Who is going to want to compete in making an alternative exchange server other than Free software proponents and perhaps Apple.
Don't underestimate Apple. They've got a server product, it supports some of the most common things served today out of the box (such as HTTP, LDAP - don't know if they've also got an IMAP server in there but I'd think so).
I reckon they'd love to implement 100% exchange compatability.
Basically HDDs = media + drive, and they are about the same price as tapes on a per GB basis if not cheaper. Multiple HDDs have better bandwidth than multiple tapes with one tape drive.
Go read the specs for a single LTO-3 U160 SCSI tape drive. It can generally sustain a transfer rate at (or close to) the speed of the bus itself - 160MB/sec.
That's SUSTAIN. Not burst. No hard disk on Earth can do that right now. (Actually, relatively few computers can because you wind up with bus contention issues).
Raid 5 is damn slow for anything that requires serious I/O rates. And if you've got a lot of drives and/or a very large array, requires a long time to rebuild if a drive fails.
If you need redundancy and speed, RAID10 is probably the way to go.
I know you should be modded insightful by rights, but a lot of how well this works in practise (particularly with large chains) is going to depend on how clued up the store manager is with regard to the law.
More often than not, the answer is "Not very". And the training is "when in doubt, refer to company policy" - and I've seen plenty of examples in recent years where company policy is in complete contradiction to statutory rights.
So instead of an emulation layer which is almost adequate for most tasks, but needs careful testing with anything you're planning on using in production because any action you try performing may or may not work as designed, we'll have a full blown OS which is almost adequate for most tasks, but needs careful testing with anything you're planning on using in production because any action you try performing may or may not work as designed.
More recently, they added the "do everything to maintain the Windows monopoly" strategy.
True. I was talking about historically.
I'm not sure Microsoft are still capable of changing direction quickly. Certainly there's been no evidence of it for some years. Without that ability, I think they'll do what IBM did - start looking more and more likely to render themselves obsolete until sooner or later a new person comes in at the top, makes major changes and reinvents the company. (Actually, I'm rather looking forward to the period just before that happens - I don't think we can truly say we're in it until such time as the Windows monoculture is well and truly dead - as I think it will herald a whole lot more innovation in IT)
I know this is absolutely begging to be modded Troll, but let's get real for a minute.
The web's been around a few years now. While they were late in recognising it, Microsoft have been taking the Internet seriously since before Google left Stanford University.
IMO, if Microsoft were able to develop "better search than Google.... better hosting than Amazon..." - they'd have done so long ago. As it stands, they can't even implement searching in their own OS (certainly not in XP - even with the Search addon, it's trivially easy to dig out something which returns zero results when it patently shouldn't) - and they've got far more control over that than Google has over the Internet.
Fact is, Microsoft's business plan has never been "build a better OS/office suite/mousetrap". It's been "build one that's good enough and market it as being better". But such marketing doesn't work so well in the Internet age because it's much easier to find out how much truth there is behind it, and AFAICT Microsoft still haven't worked that one out.
RTFA. The author said that they were told about Dell's N series (HP and Lenovo apparently have equivalents) but when they said "OK, so how do I buy one?" the answer was "Er... I don't know.... I know it's not on our website."
My guess (and it's pure speculation) is you probably need an account manager to place the order, which implies that you already have a business account. ZDnet, OTOH, seemed to be ringing up out of the blue without having an account manager.
For businesses (which, I'd point out, are the kind of people this article is aimed at - save £30 on one PC, big deal. Save £3,000 on a hundred PCs because they're going into an environment where Windows isn't required - that's more interesting, particularly if your budget is already tight) - having just one number to call when something goes wrong (rather than spend hours having Seagate tell you that your hard disk is fine, it's your motherboard which is on the fritz and the motherboard manufacturer essentially tell you the reverse of that) has great value,
Um, unless they just use the XP licenses they already have?
Anyway, MS licensing works differently if you're a 2000+ seat university compared to some lone windows fanboy running vista ultimate.... I don't think this is going to be a gain for MS at all.
Quite correct. Butthe immense likelihood is they don't have any XP licenses. They may have thousands of OEM licenses - which are tied to the PC they were bought with - and thousands of XP upgrade licenses (which are about all you can buy through Microsoft's volume licensing scheme) but without either an OEM or full retail licence to begin with, you can't legally use the upgrade license you get in the site license.
However, one must remember the old adage about "no publicity is bad publicity." The whole SCO case attracted quite a large deal of attention, and a lot of speculation on the legitimacy of linux and/or its licenses.
