration utils.
Pretty much the same reason they use their own installer which is inferior to redhat's. And their own hardware detection, which again, is inferior to redhats.
In what world is that? There is apt for rpm my friend and some HUGE repositories out there.
Yes apt is good and wonderful, deb is equal, so why not drop deb like the rest of the distros out there and keep apt?
Right now debian is not standards compliant, like the parent said rpm is specified in the LSB. This fight was fought and settled long ago... debian just won't listen.
Slashdot Mirror
The same reason OSX bogs down the G4 is why you should be running linux on the G4.
Plus there have been a barrage of OSX exploits, give it some time to mature. Just because the barrage was mostly ignored in light of the RAGING TORRENT of windows exploits, worms, and viruses, is no reason to forget it.
Best is pretty relative, I find linux gives a better combination of flexiblity and ease of use, OSX has no combination, it trades flexibility for ease of use.
OS X is the best unix desktop (atm) for joe sixpack, or a light user (tech writer, EE etc). For the more hardcore users linux is still the answer. The poster however, sounds like he'd fall into the OS X category.
"The GPL is vague enough with respect to 'a work based on the Program' that I wouldn't bet my business on it. Given Stallman's admission that the LGPL was designed to ensnare developers, so that a later conversion from LGPL to GPL would force them to GPL their own code, I wouldn't put anything past him. If he thought it would help his 'Free Software' cult, he'd be happy to sue Microsoft over GCC in Interix."
Cool, I'll start burning cd's with gcc on them and toss on whatever software I want open sourced next. You have some pretty odd ideas of what constitutes a derivative work. Being distributed on the same media as another program is REALLY stretching it. It was legally debatable whether there was even a need for the LGPL to begin with and that involves the code from a gpl'd library being physically used to provide functionality in your program!
"
If and when a court rules that UNIX is no longer a registered trademark of The Open Group, you'll have a point. 'UNIX' is arguably a 'famous' mark as well, like 'Xerox', for example, which adds further protection under American law. Note that all of this applies to commercial use, not common use, e.g. Xerox could sue Canon if the latter called its copiers 'xerox machines', but obviously can't sue someone who uses 'xerox' to mean 'copy' (which apparently some people do)."
In the case of trademark dillution courts recognize WHAT HAS ALREADY HAPPENED. So the trademark having no value is a condition which predates the courts getting ahold of it. It happens to be the case now. MOST trademark owners recognize this as well and don't throw their money away going to court over it.
"Always an easy escape route for someone who's wrong, and afraid to admit it."
No it's recognizing your dealing with someone who is arguing simply for the sake of arguing and getting a rise out of people. Which is just what your doing.
"Whether the companies I named sell hardware is irrelevant. They also sell operating systems to run that hardware. And they all sell UNIX - which is the only reason they're not pushing Linux harder - they don't want to cut off their own proprietary UNIX variants (which are doomed anyway by Linux). So their competition is very relevant. Besides, it doesn't matter WHY they're pushing Linux - to Microsoft, it's still competition."
Perhaps you missed it, so I'll say it again. These companies offer server solutions, ibm and sun have desktop solutions but those solutions come with windows installed on them. They are competiting with MS yes, but MS doesn't have a monopoly in the area they are competing in. MS is a LONG way from having anything vaguely resembling a monopoly on the server.
"However, there is nothing stopping someone from designing a completely and totally reverse-engineered Access database, regardless of MS patents. The data format does not have to be identical - merely the functionality"
That WOULD violate MS patents.
"migration tool to allow extracting Access data to the new database (and there should be the same report generation and migration ease."
Again the PROCESS of reading an access database file is patented. With patents it doesn't matter if you have a clean room implementation.
Obviously you don't understand how severe the software patent issue really is and why software patents are so evil.
And don't forget what I said about yamaha dealers either. Not everything is a matter of coding an alternative. There is no such thing as an alternative for things like automotive, motorcycle, and farm equipment parts. These packages get handed down by the vendors to the dealers and the dealers run what their handed, period. Because what their handed has all the parts, prices etc in it
"If I had $50 billion"
Actually even $50 billion isn't enough to pay for the manpower that goes into open source, wouldn't even pay it for a week. That is why MS will lose in the end. Neither they nor any other corporation can afford to compete with open source.
The death of the write once sell millions of copies software houses is inevitable, it's just a matter of time (which employ a very tiny fraction of programmers no matter what they want you to think). And the way open source works, it can keep going for decades without winning the fight if need be, it's like the energizer bunny, just keeps going and going and going.
