I don't think Red Hat's financial model relies much on people who used to buy a set of CDs for their home computer, and Oracle is even less interested in that market. The real money is in selling ES contracts to ISPs with hundreds or thousands of machines, or, especially, AS contracts with big companies.
As for RHEL/Fedora, I've been running RHEL on all my machines for the last couple of years, recently tried Fedora Core 5, and I'm no wondering why I wouldn't switch to that for most of my office machines (having one local machine running the same build as my leased webservers is IMO worth the money). I keep my downloaded Fedora CDs in one of my Red Hat 7.0 envelope for old time's sake...
And the reason it will take a lot to make me consider moving to Ubuntu or any other distro is simply that I can't bear the thought of going through the "where have they hidden this config file?" experience another time. If I'd gone with the trends as suggested by/. headlines, I would have moved from Red Hat to Mandrake to Gentoo to Ubuntu in the last four years, learned far more about the gnostic secrets of Linux than I ever want to learn, and been half as productive at my job (application programming) as a result. "Better the devil you know" counts for a lot for many OS users.
Guys, can't you see it, this article is a cunning plot by the Evil Empire to produce 3,000/. posts saying "IE7 is fine"? How devious can you get? Stick to Firefox, and then you'll never get suckered like this!
Why are webmasters who are currently ignoring XHTML going to be any more interested in Yet Another Dialect of HTML? At least XHTML has some advantages (like being a well-defined standard, being well-formed XML, and therefore being able to use it with XML technologies such as XSLT). It's not as if W3C provides the tools used by the non-conforming webmaster or anything. As long as there are significant numbers of sites that use bad HTML, and tools that produce more bad HTML, browsers will continue to parse bad HTML, and the pressure for people who don't care about standards to follow standards will remain slight. Come to think of it, even XSLT provides "un-XMLing" features when in HTML mode. If W3C announced a deal with Microsoft and Adobe to phase out bad HTML "features" from their website creation tools, there might be a chance of changing something...
Surely the bigger loser here is Novell. Oracle is competing with Red Hat for support of RHEL, but then Red Hat never had the monopoly in that market anyway: plenty of people used Red Hat without paying for AS-quality support. What this move does do is make the Red Hat flavour of Linux even more clearly the mainstream enterprise distribution. How well Red Hat will cope with competition from Oracle in offering support for that product remains to be seen, but I'd have thought that selling a non-Red Hat flavour of Linux to an IT department suddenly got a whole lot harder.
Only after Intel surpasses the best offerings from AMD, does Dell finally open up to AMD.
I suspect that a company the size of Dell takes a while to change direction, and, more importantly, that the decision to use AMD processors in some machines was not primarily about who had the most MIPS on the day of the announcement. AMD is now respectable, so Dell can sell AMD, and having both Intel and AMD know that Dell can choose from either processor range can't do any harm when it comes to negotiating prices.
Well, I would have said that making filling in a tax return over a secure connection platform-specific was pretty tricky, before trying to fill in my French tax return. (And with Windows you have to be an administrator for their java app to work properly.) As I said, just providing a certificate as a.exe is a good way to make using that certificate without Windows relatively difficult.
See here for the.exe certificate example (link 2).
This is the government whose websites offer secure certificates as.exe archives, and whose tax website advises the user to ignore any security warnings that come up while using their system. If there's a way to render ODF unreliable, insecure and/or platform-specific, I'd say that the French are the people to do it.
It's part of the same issue insofar as the ISP needs to balance its books. Their solution was effectively to tax the providers, starting with the biggest one, and it sounds like it would have boiled down to an indirect government subsidy. Getting money from the providers means they need less money from the end users. And the consumer organisation is out of touch with the realities of broadband costs, which sounds par for the course.
No, it isn't what they are selling. In the small-print of every ISP I've looked at, they say that the peak throughput is not guaranteed. If you want a 1:1 contention rate, there are ISPs that will sell it to you - at a price. My non-guaranteed ADSL package costs me a few dollars a month. The same ISP will sell me ADSL with a 1:1 contention rate for several hundred dollars a month...
The same goes for hosting packages. A lot of the "fabulous monthly bandwidth for 35 cents" offers work by throttling per-second bandwidth to make exceeding their monthly bandwidth impossible. In practice, this means that no-one will ever get close to using the headline bandwidth figure unless they set out to generate traffic with exactly the right profile, and there will be a bottleneck as soon as there is a peak in requests, even if the total monthly bandwidth is way below the headline figure. You can get fully-burstable bandwidth, and you pay a lot of money for it.
