Slashdot Mirror
What's Microsoft Up To?
So, today's one of those days when every bit of news is dominated by Microsoft. To spare you six different stories about the Borg, we'll assimilate them all into this one. You have seen the stupid Passport hole in an earlier story; also the iLoo, although that hasn't stopped you from submitting stories about it, oh no. New news: a report paid for by Microsoft shows that Windows is a better server than Red Hat. A class-action suit has been filed charging that MSN and Best Buy combined to scam customers. The WINHEC conference is ongoing - Steve Ballmer says DRM is an opportunity, not a prison, the Xbox is going to be your home communications center, Wired talks about how hardware will be changed to imprison users, and once you're locked in to Microsoft you get to pay more each year. An article describes why user desktops are locked down. Oh, and here's another on DRM, just because.
Is Slashdot running out of news?
I dont think "there has been lots of news about MS in Slashdot" counts as news.
6 months ago everything you heard from MS was about tablet PC's and how they were going revolutionize everyone's computing experience. Maybe MS figured out that not everyone wants to use a stylus with their computer or has a need to work standing up? I wouldn't mind having stylus functionality on my ultraportable laptop screen but the idea that tablets are going to create a whole new mainstream sector of computing is far fetched.
The writing stylings of michael in this story are truely childish. This is the most juvenile thing written.
If you want to attack Microsoft, this is the worst possible way to do it. You give Linux users a bad name with your elementary school-kid attitude and childish commentary.
I've seen bad stuff 'added' to submitters text, but this has to be the WORST I've ever seen on slashdot EVER.
Of course, I will be modded to -1 by a childish moderator with unlimited points, but I hope I'm not the only one that feels this way.
Good quote, too many chars. Seriously, the slashdot 120 char limit sucks!
Hmm... let me see if I can make this argument valid...
:P
Cost:
Windows: Expensive (especially if you count licenses)
RedHat: You either pay or don't (download). It's Linux.
Support:
Windows: Support costs you hell a lot of money
RedHat: If you can't afford to pay RedHat, it's Linux for God's sake. There're thousands of people on the net willing to help you.
Documentation:
Windows: None
RedHat: It's Linux, damn it. RTFM
Source code:
Windows: None
RedHat: It's Linux. You get the source code.
Patches:
Windows: Waiting for patches if Microsoft has the time and mood to fix it. Service packs come out once in a blue moon.
RedHat: It's Linux. Thousands of people have access to the source code. Bug fixes come out rapidly.
Hmm.... Windows is better than RedHat?
...they're running SAMBA. For balance I think they should test Windows 2003 throughput of NFS.
Bob
Listen to my latest album here
Microsoft pays for a test that shows that Windows 2003 is twice as fast. That's nice. But not very productive.
I see nothing in the report that they had a Red Hat guru optimize the Red Hat server. It is easy to get the results you want. If I don't see proof that Red Hat was configured by a Red Hat guru, as I am sure that Windows 2003 was optimized by a Microsoft guru, then the tests are bunk.
-Brent
put any person in the position of the high management in a company in the same position as Microsoft and they're going to start doing things like this in the persuit of profit. It is not corruption in that they go each morning thinking "what bad things can I do today" like how a psychopath would, but a self-delusional and untimately self-defeating cultistic denial of the effects of their actions in the name of "just business" or "creating great new products" or some such justification.
It would be easy to give some thing about "capitalism being the problem", but human nature is the problem when people are in a system whereby those in positions of authority are being constantly goaded to do the wrong thing whilst what they do is being portrayed by the social system as being right - the same thing happened in the USSR.
And while you o Slashdotters complain about Microsoft, the mass of open source programmers have much the same self-delusional corruption in relation to computing in a different form. You delude yourself into wasting resource by upgrading as you attach your personal worth to the number of mhz of your computer, you spend all your programming time in the pursuit of self-centred 'fun', etc. I know there are exceptions among you, as there are in some businesses, but take the beam out of your own eye before taking absolutist positions of moral authority over the likes of Microsoft.
It's no wonder that there is a community of Linux-bashers out there. Michael and the rest of the Slashdot editor cabal spew their little anti-MS tantrums on the front page and expect to be taken seriously. I challenge anyone who would even consider modding this post as -1 to *respond first*. Really, I'm waiting to hear from any slashdotter that *isn't* ashamed at Michael's rantings.
Because the test was commissioned and paid for by Microsoft. They knew the results before they paid for the test - the test was ordered in such a way as to make the results a forgone conclusion.