That's true. Around the time it started out I remember I had a conversation with someone who'd never really touched Unix professionally - he was a VMS man. Despite never really going that near anything Unix/Linux-y, the opinion he'd formed was along the lines of "SCO are a nasty bunch and I don't want anything to do with them".
Then SCO started suing their own customers. Trust me, there is such a thing as bad publicity, and "suing your own customers while shouting about it to every journalist who'll listen to you" qualifies.
TOR is a great tool, but all it would do in this context is remove one (probably not very important) piece of evidence out of many.
Things like card receipts for the weapon, CCTV evidence of the shop where the weapon was purchased, evidence gathered from whatever sources as to where the defendents car was around the time of the murder.
There's also a number of other questions which spring to mind:
1. What are the Linux servers doing? If you've got a cluster of boxes all doing the same thing, the way you administer them in Linux would be rather different to how you might do so in Windows.
2. What applications does the business use and what are their dependencies? One of the drawbacks of Linux vs. Windows is that closed-source applications (and whether you like it or not, they're a fact of commercial life) are generally compiled against a specific version of basic libraries. Not generally a big problem in Windows, but it suddenly introduces lots of questions on Linux. Like "We can't update this server because the application won't work with the update we'd like to do, but every other server should be updated because the update in question is important.... Oh look, now we've got a disparate bunch of servers". Doesn't take long before you're not sure what server is running what.
3. Group policies. Say what you like about Microsoft, GPO is slick and works out of the box on Windows. There is no immediate equivalent on most Linux distributions (though I understand Novell's ZenWorks may go some way to solving this).
Sure, you can lash together something yourself using shell scripts and cron, or other options exist if you're feeling flash, but they all require a bunch of work and none work out of the box. A basic group policy, on the other hand, can be set up in a few hours quite easily and once set up, is rolled out to everything with very little effort.
There is also the pretty big issue that every business is different. What works for me may not work for your, and the only way you'll find out what the cheapest solution for you is is by looking at your own business needs, not by looking at someone else's.
I think my best one to date is "We keep on getting pages returned with "Welcome to Internet Information Server""
ISP: Must be something wrong with the website you're visiting Me: So Google are running IIS are they? ISP: They must be. Me: Look, I've looked into this and I can see that in your cluster of proxy servers, the one with IP address 195.xxx.xxx.xxx is the one that keeps on returning that page. Every other one is fine. Would you please ask someone to look at it and let me know when it's fixed? ISP: Er....
The scary thing is that my boss at the time had signed a 1-year contract with these morons so I couldn't drop them no matter how much I wanted to. Apparently, after I left, my successor rang them up to give them a months' notice as per the contract and they'd terminated the connection before he'd even put the phone down. Seems they didn't like dealing with people who knew more than they did.
Is it just me, or doesn't anyone else thing that MS must be rather worried about the fact that a large manufacturer is looking hard at selling a non-MS operating system?
I'm sure they are. I wonder what time of year the Microsoft/Dell contract comes up for renewal?
Ship a bootable diagnostics CD. If that doesn't throw up the problem, the problem doesn't exist.
OK, that's next to useless for memory issues and dying (but still a long way off actually dead) hard disks, but quite frankly I don't see how that differs from Dell (or any other Tier 1 vendor's) current desktop support.
As an IT manager, I can quite happily confirm that in the real world, getting Windows to do what you want (rather than what Microsoft thinks is "right") is a computer science project in itself.
It's Australian dollars, but even so, the short answer is almost certainly "no".
By the time business becomes large enough to warrant the attention of the BSA, it's large enough to qualify for "volume discounts" - which are generally some absurd percentage. I'm paying about UK£120/annum for Office and I'm only buying 50 licenses. The "official" UK retail price is about UK£400 - that's if you go down the store and buy a boxed copy.
I think the logic is "we don't expect to sell a single copy at the full price, but it means that we can make it look like a bargain with generous discounts and if we take them to court, we demand full price for every copy".
Except that a Linux distribution is more than just the kernel.
All the tools used to compile the kernel are FSF-owned, and will go over to GPLv3. Things like bash, grep, gimp are also included.
A handful of other major projects have also made pro-GPL3 noises - Samba being the most well known.
