I've worked in a few companies, and in my experience what you say is true... up to a point.
"The point" in question is the point where the OS is of any importance. You may, for instance, have a core database/app server running something supported like Red Hat (or even a commercial Unix), then apache frontends which aren't really doing much running CentOS.
In at least one case I can think of - granted, this was a few years ago - everything was so tied to the database that the database was commercial, paid for and supported. But the OS was treated as a commodity. (Mind you, it probably helped that we had a fairly extensive contract with IBM, all the servers were identical and barring hardware or power failure, it was practically unheard of for anything significant to happen to the running OS).
CentOS essentially acts as advertising for the Enterprise RedHat editions. It allows sysadmins to stick with the same familiar set of tools on both systems where it is considered desirable to have a support contract and systems where this is less of an issue.
RedHat can't do much to curb this anyway - most of what they produce is standing on the shoulders of other GPL software - but if they did, I'd imagine we'd see a commensurate rise in the use of Debian, Ubuntu and (gasp!) SuSE/OpenSuSE.
interesting thinking to claim that murder is somehow forgivable.
Excuse me while I play devil's advocate here - am I to understand that your point of view is that life imprisonment should mean life?
Because if not, then the idea that society should never forgive a murderer to my mind implies that anyone who is let out of prison after serving such a term should never be able to find work, should be denied even the most basic of social housing or benefits and should essentially have no choice but to wind up living as a tramp, wandering the streets, drinking methylated spirits and shouting at people who aren't there.
As for the reliability of the information gathered: Is information gathered in this way admissible? It would seem that there are to many potential snags with this that it would never be able to be relied upon by itself.
Does it need to be? I know nothing of German law, but in many western countries this would be the first step of intelligence gathering prior to mounting a raid, not the be all and end all.
The problem likely stems from hiring programmers that aren't familiar with porting software to the Linux platform.
There's a simple reason for that.
They don't have an IT department any more. It was outsourced some time ago - shortly before it became apparent that the Internet was likely to become if not the future of media distribution, then at least very significant.
The development process seems to have been "throw something cheap and cheerful together at the last minute". It had to include DRM because pretty much everything the BBC produces is wrapped up in all sorts of licensing. Heck, even Tuesday's episode of Eastenders used some recent commercial music in the background at one point - that would have to be licensed under specific terms, and it's exactly that kind of thing which prevents the BBC from making all their content free to all.
You misunderstand me, but then I could perhaps have worded it more clearly.
In answering the question "do Apple want to let OS X run on third-party hardware?", the fact that Dell wouldn't be shipping OS X preinstalled doesn't really enter into the matter. Apple doesn't make enough money out of OS sales to survive on that alone, so they don't want people taking the cheaper option of "cheapest Dell available plus OS X".
There are two things Apple can do to prevent people doing this:
1. Print on the box "Not to be installed on non-Apple hardware". Dubious how legally effective this would be, would result in them playing whack-a-mole with casual end users which would endear them to no-one.
2. Make an effort to render this difficult/impossible.
that it is a non-profit established to help with Open Document Format
Stop right there. If that is the sole purpose for the organisation to exist, then it makes no sense at all for it to start promoting an alternate format.
The most logical reason for this change of heart I can think of - given that nobody seriously expects "compatability with Microsoft formats" to ever be anything more than a pipedream - is a big bag of cash.
had apple continued with the clone program odds are I would own a clone.
Had Apple continued with the clone program, odds are you would own nothing. At the time, most of their cashflow came from hardware sales and the clones were significantly eating into that.
2. Being indifferent to third-party hardware 3. Actively interfering with attempts to run on third-party hardware
Please excuse my ignorance in these matters, because I genuinely don't know. Is Apple doing #2, or #3? It's plausible that, as people claim, #1 interferes with Apple's desire to guarantee quality. But #2 and #3 should be essentially equivalent in terms of the quality that Apple can deliver for its customers, and hobbyists would be a lot happier with #2.
The problem is twofold.
Firstly: Apple is all about a brand, an experience if you like. It's a bit hard to explain to an IT crowd who are used to being able to mix and match what they like and don't mind too much if something breaks, but the whole point of Apple as a company is "sell elegant stuff which JFW". The "don't care if it breaks, I'll just fix it" customer mentality has never been particularly important to Apple.
If someone's experience of Mac OS is "oh, that's the thing the kid down the road installed on my PC and it never really worked properly", then it's very hard for Apple to get the message across that they sell elegant stuff which JFW.
Secondly: If Mac OS can be made by hobbyists to work well with non-Apple hardware, suddenly Apple finds that every PC OEM on the planet has just become an Apple-cloning company. Something similar almost destroyed Apple some years ago, they're not about to make the same mistake again.