I agree that there's a problem in how broadband is sold, but I really don't know how an ISP would translate, say, a 50:1 contention rate into terms that my grandmother would understand or care about.
Surely giving customers what they pay for is not only reasonable but the only sane way to run a business? The basic problem here is the expectation that domestic broadband should be able to run at peak throughput, 24/7, for an attractive flat-rate price. That just isn't viable: the only reason it works is because most broadband users don't generate much traffic, yet.
The two obvious solutions to this are to bill bandwidth-hogging consumers for the bandwidth they consume, or to increase the flat-rate cost to a point where it covers a much lower contention rate than at present. Both those options are desperately unpopular, so it looks like the ISP tried making the figures add up another way.
And there won't always be a cheaper solution around the corner, not if you care about reliability and actual (rather than published peak) bandwidth. One of my broadband connections is astoundingly cheap 10MBit cable, and I think I've managed to get it close to 10MBits about once in the past year. (This is connecting to my own remote servers that clock ten times that speed when transferring data to other remote data parks, so the bottleneck is definitely nearer the broadband end of the connection.)
I know a lot more about perl and about python, but I've just spent half an hour looking around, and I can't find anything remotely resembling CPAN, which seems like one extremely good reason to stick with Perl. The nearest I can find is a static list of modules on the python.org site that, in total, is about a tenth the length of the CPAN list of XML-related modules alone.
CPAN is another reason why your "10,000 lines of code" thing doesn't work. Larry Wall's "laziness" principle says that you use other people's modules whenever you can, and the result is often a script consisting of a list of includes of well-documented third-party modules followed by a very short piece of code that uses those modules.
If $_ is behaving in an unexpected way, I use variable names. K&R is almost completely useless IMO because it hardly mentions any of the optional extras like, um, input and output. If you want to implement encryption, K&R gives you a totally generic and unbelievably low-level set of tools to do it with. The Camel book points you to CPAN. The difference is best measured in man-months...
The large square panels are presumably slides from his live presentation. They aren't supposed to stand alone, in fact I suspect that the whole presentation would make (as much) sense without any of them.
Personally, I find Wall's prose simply wonderful: I've been known to read entire chapters of the Camel book just for the asides. I think that to judge this presentation you have to imagine the equivalent speech about "the future of typing in C++" or "the evolution of the object model in java" and ask yourself if anyone would be awake by the end of page one. Wall is one of the few people I've come across who stands far enough back to get beyond "Side Effects Good/Bad", "Parentheses Good/Bad", "Strict Typing Good/Bad" and so on.
And at least the man does seem to have a nodding acquaintance with other languages. Try Learning Java for the other approach, where the opening chapters compare Java with other languages, and point to several weaknesses or omissions in perl. Unfortunately, perl does all the things the author thinks it doesn't, and it did some of them before Java had crawled out of the C.
I rewired our flat last summer and put CFL's in almost every socket. All of them were known brands. I've replaced several of them, and, in the lights that get the most use, I've replaced them twice. They last longer than filament bulbs, but they don't last anything like the ten years in TFA.
Most of my bulbs are in enclosed fittings, which I gather from discussion here has some effect (why?), but I suspect the main problem is that they don't like power surges (of which we get quite a lot).
In the last month I've been experimenting with LED lights, which are cute, but sourcing replacements for 100w bulbs with LEDs is a problem: the ones that arrived today feel like they are equivalent to 15w tungstens. LED spots are better.
I ran a non-gaming cybercafe for 3 years, and, around the same time, someone else started a cybercafe that was mainly about gaming. I never got rich, but we lasted longer than the gaming cafe.
One problem is simply per-hour income. For the vision to work, you need people to spend hours per week in front of one of your computers, and the hours they have available will be the same as other people's. Their total spend per month on doing this is limited. So you are going to be lucky to be very get a couple of dollars an hour out of people long term. What hourly salary do you want, and what are the overheads of the building, taxes, electricity, heating, insurance, breakages... If you count your time, I'd say you need a dozen customers all the time you are open, and that means a much larger capacity (because if regulars can't get in when they want to they won't be regular for very long).