Now, there's nothing to stop me taking the last GPLv2 version of these, entering into an agreement with someone which would stop me distributing GPLv3 code but not v2 and simply never upgrading the versions I ship - or alternatively maintaining my own versions complete with clean-room reimplementations of any new features in the GPLv3 software as and when it's updated - but that immediately puts me at a massive commercial disadvantage to anyone else who's not made such an agreement.
If I never update the versions of software I ship, then my distribution starts to look rather dated inside a year, and positively antiquated inside 3.
If I go down the "reimplement every new feature" route, my software development costs go through the roof because all of a sudden I have to keep up with all this GPLv3 software but I don't have anything like the number of volunteer developers working on my code that I did before. There's also a strong chance that the agreement I made earlier will effectively make a lot of project leaders who are using GPLv3 think twice before accepting patches from me - that's assuming I can submit such patches in the first place.
The issue with RFID passports would be if they could be/forged/... it doesn't matter if they can be duplicated.
Not true.
There's a lot to be said for not bothering to forge passports anyway - sooner or later customs at most first-world countries will probably link up, so the passport number can be checked instantly against a database to make sure the details match up. The only way a "forged" passport will work then is if it's not forged at all, but rather made with the collusion of someone at the passport office.
However, if you can duplicate a passport, you can pretend to be someone else. Someone who (you hope) has no criminal record and is not even vaguely interesting to the authorities. With access to a crooked person in authority, you can confirm this. Without such access, you simply make a few flights and see if you get stopped. The only way I can see around this is if government starts tracking where everyone is, and if the passport handed over at customs belongs to someone you know for a fact was a thousand miles away only ten minutes ago, you know something fishy's going on. But we're a long way from having that level of technology - and while I absolutely hate the sound of it, I wouldn't be even remotely surprised if someone in government is mulling it over right now.
Firstly, I agree with earlier suggestions that this is just a ruse to get cheaper prices from MS.
Even so, with the kind of money the government has available, "getting Google to put their apps onto an appliance and selling the FAA a few dozen|couple of hundred such appliances" shouldn't be too difficult.
Slashdot Mirror
Who is going to want to compete in making an alternative exchange server other than Free software proponents and perhaps Apple.
Don't underestimate Apple. They've got a server product, it supports some of the most common things served today out of the box (such as HTTP, LDAP - don't know if they've also got an IMAP server in there but I'd think so).
I reckon they'd love to implement 100% exchange compatability.
Basically HDDs = media + drive, and they are about the same price as tapes on a per GB basis if not cheaper. Multiple HDDs have better bandwidth than multiple tapes with one tape drive.
Go read the specs for a single LTO-3 U160 SCSI tape drive. It can generally sustain a transfer rate at (or close to) the speed of the bus itself - 160MB/sec.
That's SUSTAIN. Not burst. No hard disk on Earth can do that right now. (Actually, relatively few computers can because you wind up with bus contention issues).
Raid 5 is damn slow for anything that requires serious I/O rates. And if you've got a lot of drives and/or a very large array, requires a long time to rebuild if a drive fails.
If you need redundancy and speed, RAID10 is probably the way to go.
I know you should be modded insightful by rights, but a lot of how well this works in practise (particularly with large chains) is going to depend on how clued up the store manager is with regard to the law.
More often than not, the answer is "Not very". And the training is "when in doubt, refer to company policy" - and I've seen plenty of examples in recent years where company policy is in complete contradiction to statutory rights.
WINE isn't good enough for everyone.
So instead of an emulation layer which is almost adequate for most tasks, but needs careful testing with anything you're planning on using in production because any action you try performing may or may not work as designed, we'll have a full blown OS which is almost adequate for most tasks, but needs careful testing with anything you're planning on using in production because any action you try performing may or may not work as designed.
Great. Thanks.
I've got a file here which isn't secret, existed before the indexing tool was run and still isn't found after running it.
I know one anecdote isn't data, but please don't imagine I haven't taken account of that. I have.
More recently, they added the "do everything to maintain the Windows monopoly" strategy.
True. I was talking about historically.
I'm not sure Microsoft are still capable of changing direction quickly. Certainly there's been no evidence of it for some years. Without that ability, I think they'll do what IBM did - start looking more and more likely to render themselves obsolete until sooner or later a new person comes in at the top, makes major changes and reinvents the company. (Actually, I'm rather looking forward to the period just before that happens - I don't think we can truly say we're in it until such time as the Windows monoculture is well and truly dead - as I think it will herald a whole lot more innovation in IT)
I know this is absolutely begging to be modded Troll, but let's get real for a minute.