I am given to understand that in flagrant cases like this, there are similar laws in many US states.
However, it does not do you a shred of good with large companies which have a policy of not training their own staff in small matters like the Sale of Goods act, instead telling them "this is the company policy on returns and You Must Not Deviate From It".
Pretty much your only choice then is:
* Call head office and speak to someone who does know this. Practically impossible when the only telephone numbers you can get are to call centres manned with equally uninformed staff.
* Stand there at the checkout queue until they get fed up of you.
* Inform the press/trading standards/take them to court.
Traditionally, supercomputers were only used to deal with very specific problems which you'd probably write your own software for. They had a lot of very specific hardware designed from the ground upwards for such problems. An algorithm which will get real benefit out of such a system may well perform surprisingly poorly on your dual core laptop.
However, the amount of R&D going into x86 and related architecture has meant that the likes of Cray had trouble keeping up - so many of their latest systems are clusters of opterons.
I'm sure your laptop is faster than a supercomputer of 15 years ago, but for the specific task that supercomputer was intended for I don't think the performance gain would be anything like as large as you'd expect.
Not really a lot of point in doing that if you've got a site license for XP - it doesn't really "stick it" to them.
It would also limit their options and mean they would have to be very careful how PCs get used - the site license is an upgrade-only license and you still need an OEM copy to be properly licensed. If you've already got XP OEM licenses, the main benefit it gets you is a CD and license key which bypasses all this activation guff and can be legitimately installed on many systems. And when all new business PCs ship with Vista, you'll still be able to legitimately downgrade them.
I think they mean an SSH server rather than client, and it certainly is something that has historically been missing.
It doesn't surprise me much, because Microsoft have spent the last 10 years trying to drive everyone away from command line use - generally speaking, you either use Remote Desktop or one of the MMC plugins over RPC to admin a Windows server. However, they are now (finally) producing proper scripting mechanisms for things like Exchange - clearly the fact that every Windows shop of any significant size inevitably winds up doing scary things with batch files, AutoIt, KIX32 and the like has finally sunk in.
Microsoft being what they are, I expect at least 2 generations to go by before these scripting tools are worthy of being taken seriously. I would expect an SSH server in either Windows Server 2008 R2 or the next version after that - and very likely available as an extra download for existing 2008 servers at the same time.
Postgres is completely different under the hood, so I don't think that's terribly likely.
Bit of a shame in many ways, because while there are commercial addons to Postgres which give decent clustering support - either for HA or performance - AFAICT the free Postgres server can offer very little in the way of clustering.
You're assuming that time machine works over a shared network folder.
I very much doubt this will be the case. To my mind, Time Machine looks an awful lot like a pretty wrapper around a snapshot function, similar to that found in modern logical volume managers and SAN products. Sun's ZFS has such a function, and Apple have licensed ZFS for inclusion in Leopard.
Such a system generally works at the block level (with LVM), though with the filesystem integration ZFS gives it could probably operate more efficiently. In any case, the only way to get at earlier snapshots is to be able to run an application on the machine itself - and if you can do that, you can do more or less anything.
The whole thrust of the GPs argument stems from reading sites like/. and ZDNet announce "Finally! Linux is ready for the masses" and taking this comment at its face value.
Nobody double-checks that their hardware works with Windows. The very idea that it might not is completely alien. And when Windows users try Linux, having been told that "Finally! Linux is ready for the masses", they don't expect to have to check that their hardware will work.
Linux doesn't have this level of hardware support yet. Sure, basic things like networking, hard disks and that are fine - but times have changed in the last 10 years. For your average desktop user, "basic things" now includes WiFi, suspend/hibernate and that cheap nasty webcam they picked up in Walmart for $9.95.
Right.... so where's the standardised interface for the SNMP userland tools to speak to the hardware to find out details like "is the RAID OK?" "are all power supplies present and functioning?" "are all the fans spinning?" ?
IPMI is a good start, but it's far from perfect. Not all implementations of IPMI even make that much detail available.
Put another way, $15k for a perl script which spiders torrentspy.com, parses all the IP addresses in all the torrent files it can find, does a whois on each of them and returns the results - a few hours work at most - seems like a pretty sweet deal.
Slashdot Mirror
I've worked in a few companies, and in my experience what you say is true... up to a point.
"The point" in question is the point where the OS is of any importance. You may, for instance, have a core database/app server running something supported like Red Hat (or even a commercial Unix), then apache frontends which aren't really doing much running CentOS.
In at least one case I can think of - granted, this was a few years ago - everything was so tied to the database that the database was commercial, paid for and supported. But the OS was treated as a commodity. (Mind you, it probably helped that we had a fairly extensive contract with IBM, all the servers were identical and barring hardware or power failure, it was practically unheard of for anything significant to happen to the running OS).