If you were going to contruct a biased test of Linux vs. Microsoft, wanting Linux to prove better - you'd choose products and tests that would favor Linux... just as Microsoft did here. Take a native protocol to Microsoft and stack it up against a re-implementation on another operating system... sounds pretty weighted against Linux to me.
Of course there are lots of ways to make an unbiased performance comparison - such as using multiple protocols, etc... but why would Microsoft pay for that? I'm not saying that either system was better or worse than the other - just that this particular test doesn't prove anything except that Microsoft know how to spec a test so that their products appear faster.
-- "Other than that, how was the play Mrs. Lincoln?"
...With regard to IT's (real!) need to lock down desktops...
We need a new definition of OS stability.
Today, "stability" basically refers to the ability of an OS to run without crashing _in the absense of configuration changes_.
In the real world, there are ongoing needs to install new software, apply patches, updates, etc.
In a system that had proper modular design, it should be possible to install something new or change a legitimate setting without feeling that you're playing Russian Roulette.
CERTAINLY it should be possible to install vendor-recommended updates with a high level of confidence that it's not going to break something.
Remember all that stuff a few years back, that implied that the problem with stability was that people weren't keeping their systems properly updated and that "self-healing" systems would fix that? Well, now, we all but have them, and, in fact, it's made things worse.
"How to Do Nothing," kids activities, back in print!
Its so horrible to see a company like this doing dirty business everywhere, getting sued like mad, and yet, their stock is UP!!!!!
But no, this company is not a monopoly at all.
Get paid to code OSS
From what they say in the wired article I can see DRM being extremely dangerous. If you can sign an email, making it only viewable by the intended recipient and stopping them from print/forwarding/saving, then you could very easily mount a campaign of e-stalking, sendingv vulgar and/or threatening emails and the person receiving them unable to remove them or forward them to anybody who can help.
"Of course, this is just one part of Microsoft's strategy against Linux and OSS. But I'm pretty sure that this salvo will fall on deaf ears."
You're right. Nobody has been listening to the Samba people claim their server was faster.
This was more of a pride thing to Microsoft. They saw an article showing their server was not the fastest one out there, and so they went to their developers and said "fix this."
And so they did. Now it's not just faster, it's signifigantly faster.
That's the nature of competition!
I used to work for Best Buy. We were given the whole deal about getting out as many MSN disks as we could. The thing about that is, MSN kicks back a good amount of money for each free account that someone signs up for. Whether or not it's the 30 free days, 2 months, or 9months (or whatever). Heres what happens. Say a customer buys a computer at Best Buy, they automatically get 6 free months (if they pay with a credit card or major debit). If the customer doesn't sign on and activate that account, they never get billed. However, even if they sign on just once, even for just 30 seconds, they have to call and cancel the account.
Now, if a customer signs up for the 30 day free trial, yes they have to give a creditcard number, but they have to call and cancel the account even if they don't sign on. I had that problem but I called MSN and got this resolved and my money refunded immediately.
Now, the problem is that some bestbuy clerks don't bother to tell the customers this. I've had customers outright refuse the msn and i said that's fine and just don't go through with ringing it up. I always mentioned it to all the customers I talked to, and if they just didn't want it, I never scanned it. This class action probably won't work, but I hope this guy gets his money back.
Anyway that's my piece!
It's already been put out there that Microsoft wants a piece of the search engine pie. Their current MSN search is horrible, and Google is far superior. Is it competition for Google? Maybe. And I welcome it. Google's done some amazing things with their search engine, and I'd love to see what they can do with a hefty bit of competition. Because right now they really have none. If Microsoft can build a better search engine then let them try.
"I'm a leaf on the wind. Watch how I soar."
-Hoban Washburn
[Just to be clear, this is the Wired article talking aobut Longhorn]
"This is scary stuff," said a developer who asked that his name be withheld. "I could see a lot of people sticking with their old computers, operating systems and media players to avoid all this permission crap. Any geek who does use Windows is going to stick with Windows 2000; most of them are already not thrilled with XP anyway."
Mmm hmm...yea. The same thing was said about Internet Explorer 4.0 & Windows 98 (just substitute the words "web browser integration" in for "permission", and it should bring back memories). No one was going to upgrade because no one wanted their internet integrated into the operating system. But whoever was saying such a thing didn't think about this key issue:
The average joe does not care, let alone know about integrating a web browser into an OS. It doesn't matter if the nerd police showed up on the opening day of Windows 98 to tell people how evil it was.