Alas my friend, I can't shake the feeling we are debating semantics anyway. If we end the thread disagreeing, neither of us leaves better off. If we end the thread agreeing either way, the result is the same, neither of us is ahead in any way.
apt is definately the way to go, generally I install from the dag repository on a fedora system, which loads you up with 3 or 4 very large compatible repositories.
Sun, HP, and IBM sell hardware for starters. And the hardware they sell running linux is in the server market where Microsoft has no monopoly. It's not really relevant in this discussion.
;) Temporary is a given, everything is temporary and history has taught us that all empires crumble. The question is how long will it take on it's own, because each day they exist is a day in which technology is not progressing at anywhere near the rate it should be.
"The bottom line is that Microsoft has no state support for, regulatory or other legal basis for its "monopoly". Therefore it is not a true monopoly."
Well not a Constitutional monopoly anyway. As per the constitution a government granted 3rd party monopoly is illegal in and of itself without any abuse whatsoever.
"Any monopoly must be a coercive monopoly, unless it is a "natural" monopoly."
There has to be a punch behind it's push, this is not milk or some other commodity, this is software and it is designed for incompatiblity. Almost every example of linux breaking these barriers through reverse engineering predates issues regarding software patents as well. All of the new "technologies" Microsoft puts out are patent encumbered and CANNOT be reverse engineered for interaction. Without interaction and a slow transition it is NOT viable to replace windows. Again the best example off the top of my head is an access database, there are no shortage of companies that use them for everything and would crumble if they didn't have the data in them.
The software itself is replaceable, the data however is not and is locked in a propietary format which is patent encumbered so it cannot be legally reverse engineered. Some of the new things coming in terms of trusted computing and such WILL be further protected by law under the DMCA. Microsoft is using it's monopoly to bully vendors into selling windows only motherboards... and succeeding. Workarounds will be DMCA violations and therefore illegal.
"Microsoft has a TEMPORARY natural monopoly position because of the factors you cite and others. It is highly unlikely to last."
At least we are past quasi-monopoly
"Incidentally, with a few exceptions like the compiler, the open-source code in SFU is BSD-licensed, not the GPL-licensed."
RTFA, it specifically states that Microsoft is replacing the gpl'd code in SFU with SCO Licensed code. And uses the excuse that commercial code can't be shipped alongside gpl'd code as the reason it's not included with windows (which is ridiculous of course).
"In other words, you can't deny UNIX is a registered trademark of The Open Group, so you have to resort to childish insults. Not really surprising for a Linux zealot, I'm afraid."
lol so we really have reached a time when being suggested to be a lawyer is considered an insult.
Yes there is a registered trademark, yes it's meaningless. Yes there is a generic term unix that has been in use as a generic term to refer to any unix-like system for over a decade.
Yes by law that dillutes and invalidates the trademark whether a court has confirmed it yet or not.
If Unix were merely a trademark, then you'd have no point. If Unix is the generic term then you have no case.
So what exactly is it your trying to prove in all this? NM, I've been baited by a troll, it's time to accept it.
Although if you want to play dvd's, divx, or any other type of worthwhile media aside from mp3's you'll have to hunt down 3rd party sources.
The good thing is that, unlike fedora where 3rd party sources containing MASSIVE repositories of rpms are all over the place confusing you. With SUSE there are like two, and they have less packages than the cd's do not more. In fact, if you actually manage to find a 3rd party repository for yast that has something in it besides the base distribution, let me know.
Are you a lawyer? Perhaps a judge? Or some other form of blind idiot?
Walk into a linux conference (take your pick) or random lug meetings and start asking them if they work with linux. If 9 out of 10 answers is yes (and refering to linux) then the trademark is IN REALITY dilluted, whatever a court rules.
Last I checked the trademark was something pretty much anybody could buy a stake in, as opposed to fitting the generic term, which ACTUALLY REQUIRES a unix system.
BTW are you sure SFU has been certified and not just the code MS licensed to replace the gpl'd stuff in it? There is a big difference and the Open group will tell you as much.
Mozilla passes it with the appropriate security level, indicated it's unfiltered unchecked data coming from the web. They are doing exactly what they are supposed to.
And the whole reason the browser is passing it is because it's NOT a known uri type (who would expect there to be a shell uri, what kind of idiot comes up with the brilliant idea for a shell uri to begin with?).