CentOS essentially acts as advertising for the Enterprise RedHat editions. It allows sysadmins to stick with the same familiar set of tools on both systems where it is considered desirable to have a support contract and systems where this is less of an issue.
RedHat can't do much to curb this anyway - most of what they produce is standing on the shoulders of other GPL software - but if they did, I'd imagine we'd see a commensurate rise in the use of Debian, Ubuntu and (gasp!) SuSE/OpenSuSE.
interesting thinking to claim that murder is somehow forgivable.
Excuse me while I play devil's advocate here - am I to understand that your point of view is that life imprisonment should mean life?
Because if not, then the idea that society should never forgive a murderer to my mind implies that anyone who is let out of prison after serving such a term should never be able to find work, should be denied even the most basic of social housing or benefits and should essentially have no choice but to wind up living as a tramp, wandering the streets, drinking methylated spirits and shouting at people who aren't there.
A lot less than that when you're getting them masted by the million, I'm sure.
For an industry that regularly has to deal with and resolve ambiguities, it's surprising to me how inert it seems to have been on this one.
There wasn't an ambiguity before hard disk manufacturers decided to invent one.
As for the reliability of the information gathered: Is information gathered in this way admissible? It would seem that there are to many potential snags with this that it would never be able to be relied upon by itself.
Does it need to be? I know nothing of German law, but in many western countries this would be the first step of intelligence gathering prior to mounting a raid, not the be all and end all.
The problem likely stems from hiring programmers that aren't familiar with porting software to the Linux platform.
There's a simple reason for that.
They don't have an IT department any more. It was outsourced some time ago - shortly before it became apparent that the Internet was likely to become if not the future of media distribution, then at least very significant.
The development process seems to have been "throw something cheap and cheerful together at the last minute". It had to include DRM because pretty much everything the BBC produces is wrapped up in all sorts of licensing. Heck, even Tuesday's episode of Eastenders used some recent commercial music in the background at one point - that would have to be licensed under specific terms, and it's exactly that kind of thing which prevents the BBC from making all their content free to all.
They're seriously talking about extending the license to anyone who owns a computer.
Yet I have only Apple Macs in my house.
I will gladly go to court and explain yes, I have a computer.
But owing to the BBC's own decisions I am not able to use their service on it.
No - selling Mac clones would still be illegal.
You misunderstand me, but then I could perhaps have worded it more clearly.
In answering the question "do Apple want to let OS X run on third-party hardware?", the fact that Dell wouldn't be shipping OS X preinstalled doesn't really enter into the matter. Apple doesn't make enough money out of OS sales to survive on that alone, so they don't want people taking the cheaper option of "cheapest Dell available plus OS X".
There are two things Apple can do to prevent people doing this:
1. Print on the box "Not to be installed on non-Apple hardware". Dubious how legally effective this would be, would result in them playing whack-a-mole with casual end users which would endear them to no-one.
2. Make an effort to render this difficult/impossible.
that it is a non-profit established to help with Open Document Format
Stop right there. If that is the sole purpose for the organisation to exist, then it makes no sense at all for it to start promoting an alternate format.
The most logical reason for this change of heart I can think of - given that nobody seriously expects "compatability with Microsoft formats" to ever be anything more than a pipedream - is a big bag of cash.
had apple continued with the clone program odds are I would own a clone.
Had Apple continued with the clone program, odds are you would own nothing. At the time, most of their cashflow came from hardware sales and the clones were significantly eating into that.
2. Being indifferent to third-party hardware
3. Actively interfering with attempts to run on third-party hardware
Please excuse my ignorance in these matters, because I genuinely don't know. Is Apple doing #2, or #3? It's plausible that, as people claim, #1 interferes with Apple's desire to guarantee quality. But #2 and #3 should be essentially equivalent in terms of the quality that Apple can deliver for its customers, and hobbyists would be a lot happier with #2.
The problem is twofold.
Firstly: Apple is all about a brand, an experience if you like. It's a bit hard to explain to an IT crowd who are used to being able to mix and match what they like and don't mind too much if something breaks, but the whole point of Apple as a company is "sell elegant stuff which JFW". The "don't care if it breaks, I'll just fix it" customer mentality has never been particularly important to Apple.
If someone's experience of Mac OS is "oh, that's the thing the kid down the road installed on my PC and it never really worked properly", then it's very hard for Apple to get the message across that they sell elegant stuff which JFW.
Secondly: If Mac OS can be made by hobbyists to work well with non-Apple hardware, suddenly Apple finds that every PC OEM on the planet has just become an Apple-cloning company. Something similar almost destroyed Apple some years ago, they're not about to make the same mistake again.
I am given to understand that in flagrant cases like this, there are similar laws in many US states.