Joe wants a new computer with all the new bells and whistles. If Longhorn says that "it will make the internet come alive with all sorts of new technologies...all you need is Longhorn", then so be it, Joe's going to get Longhorn, because the internet is "cool".
Take a native protocol to Microsoft and stack it up against a re-implementation on another operating system... sounds pretty weighted against Linux to me.
I just realized: does this mean Microsoft officially recognizes Samba as legitimate competition?
Healthcare article at Kuro5hin
Is Microsoft working on supporting PNG in Internet Explorer? This is the single most significant feature that other browsers have over IE, from a web developer's viewpoint.
... If this is indeed what happened, it constitutes a criminal offense, and involves organized crime and banks too, making it a Federal crime.
However, the article also says that the guy "has not been unable" to get a rebate from either company. That's more than a teeny bit unclear thanks to the (probably accidental) double negative - I bet neither of them have given him a rebate yet. ha.
In times past, Linux (or *BSD) plus Samba has outperformed MS-Windows on the same hardware. This is quite important, and the reason this test was commissioned.
Secondly, note that no real test results were provided; the report merely states that MS-Windows provided a higher *peak* throughput. Please realize that real-world performance does not rely on peak throughput as much as it relies on aggregate *sustained* throughput.
It could be that Samba still knocks the socks off MS-Windows in that more-important category. But, until some legitimate benchmarks are run, Microsoft will continue to pay for FUD.
BTW: several quibbles with testing methodology. First, no optimisations were done to the Linux box (no noatime option on the filesystem mount, for instance). Second, they didn't test against an optimized kernel (which is fair, I guess, as most people will stick with a stock install; however, most people won't do those MS-Windows tweaks, either). Finally, this was tested against an aging 2.4 kernel, and not against either the newer 2.4 kernel, or against any of the later 2.5 builds. With the SMP, low-latency, and I/O buffs in the new 2.5 series, I imagine the outcome would be quite different.
But, that's just my opinion. I could be wrong.
Microsoft is to software what Budweiser is to beer.
My main problem with using MS for everything is that their patches come in service packs and not singly. I know of a more than a couple Win32 servers which are firewalled instead of completely patched because the admins just aren't totally sure what all a hotfix/service pack/update will do to their machine (why they don't have other hardware they can test on, I don't know; I suspect time is a limiting factor). This is part of the reason why slapper was so big (the update being a pain to install also helped). That critical updates come with other choices which have been made for you is a hard pill for me to swallow. The Win32 servers at work also seem to have more downtime, but I don't have any numbers compiled so i can't really argue the point.
I'm no great fan of MS, but you are right: thery are doing some pretty impressive stuff. For a small to medium-sized workgroup, I think the two are roughly equivalent. Given the choice I'd personally pick Red Hat primarily because of the patching issues, licensing/cost, familiarity in our workplace, and freedom of choice. But just because some other group doesn't make that same set of choices doesn't mean they are necessarily "losing" anything or are worse off.
-B
Ash and Hickory, straight-grained and true, make excellent bludgeons, dandy for the cudgeling of vegetarians.
EvoX and XBox Media Player have turned my XBox into the first Micro$oft product I actually enjoy using! Streaming video, audio, and pictures over my home LAN and playing DVDs without a dongle is awesome! And then, of course, there is the satisfaction of playing Super Mario Brothers and Metroid on a big-old console controller again. Takes me back to 1987!
If you read the actual doc, it says Windows is a better CIFS server than RedHat. Very different conclusion than the very broad "Windows is a better server than Red Hat".
The sad truth is that Slashbork would be exactly nowhere without Microsoft.
No, you're not even close. /. would be a quirky online news site where there would be holy wars about trivial topics (Linux vs BSD vs UNIX) and other things.
MS has only caused a vast majority of those quirky people to come together in jihad against a company whose business ethics are non-existant and whose products are of dubious quality. Although they have a STELLAR marketing group.
Computer Science is Applied Philosophy
Except the benchmark uses questionable methods, so most people who this could matter to will discount it.
Thats the nature of fabrication!
In addition, note that the performance of Windows 2003 server varied a whole lot depending on the number of users, processors, etc. while the Linux boxes stayed relatively steady and relatively consistent.
/etc/fstab, as you allude to in your message.