This is windows, remember that most uri types aren't documented. Since we are only talking about unknown datatypes, it's a safe bet the browser will never know how to treat them. Which is why it does what it's supposed to do and passes it to the OS clearly labeled as hazardous nuclear waste.
Rarely useful?! At least half the media types you load are able to function due to this feature? (assuming you use Mozilla).
There is a security scheme in place in windows for this type of content for just this purpose. Mozilla handles this the way a windows application is supposed to do it. If the security scheme is broken it's 100% a microsoft issue.
The browser checks to see if it knows what to do with it, generally if it does, it blindly passes it to another application (plugin or whathaveyou), if not it blindly passes it to the OS which may or may not have a handler for that type.
First there shouldn't even be a shell uri in the OS! Second, there is a vulnerability IN THE SHELL URI which escalates the priv level to that of the user.
If Mozilla passed the data along and said, here ya go it's good stuff, completely trusted. That would be one thing, but mozilla passes it along and says I have no clue what this is or where it's coming from and have no reason to believe it safe in any fashion. You have any ideas?
If it's the RIGHT data, then windows tells itself it was the current user and not some untrusted guy off the web who gave it that data. The bug is in windows!
Hell the entire scheme or uri handling in windows is fscked up. There shouldn't be any uris which cause local execution!
Unix has become a generic term, the trademark is diluted.
A unix system represents a philosophy and level of posix compliance.
Linux IS Unix, and Darwin IS Unix. And yes, debian downloads should certainly count.
And here I was under the impression they contained clients not servers, just like the misnamed netware package?
"PS: Let us when Wine runs MSSQL with any sort of reasonable performance. Oh, sorry, that won't happen."
Why would anyone want to do that? I mean seriously, I can't think of an SQL database that doesn't womp MSSQL in almost every respect... I also can't think of any significant sql offerings that don't run natively on linux?
I've been baited by a troll haven't I?
"Ask the Wine guys and they'll admit that Win32 on Unix is not optimal"
Ask the wine guys and they'll admit that win32 on nt or dos isn't optimal either. It's win32 that isn't optimal not the underlying system.
Your right to a degree, I've seen the windows code (good thing I don't work on wine eh?) and it's not properly modular, there are pieces which spill over where they should be totally abstract.
Basically it's spaghetti code. It's worth noting though, that about half the win32 api runs FASTER on wine than on the MS implementation. Probably due to better implementation as much as anything, but worth thinking about.
As for kernel modules, win32 just isn't fast, it depends on scheduler behavior like that in windows, robbing everything else of performance to speed up gui performance (which is one of the reasons everything but the gui is faster in *nix).
The win32 code still isn't as stable, this is evidenced with something we've all seen. Start an app that does alot of processing, move the window around rapidly.
In windows the window won't keep up with your cursor, in fact the app will usually freeze, sometimes even taking the gui with it (usually explorer reloads to try to prevent you from ever knowing it crashed but your system tray reveals the lie with missing icons).
In linux the window will keep up with your cursor but may leave a trail, even if the window doesn't keep up where you let go of the cursor is where it will be eventually. The display will catch up with it as soon as it can but neither the app or the gui will crash.
Hopefully with the new scheduler tools we've seen recently we'll no longer have to make the trade (overall anyway). You'll pick a scheduler at install time.
But here is something some don't know that will result in the linux gui blowing windows away. x.org is going OpenGL for xwindows. The gui and rendering work will be offloaded to the video card which is fully optimized for opengl (they all are thanks to the beloved quake benchmarks).
Although the method is different, the overall effect will be the same as the mac, the gui will be faster because the videocard is faster at graphics calculations. The cpu won't be bogged down by a pretty gui (unless you use a video card without hardware 3d acceleration but who does?).
Anyway, I've strayed a bit from wine performance (although this will help some aspects of it).
"Windows machine up to date is by no means a chore"
Downloading everything from windows update won't have you fully up to date. There are numerous outstanding critical vulnerabilities.
"My *free* (as in beer) virus checker updates and runs itself on a daily schedule"
Too bad there is no single virus checker which catches all the viruses. When doing a virus removal we typically run several, and usually find it was a good idea.
Because the viruses checkers generally aren't happy with one another, that means uninstalling and updating several scanners everytime you suspect a virus (about once a week for the typical windows user).