However, it does not do you a shred of good with large companies which have a policy of not training their own staff in small matters like the Sale of Goods act, instead telling them "this is the company policy on returns and You Must Not Deviate From It".
Pretty much your only choice then is:
* Call head office and speak to someone who does know this. Practically impossible when the only telephone numbers you can get are to call centres manned with equally uninformed staff.
* Stand there at the checkout queue until they get fed up of you.
* Inform the press/trading standards/take them to court.
Yes and no.
Traditionally, supercomputers were only used to deal with very specific problems which you'd probably write your own software for. They had a lot of very specific hardware designed from the ground upwards for such problems. An algorithm which will get real benefit out of such a system may well perform surprisingly poorly on your dual core laptop.
However, the amount of R&D going into x86 and related architecture has meant that the likes of Cray had trouble keeping up - so many of their latest systems are clusters of opterons.
I'm sure your laptop is faster than a supercomputer of 15 years ago, but for the specific task that supercomputer was intended for I don't think the performance gain would be anything like as large as you'd expect.
For every Linux job advertisement you can show me for Linux, I'll show you 100 for windows.
That cuts both ways though.
For every 100 experienced applicants for a Windows-centric job, you'll probably get 1 experienced applicant for a Linux-centric job.
Not really a lot of point in doing that if you've got a site license for XP - it doesn't really "stick it" to them.
It would also limit their options and mean they would have to be very careful how PCs get used - the site license is an upgrade-only license and you still need an OEM copy to be properly licensed. If you've already got XP OEM licenses, the main benefit it gets you is a CD and license key which bypasses all this activation guff and can be legitimately installed on many systems. And when all new business PCs ship with Vista, you'll still be able to legitimately downgrade them.
I think they mean an SSH server rather than client, and it certainly is something that has historically been missing.
It doesn't surprise me much, because Microsoft have spent the last 10 years trying to drive everyone away from command line use - generally speaking, you either use Remote Desktop or one of the MMC plugins over RPC to admin a Windows server. However, they are now (finally) producing proper scripting mechanisms for things like Exchange - clearly the fact that every Windows shop of any significant size inevitably winds up doing scary things with batch files, AutoIt, KIX32 and the like has finally sunk in.
Microsoft being what they are, I expect at least 2 generations to go by before these scripting tools are worthy of being taken seriously. I would expect an SSH server in either Windows Server 2008 R2 or the next version after that - and very likely available as an extra download for existing 2008 servers at the same time.
Postgres is completely different under the hood, so I don't think that's terribly likely.
Bit of a shame in many ways, because while there are commercial addons to Postgres which give decent clustering support - either for HA or performance - AFAICT the free Postgres server can offer very little in the way of clustering.
You're assuming that time machine works over a shared network folder.
I very much doubt this will be the case. To my mind, Time Machine looks an awful lot like a pretty wrapper around a snapshot function, similar to that found in modern logical volume managers and SAN products. Sun's ZFS has such a function, and Apple have licensed ZFS for inclusion in Leopard.
Such a system generally works at the block level (with LVM), though with the filesystem integration ZFS gives it could probably operate more efficiently. In any case, the only way to get at earlier snapshots is to be able to run an application on the machine itself - and if you can do that, you can do more or less anything.
I then immediately took it on a multi-week vacation.....I was able to conduct business as usual,
;)
Well, it sounds like you had a nice relaxing holiday
What other industry is there that abuses their customers like this?
Prostitutes specialising in S&M?
I doubt it. Myself, I rather like the idea of decorating my house in it.
Once complete, I could completely redecorate my house without having to spend hours moving furniture and spilling paint.
The whole thrust of the GPs argument stems from reading sites like /. and ZDNet announce "Finally! Linux is ready for the masses" and taking this comment at its face value.
Nobody double-checks that their hardware works with Windows. The very idea that it might not is completely alien. And when Windows users try Linux, having been told that "Finally! Linux is ready for the masses", they don't expect to have to check that their hardware will work.
Linux doesn't have this level of hardware support yet. Sure, basic things like networking, hard disks and that are fine - but times have changed in the last 10 years. For your average desktop user, "basic things" now includes WiFi, suspend/hibernate and that cheap nasty webcam they picked up in Walmart for $9.95.
Right.... so where's the standardised interface for the SNMP userland tools to speak to the hardware to find out details like "is the RAID OK?" "are all power supplies present and functioning?" "are all the fans spinning?" ?
IPMI is a good start, but it's far from perfect. Not all implementations of IPMI even make that much detail available.
Put another way, $15k for a perl script which spiders torrentspy.com, parses all the IP addresses in all the torrent files it can find, does a whois on each of them and returns the results - a few hours work at most - seems like a pretty sweet deal.