The web's been around a few years now. While they were late in recognising it, Microsoft have been taking the Internet seriously since before Google left Stanford University.
IMO, if Microsoft were able to develop "better search than Google.... better hosting than Amazon..." - they'd have done so long ago. As it stands, they can't even implement searching in their own OS (certainly not in XP - even with the Search addon, it's trivially easy to dig out something which returns zero results when it patently shouldn't) - and they've got far more control over that than Google has over the Internet.
Fact is, Microsoft's business plan has never been "build a better OS/office suite/mousetrap". It's been "build one that's good enough and market it as being better". But such marketing doesn't work so well in the Internet age because it's much easier to find out how much truth there is behind it, and AFAICT Microsoft still haven't worked that one out.
RTFA. The author said that they were told about Dell's N series (HP and Lenovo apparently have equivalents) but when they said "OK, so how do I buy one?" the answer was "Er... I don't know.... I know it's not on our website."
My guess (and it's pure speculation) is you probably need an account manager to place the order, which implies that you already have a business account. ZDnet, OTOH, seemed to be ringing up out of the blue without having an account manager.
For home users, this is no big deal.
For businesses (which, I'd point out, are the kind of people this article is aimed at - save £30 on one PC, big deal. Save £3,000 on a hundred PCs because they're going into an environment where Windows isn't required - that's more interesting, particularly if your budget is already tight) - having just one number to call when something goes wrong (rather than spend hours having Seagate tell you that your hard disk is fine, it's your motherboard which is on the fritz and the motherboard manufacturer essentially tell you the reverse of that) has great value,
Um, unless they just use the XP licenses they already have?
Anyway, MS licensing works differently if you're a 2000+ seat university compared to some lone windows fanboy running vista ultimate.... I don't think this is going to be a gain for MS at all.
Quite correct. Butthe immense likelihood is they don't have any XP licenses. They may have thousands of OEM licenses - which are tied to the PC they were bought with - and thousands of XP upgrade licenses (which are about all you can buy through Microsoft's volume licensing scheme) but without either an OEM or full retail licence to begin with, you can't legally use the upgrade license you get in the site license.
However, one must remember the old adage about "no publicity is bad publicity." The whole SCO case attracted quite a large deal of attention, and a lot of speculation on the legitimacy of linux and/or its licenses.
That's true. Around the time it started out I remember I had a conversation with someone who'd never really touched Unix professionally - he was a VMS man. Despite never really going that near anything Unix/Linux-y, the opinion he'd formed was along the lines of "SCO are a nasty bunch and I don't want anything to do with them".
Then SCO started suing their own customers. Trust me, there is such a thing as bad publicity, and "suing your own customers while shouting about it to every journalist who'll listen to you" qualifies.
TOR is a great tool, but all it would do in this context is remove one (probably not very important) piece of evidence out of many.
Things like card receipts for the weapon, CCTV evidence of the shop where the weapon was purchased, evidence gathered from whatever sources as to where the defendents car was around the time of the murder.
Also consider that in any murder case, the prime suspects are the victim's immediate relatives.
There's also a number of other questions which spring to mind:
1. What are the Linux servers doing? If you've got a cluster of boxes all doing the same thing, the way you administer them in Linux would be rather different to how you might do so in Windows.
2. What applications does the business use and what are their dependencies? One of the drawbacks of Linux vs. Windows is that closed-source applications (and whether you like it or not, they're a fact of commercial life) are generally compiled against a specific version of basic libraries. Not generally a big problem in Windows, but it suddenly introduces lots of questions on Linux. Like "We can't update this server because the application won't work with the update we'd like to do, but every other server should be updated because the update in question is important.... Oh look, now we've got a disparate bunch of servers". Doesn't take long before you're not sure what server is running what.
3. Group policies. Say what you like about Microsoft, GPO is slick and works out of the box on Windows. There is no immediate equivalent on most Linux distributions (though I understand Novell's ZenWorks may go some way to solving this).
Sure, you can lash together something yourself using shell scripts and cron, or other options exist if you're feeling flash, but they all require a bunch of work and none work out of the box. A basic group policy, on the other hand, can be set up in a few hours quite easily and once set up, is rolled out to everything with very little effort.
There is also the pretty big issue that every business is different. What works for me may not work for your, and the only way you'll find out what the cheapest solution for you is is by looking at your own business needs, not by looking at someone else's.
AOL.