And last but certainly not least, pretty much every virus nowdays disabled the virus scanners. Since the viruses come out before the updates a fast spreading virus will hit the user and disable his virus scanner before he ever knows about it. If the virus exhibits no other obvious and visible effect the user, grown complacent by his virus scanner he ignores because it does everything on autopilot, generally won't know until the payload has already hit.
"Adaware (were I to pay for it) will similarly update and run itself, cleaning my machine of any malware that I inadvertently install (I don't, but I digress)."
99% of malware on your pc comes from vulnerabilities in IE. If your not getting much, it just means your going to websites that give you much. It's always odd to me when people seem to have this crazy idea it comes from software you install (that happens what, once or twice a year?). Also malware scanners are even worse than virusescanners, ad-ware doesn't even begin to catch all the malware on a windows pc. I've NEVER scanned a pc for malware using adware and not found more when I ran other scanners. There are about 5 we run on every pc through the shop, if we find nothing it means the pc wasn't on the web.
None of that is the real maintenance cost of windows though. All that is fine and dandy but there are several other critical problems. Bitrot, for one, over time applications installed will install thier own preferred versions of various system dlls along with their own dlls. When you install another application it might replace 3 of the 10 the last application changed. When you windows update it might replace a couple more. The next app replaces 1 of those. So on and so forth.
The net result is a complete hodgpodge of crap. Pretty soon numerous applications have problems because not all the libraries they use are 100% mesh with the behavior they need. This result in unusual application behavior and strange things happening, sometimes in rarely used functions and sometimes in commonly used ones. Usually this is resolved by reinstalling that application, only to be replaced by problems the next time around.
If the bitrot has gotten too bad the tech will get called. He will do one of two things, perfom an over the top installation (the method varies between windows version, usually an "upgrade" to the same version) replacing all the dlls to their shipped version or he'll just dump and reinstall.
Note your beloved windows update generally will still consider the updates to be installed, even though insecure versions may have been put back in place.
Of course this occurs with any windows system over time, it's most prevalent with those who install lots of applications (thus coupled with spyware, since their webbers, and viruses). Given enough time however, EVERY windows system suffers from bitrot before becoming obsolete. Even servers.
The next biggest reason for windows failure and bitrot is that windows assumes the hardware works, and that the other pieces of itself work!
"SFU is not shipped with Windows because SFU currently contains open-source software, such as the GNU C compiler, which cannot be distributed with commercial software. Zions confirmed that Microsoft is working to replace all open-source code in SFU with commercially licensed alternatives. Last year it licensed Unix software from SCO."
Since when can't open source software be shipped side by side with commercial software?
What Microsoft really means is that they don't want the fact that they use gpl'd software becoming very public.
"but oddly focuses on male homosexual relationships"
Probably because the largest audience for games at the moment is straight men and straight men are turned on by bi-sexual, lesbian, and straight women? Most straight men would not however be comfortable playing a game with a gay male theme.
Most women I've known on the other hand, just want there to be pics of naked studs as well as bimbos, and wouldn't be uncomfortable regardless?
Of course their answer might be different in a forum asking their opinion publically where they feel they "represent" the female view, as opposed to what they'll tell you one on one eating pizza and talking about life.
"Instead of having 10 people constantly re-inventing the wheel, you can have 1 person doing a job, and freeing up the other 9 people to do other innovative things."
If you have a monopoly, and colleges turning out 500 people qualified for a job which the monopoly has streamlined to needing 1 person to fill.
Maybe half of those people will find other similar jobs, the other half will be adding to the unemployment rates.
The one who does get the job, will be paid less (after all, they don't have to compete with other employers, your clamoring for the job instead of their clammoring for the employee), have less benefits, etc.
"Sure, with 10 peopld doing the exact same job, over and over, you might end up with one finding a truly innovative way of doing something, but you might also just end up with 10 more or less equal "wheels"."
Right, so worse case nothing gets better, best case something does. And if one does, there is a push for other 9 to come up with something better than that one did, it's called competition.
With one person doing it and no competition there is no push or incentive and nothing gets better period. If the one finds a better way or something innovative, the company isn't likely to be interested because they don't have to keep up with the joneses.
"Think, for example, about cell-phone service. In many places in the world, there is only 1 provider. This means the consumer doesn't have to worry about roaming to other networks, or whether provider X has better coverage in their area."
In this area there are 4. Because none of them can provide decent coverage on their own they solve the problem by sharing towers and because they have to compete with landlines, radio, etc they provide a large area with no roaming charges despite using eachother's towers.