That tells me that Linux is really the better performer and the only reason its running slower is some artificial performance limitation, like the absence of setting the 'noatime' directive in the mount options in
Also, what file system was in use? I don't have Red Hat Advanced Server 2.1, but I'll bet it installs ext3 by default, rather than the higher performance reiserfs or xfs. Now, I wouldn't install xfs, it's too unstable, but reiserfs has good performance along with rock-solid reliability on 2.4.18 and later. ext3 is slllloooowwww.. mostly because it journals metadata *and* data, while reiserfs only journals metadata and uses B* trees.
My journal has hot
arent you doing that already?
[sorry it was a cheap slab]
I know you are psychotic, but please make an effort.
Strictly speaking a rather large flaw was found in the testing protocols. Now he is going out on a limb with: "This is enough to account for these differences, I suspect." He's bang on with: "They disabled last access time updating under windows. They didn't under Linux."
No, he can't say from that Windows and Linux have equivalent performance under those circumstances. However, neither can MS use these results to confirm Windows superiority. Disabling access time updating is known to be a large optimization for those who can do without it. If that methodology is typical of this study then it's likely that the claimed results are utterly worthless for any conclusion whatsoever.
I understand that the employees are under pressure from their bosses to sell all of this stuff. But you work at a store that sells something that I want, and I have the right to buy that without having to buy all this extra stuff. If you don't like having to deal with customers, then perhaps you should work at a different job. Basically, what you are saying is that customers should spend more of their own money so that you don't have to deal with the inconvienience of your job.
The Samba team got a hold of this about a week ago. These benchmarks are a little off.
For instance, they're comparing Win2k3 vs. Samba 2.2.7. We're rather close to the 3.0 release of Samba and the 2.2 base hasn't really been worked on in a long time.
Moreover, RHAS is actually slightly older than RH8.0 (a lot older than RH9.0). That's why the one benchmark with all three systems showed RH8 beating RHAS. I believe that RHAS didn't ship the O(1) scheduler.
I've also heard claims that the real reason behind the difference in throughput was the poor software raid used in the benchmark machines. Had a supported hardware RAID been used, things would have been pretty different.
Not to mention the "tuning" done to the two systems. The socket buffers were tweaked and the file descriptors increased on the linux side while a bunch of strange registry options were set on the Windows side. There could have been a lot more tuning done on the linux side to improve performance.
Of course, what would you expect from a study commissioned by Microsoft. What someone should do is let the Samba team set up a machine and some Microsoft folks set up another machine. Then we'll see who outperforms who.
int func(int a);
func((b += 3, b));
I noticed a lot of benefit given to microsoft in that test. Comparing w2k3 (not yet released?) with red hat 8 and as2 (both out of date). As others have pointed out, using SAMBA/microsoft networking. What about all those other protocols? Servers usually do http and such. I didn't notice this, but since I'm listing, w2k3 was tuned and redhat wasn't. And of course, microsoft paid for the test. We should have all known that since it was a microsoft test they would win, and be given whatever benefit necissary.
This is such bullshit. You have a job. If you do not like your job, quit. If you cannot do you job and get bitched out by your supervisors, quit. It is not the customers fault that you cannot do your job. It is not your supervisors fault that you cannot do your job. It is your fault you cannot do your job. Your job is to sell things to customers in such a way that your supervisors will be happy. If you cannot do that, then you do not deserve the job. Your hate customers because you know you do not do the job that is required. I feel for you, because perhaps this is the only job you can have. But I would think that perhaps you might feel pity for the customers that you lie to, feel relief when customers do not fall for the lies, and responsibility for perpetuating the lies.
"She's a scientist and a lesbian. She's not going to let it slide." Orphan Black
I would believe this if this was the only complaint I had ever seen. I work at MSN Customer Service and we litterally get hundreds if not thousands of complaints a month. And we're just one call center. And the complaints don't come from every retailer. We don't hear a peep out of people who were signed up at Staples, or Circuit City, or Radio Shack (well, unless it's a poed customer with a term commit). How can it be that only Best Buy customers complain? Are they somehow more stupid/evil than other consumers?
Now, perhaps everything IS on the up and up at your particular store. I've ALSO noticed we only get complaints from certain regions (OC California seems to be popular for some reason). This leads me to believe the problem (always assuming there is one) is only with a few stores, and not universal. As the saying goes though, it only takea a few bad apples...
Regardless, I know the paperwork IS supposed to be included. The username, password AND terms and conditions are printed right on a receipt. It's unbelievable how few people actually look at these things. Cripes, I've had customers call up and complain about being billed FOR OVER AN F'ING YEAR and claim they just noticed. Either they're trying to scam us, or people really are that stupid, or most likely both.