I think my best one to date is "We keep on getting pages returned with "Welcome to Internet Information Server""
ISP: Must be something wrong with the website you're visiting
Me: So Google are running IIS are they?
ISP: They must be.
Me: Look, I've looked into this and I can see that in your cluster of proxy servers, the one with IP address 195.xxx.xxx.xxx is the one that keeps on returning that page. Every other one is fine. Would you please ask someone to look at it and let me know when it's fixed?
ISP: Er....
The scary thing is that my boss at the time had signed a 1-year contract with these morons so I couldn't drop them no matter how much I wanted to. Apparently, after I left, my successor rang them up to give them a months' notice as per the contract and they'd terminated the connection before he'd even put the phone down. Seems they didn't like dealing with people who knew more than they did.
Is it just me, or doesn't anyone else thing that MS must be rather worried about the fact that a large manufacturer is looking hard at selling a non-MS operating system?
I'm sure they are. I wonder what time of year the Microsoft/Dell contract comes up for renewal?
Ship a bootable diagnostics CD. If that doesn't throw up the problem, the problem doesn't exist.
OK, that's next to useless for memory issues and dying (but still a long way off actually dead) hard disks, but quite frankly I don't see how that differs from Dell (or any other Tier 1 vendor's) current desktop support.
This doesn't exactly produce a fantastic advert for Linux.
"Use Linux on the Desktop! It's great, except that some very basic desktop functionality has clearly had zero testing as IT DOESN'T WORK!!"
That may be, but you can't really hold Windows responsible for HPs software.
As an IT manager, I can quite happily confirm that in the real world, getting Windows to do what you want (rather than what Microsoft thinks is "right") is a computer science project in itself.
It's Australian dollars, but even so, the short answer is almost certainly "no".
By the time business becomes large enough to warrant the attention of the BSA, it's large enough to qualify for "volume discounts" - which are generally some absurd percentage. I'm paying about UK£120/annum for Office and I'm only buying 50 licenses. The "official" UK retail price is about UK£400 - that's if you go down the store and buy a boxed copy.
I think the logic is "we don't expect to sell a single copy at the full price, but it means that we can make it look like a bargain with generous discounts and if we take them to court, we demand full price for every copy".
Except that a Linux distribution is more than just the kernel.
All the tools used to compile the kernel are FSF-owned, and will go over to GPLv3. Things like bash, grep, gimp are also included.
A handful of other major projects have also made pro-GPL3 noises - Samba being the most well known.
Now, there's nothing to stop me taking the last GPLv2 version of these, entering into an agreement with someone which would stop me distributing GPLv3 code but not v2 and simply never upgrading the versions I ship - or alternatively maintaining my own versions complete with clean-room reimplementations of any new features in the GPLv3 software as and when it's updated - but that immediately puts me at a massive commercial disadvantage to anyone else who's not made such an agreement.
If I never update the versions of software I ship, then my distribution starts to look rather dated inside a year, and positively antiquated inside 3.
If I go down the "reimplement every new feature" route, my software development costs go through the roof because all of a sudden I have to keep up with all this GPLv3 software but I don't have anything like the number of volunteer developers working on my code that I did before. There's also a strong chance that the agreement I made earlier will effectively make a lot of project leaders who are using GPLv3 think twice before accepting patches from me - that's assuming I can submit such patches in the first place.
The issue with RFID passports would be if they could be /forged/... it doesn't matter if they can be duplicated.
Not true.
There's a lot to be said for not bothering to forge passports anyway - sooner or later customs at most first-world countries will probably link up, so the passport number can be checked instantly against a database to make sure the details match up. The only way a "forged" passport will work then is if it's not forged at all, but rather made with the collusion of someone at the passport office.
However, if you can duplicate a passport, you can pretend to be someone else. Someone who (you hope) has no criminal record and is not even vaguely interesting to the authorities. With access to a crooked person in authority, you can confirm this. Without such access, you simply make a few flights and see if you get stopped. The only way I can see around this is if government starts tracking where everyone is, and if the passport handed over at customs belongs to someone you know for a fact was a thousand miles away only ten minutes ago, you know something fishy's going on. But we're a long way from having that level of technology - and while I absolutely hate the sound of it, I wouldn't be even remotely surprised if someone in government is mulling it over right now.
Firstly, I agree with earlier suggestions that this is just a ruse to get cheaper prices from MS.
Even so, with the kind of money the government has available, "getting Google to put their apps onto an appliance and selling the FAA a few dozen|couple of hundred such appliances" shouldn't be too difficult.