One huge overhead is the cost of games licences. The people who looked for games in our cybercafe didn't say "Do you have games?" they said "Do you have version Y of game X?". I'm no expert on the gaming market, but you are going to need several different games, and you are going to have to buy new licences pretty often.
Another is the hardware. Our cybercafe used neolithic PCs and diskless LTSP, which was fine for browsing and WP. Even then, we found that people expected decent screens. If you want to provide a good gaming experience, you are going to need expensive kit and regular hardware upgrades.
And your competitor is what the kids could do themselves with their laptops and $20 of networking kit, or their desktop machines and ADSL. AFAICT, the cybercafe down the road folded when the relatively small client base worked out they could network their own computers and cut out the middle man.
Yes to everything that has been said about decent coffee, decor, customer relations, but I don't think that will pay the bills. I'd say it's a hygiene factor, ie lousy environment = no customers but spend on better environment does not give a linear increase in revenue.
Finally, don't count on mixing serious Internet use with serious gaming unless you have two rooms, because the clientele is different and the atmosphere people want is totally different. We picked up lots of Internet customers from our competitors because of the noise level generated by games and gamers.
We just pay for the basic package which gives us 30 days installation support and up2date. I failed to find the email I mentioned on my first sweep - maybe it was a support ticket. The context was that RHN got into a knot, so I was actually asking them to make 2 subsciptions work with 2 computers, and they suggested swapping the 1 working subscription between machines before they grasped the nature of the problem. I'll have another look for it later. We only have 2 or 3 licences at any one time (excluding our leased servers), so maybe the story is different for large corporate users.
I simply don't wish to spend $800+ per server per year
I pay less than half that per server
No, you can't install it 10 times and only pay support on 1 copy to get updates)
Yes you can: I have an email from RH somewhere advising me to use one RH account to update several machines. I have one account per working machine because life is just too short, but it's perfectly permissible.
Alternatively, has anyone ever actually used RedHat support? *I* wasn't impressed...
I've used it a couple of time, for installation issues, and found it to be as good as any phone support I have used (except that I ended up phoning the States). In each case, the first person I talked to seemed to know what a computer is, which is a step up on, say, the AOL people I spoke to yesterday (it was for a friend, honest).
But that's not why I used Redhat. It's because
It's what our leased servers use
Keeping up to date with patches is a no-brainer
Most third-party rpms work
That last reason is why I ditched SUSE a few years ago, even though I liked their distro. I'm an applications programmer. I don't want to spend my time fiddling with make files, I want to grab the modules I need to do the job and then do the job, and RHEL generally lets me do that. And I'm a lot more Linux-savvy than your average beancounter.
I'd say that the main thing going for Ubantu is that it isn't Gentoo.
Most of your points are actually good reasons for most corporations not to touch ReactOS with a bargepole:
They are restricted to a far smaller number of third party companies who can provide support
They have to install the OS themselves, rather than just unpacking a new PC and switching it on
If there's a problem with, say, software RAID, they can play ping-pong between the hardware vendor and the OS maintainer, rather than phoning Dell and saying "fix it NOW"
They have enough problems already with every department writing their own undocumented Access applications containing half the company's valuable data, so they really want everyone to start forking their own version of the OS, not
Upgrading old hardware is only free if they don't pay their staff - one day's fiddling with downloaded images costs about as much as a new PC with Windows pre-installed
In addition, they have to retrain all their in-house staff, which is tricky given that there are no training courses and no manuals, and they have to retrain every MS Certified Engineer who joins them from that point on.
Whenever they have a problem with any third-party application, Adobe, Lotus or the guy down the road who writes VB in his lunch break will all swear blind that the problem is down to the unsupported OS, not their application, and point them to the EULA of the third-party software before hanging up.
MS stuff is expensive, late and sometimes buggy, but MS is much better at judging what combination of late, expensive and buggy the corporate market will bear than the OSS community. Corporate IT departments will moan about the cost of Vista, think through the implications of changing, and buy Vista. Or, maybe, they'll wait a couple of years and then buy Vista. Or, possibly, they'll go for a radical move to, say, Linux thin clients. But moving to a DIY clone of an obsolete version of Windows just doesn't work, mainly because of support issues.
I bet you said the same thing about linux (or samba, or bsd, or whatever) back in the day hey?:-)
If he did, he has been proved right. None of those OSs has made a serious dent in Microsoft's core market. (Linux is strong in the server market where Windows was initially not present at all.)