When there is competition, these things get worked out. When it comes down to it the entire cell industry has to compete against other solutions to make Cell phones viable. Competition is good not bad.
"Even better, is that in many places, monopolies are regulated by the government and the government can tell the monopoly "Hey, you guys don't have to worry about competition, so in order to keep your sweet deal, you have to do X, Y, and Z". In the cell-phone example, that can mean forcing them to put up towers to provide coverage for more people, or keeping the rates under control, or many other things."
Government is slow, government regulation is the reason the phone company drags feet and most lines are outdated. Government regulation is the biggest reason for the last mile problem. A monopoly which is government regulated pretty much does nothing unless the government forces them to do it. In fact, if they try to do something to benefit the consumer, it's twice as expensive and takes several times longer because of the government red tape.
Look how long it was after the fact before the anti-trust case against microsoft even happened, look how long the trial took, then look how long it took for it to end. It's not just the legal system that is this way, it's everything done by government. There are checks, upon checks, upon reviews and at the end of it all the finish product is STILL screwed up because none of the people doing the checks and reviews actually care.
"Sometimes monopolies reduce prices, but often they don't. When there is fierce competition, companies spend a lot of money on advertising. Who ends up paying for the advertising? The consumers. Under a monopoly, there's no need for advertising, so those costs aren't passed on. A bigger problem is dealing with all the paperwork. Hospitals in the US have to have enormous paper-pushing departments to deal with the hundreds of insurance companies, health-care plans, and other annoyances resulting from competition. In countries with government-provided healthcare, this process is competely streamlined, and the costs of providing health care are greatly reduced."
The countries I kno
Actually KDE takes more like 8-10hrs and as much again if want gnome also (which most do) on an Athlon XP 2600 with 512mb DDR.
I was exaggerating a tad. But your average computer in use today is MUCH slower than my athlon, in fact it's more like p3'ish to t-bird.
Don't get me wrong, I love gentoo. Although there are definately holes in the documentation, what is there is an example how documentation should be done. And portage is a dream in terms of doing what it does and doing it well. A gentoo system is certainly a hell of alot easier than a LFS system.
3hrs for open office compared with 30min (assuming it has to be downloaded as well, probably less) for the rpm, it's a hell of a difference.
The worst thing though, is after spending all the time to compile my gentoo system completely from scratch, along with all the applications I wanted on it. And then to configure everything (most of the confs are in the raw form they are when you download the tarball, which is almost never usable out of the box).
Anyway, after spending the time to get everything installed the way I like it, a full weekend of nonstop compiling and then tweaking and app installing over the next week. The system lasted a couple months, rock solid and very fast (no matter what anyone says about minimal gains compiling from source). My motherboard then died, when I ordered a replacement board they sent me the wrong one... instead they sent me what was the top of the line p4 board at the time.
By this point I was starving for computer and decided to give a p4 a shot (not even thinking about my fully athlon optimized OS). Naturally it wouldn't even run the bootloader.
Somewhere in all that I figured out, I uninstall and install applications daily. I change hardware very very frequently. As much as I loved the performance and pride that came with a gentoo system, I just didn't have time for one.
"Corporate users have lock-in because they don't want to re-train."
Corporate users are a tiny pretty much insignificant factor in the overall scheme of things. They represent the biggest sales in a single shot but overall I'd be suprised if they amounted to double digits of the market.
A far far larger chunk is small to mid-sized businesses. Yup the 99% of businesses that move less than a million annually. Also schools.
Schools have to run a platform which supports the software they have classes for (for instance I haven't heard of a highschool without an autocad class).
Small business is more the concern here. For instance, no matter what dell preinstalls on it's computers, any shop that sells or services yamaha motorcycles will be buying a windows pc. If they currently use dell, next year they might be buying local or using an HP.
As for retraining, the numbers are bigger for a corporation but the impact is much bigger for a small shop. Retraining in corporate america can be done a piece or a few teams at a time, at no point griding them to a halt.
In a small business they can't afford a training lab, and they can't afford to have all their people off the floor. Everything has to be switched at once and everybody needs to be able to use it at once. Not having this condition met doesn't just mean a loss of profit, it means they go bankrupt.
They don't have an IT staff, they hire the local tech shop. They might have one person who knows enough to be able to install acrobat reader if someone walks them through the download over the phone. On a bad day it could kill this months profit if that person is out for retraining.
The people who own and run these shops are the meat of america. They are the ones who actually pay for software, the average home user certainly doesn't. Because the average home user is pulling less than $25k/yr and can barely afford to eat. Joe sixpack doesn't buy MS Office or windows unless he happens to be the joe sixpack who owns joe's pizza corner.
Joe's pizza corner does it's order with a proprietary application which only runs on windows. And Joe's pizza corner can't afford to buy a computer just to find out if that application will work in wine. And guess what, installing that application on another computer to test that is called software piracy and illegal.
You'd say this would be a problem with the software vendor not Microsoft. I'd say this is a problem with the win32 api not being fully and properly documented.
Someone else said something about "if they can upgrade to xp...". Most of them HAVEN'T upgraded to XP. Most are still running win98.
"Dell and HP (and of course IBM) can accelerate that process by providing Linux at low-cost in new machines with more support."
More than anything though, this is the point. You view MS as a quasi-monopoly because there is competition. But there isn't, linux doesn't qualify, it is NOT a competitor who is successfully competing against windows in the marketplace.
Lets give an example. Does anyone question AT&T was a monopoly? I don't think so, no COMPANY could successfully compete in the marketplace against them.
But lets say that jed designs his own phone equipment, jed gets together a group of people who have all the resources and knowledge needed, iron mines, the whole works to build everything from scratch needed to run the phone systems. Jed and buddies all get together and produce the equiptment and run the service all on their own, everyone on the service just has to pay the expenses to maintain the equipment at their own house, no monthly fees, nothing.
Could Jed and his buddy compete with AT&T in it's prime, slowly chipping away at it's marketshare? Certainly they could. Does it make AT&T any less of a monopoly? Of course not, Jed is not AT&T's peer, jed doesn't have to turn a profit. The companies that work with linux to make their profit aren't competing with Microsoft either. Either
"Why do people think that Install Wizards are so easy?"
Because they are time saving. The mac handles installs very poorly for instance. One click is a beautiful dream but not worthwhile in practice. Even intermediate users want to change options alot of the time.
What is poor is your installation class choices. With linux we don't have to care about sounding professional like commercial offerings.
A few sample possibilities:
"I Intimidated, figure all this out for me."
Picking this and clicking next will install using all defaults without asking another question unless it cannot continue (lack of diskspace or some such).
"Beginner"
Basically the same sort of simple questions asked by a windows installer now. Where to install it, do you want a shortcut on the desktop.
"Intermediate"
Configuration choices, both general and specific, anything that doesn't take knowledge of programming to understand.
"Advanced"
All available options, as well as general config schemes (advanced doesn't mean i should have to set EVERY option, just that I want them at my fingertips).
"But some people would prefer to take the easy way out, and require the user to make all the decisions."
The problem with this logic is that linux packages require all the configuration to be done after the fact. Sometimes hours of pointless reading through options in conf files to set one or two of them.
And because there are so many, you might do this several times before remembering what the option you needed to set is without going through most of the conf.
With an install wizard, postfix for instance could ship ready to go out of the box with no additional configuration by asking just 1 question, local subnet. Now you have to find and edit the conf file, if your not particulary familiar with postfix that means reading about 100 sometimes ambiguous options to determine if you need to set them. For most people the end result is that they needed to set ONE thing out of all those options.
Just about every package should be fully functional and configured upon installation, not just chucked on the drive.
And almost all sourcecode comes with those instructions, most don't give much documentation beyond that. And 98% fail EVERY time if you do exactly that and don't use any additional options or provide additional dependencies.
Yeah, and if only compiling everything didn't take up to 24hrs for just one package and it's dependencies!
Unfortunately mandrake has chosen to rename their core system packages and libraries in such a fashion that a redhat rpm won't recognize them as dependencies and vice versa.
Do they gain anything by this? NO Did it involve lots of work with absolutely no gain whatsoever to go from standard redhat naming to the Mandrake scheme? Yes.
why? So they could claim they were their own distro and not just a redhat knockoff with their own logos and some graphical configuration utils.
Pretty much the same reason they use their own installer which is inferior to redhat's. And their own hardware detection, which again, is inferior to redhats.
"RPM has long lacked the convenience of apt"
In what world is that? There is apt for rpm my friend and some HUGE repositories out there.
Yes apt is good and wonderful, deb is equal, so why not drop deb like the rest of the distros out there and keep apt?
Right now debian is not standards compliant, like the parent said rpm is specified in the LSB. This fight was fought and settled long ago... debian just won't listen.