When corporations don't like the new Windows release, they tend to hang onto their existing one until their concerns are addressed. So, depending on the issues related to Vista, corporations might decide to stick with XP until 2008 or 2009. The longer they wait, the more the pressure grows to be compatible with Vista, and that's why they are likely to get Vista eventually (as well as the fact it turns up on all their new PCs whether they like it or not). Moving to a system that does some of what XP does when they have XP already really doesn't make a lot of sense, unless they want to get out of the whole MS system (in which case the existence or otherwise of Vista is largely irrelevant, and an XP clone would be an odd way to go).
Real coders? So there is more than one person doing this?
And both amazingly powerful.
I don't doubt that Scheme at least is a wonderful language (although personally I prefer zetalisp). But how many libraries are there for, say, manipulating PDF, driving graphics packages or doing cryptography? Any coder who is reinventing every single wheel from scratch isn't a real coder. And how well does Scheme scale for web apps? How much native support does it have from, say, Apache? Can you preload? Can you compile at all for that matter, and, if so, do you have to do that offline and then fiddle around with binaries on the live server? Can you cluster? Can you serve a non-trivial page fast enough to cope with a busy website? Or are these real coders producing web apps for themselves and three dorm-room friends?
PHP is a lost cause... please google for "Ruby on Rails". You'll find no bullshit coders there.
Ok, I'll bite...
PHP is not a lost cause precisely because of your last sentence - what people love about PHP is that you can do whatever you like, armed with nothing more than a few HTML tags, whether or not you have a clue what the implications are, whether or not you can spell "security hole", whether or not there is any prospect of your code being maintained, even by the original coder, a week after it has been written. The stated aim of version 1 of Personal Home Page was to circumvent the unix user-based security system - the whole language is actually an exploit!
It's the rails of Ruby on Rails that puts off PHP programmers addicted to "I started typing HTML and accidentally created a banking website, but someone hacked it before I could show it to you" (and Ruby is a lovely name for dumbed-down perl without a few gigabytes of CPAN modules).
Cayman Islands is a nice place to go for diving and sun, but not for internet based business.
Surely it depends rather a lot on what sort of Internet-based business we are talking about. Running a spam empire only means sending one relatively short bit of text once - the machines doing the spamming could be anywhere in the world, and, indeed, if I was planning a semi-legal or illegal business, I'd be keen to keep the servers as far away from me (both physically and in terms of hops and audit trails) as possible.
Slashdot Mirror
I don't think Red Hat's financial model relies much on people who used to buy a set of CDs for their home computer, and Oracle is even less interested in that market. The real money is in selling ES contracts to ISPs with hundreds or thousands of machines, or, especially, AS contracts with big companies.
As for RHEL/Fedora, I've been running RHEL on all my machines for the last couple of years, recently tried Fedora Core 5, and I'm no wondering why I wouldn't switch to that for most of my office machines (having one local machine running the same build as my leased webservers is IMO worth the money). I keep my downloaded Fedora CDs in one of my Red Hat 7.0 envelope for old time's sake...
And the reason it will take a lot to make me consider moving to Ubuntu or any other distro is simply that I can't bear the thought of going through the "where have they hidden this config file?" experience another time. If I'd gone with the trends as suggested by /. headlines, I would have moved from Red Hat to Mandrake to Gentoo to Ubuntu in the last four years, learned far more about the gnostic secrets of Linux than I ever want to learn, and been half as productive at my job (application programming) as a result. "Better the devil you know" counts for a lot for many OS users.
Guys, can't you see it, this article is a cunning plot by the Evil Empire to produce 3,000 /. posts saying "IE7 is fine"? How devious can you get? Stick to Firefox, and then you'll never get suckered like this!
Why are webmasters who are currently ignoring XHTML going to be any more interested in Yet Another Dialect of HTML? At least XHTML has some advantages (like being a well-defined standard, being well-formed XML, and therefore being able to use it with XML technologies such as XSLT). It's not as if W3C provides the tools used by the non-conforming webmaster or anything. As long as there are significant numbers of sites that use bad HTML, and tools that produce more bad HTML, browsers will continue to parse bad HTML, and the pressure for people who don't care about standards to follow standards will remain slight. Come to think of it, even XSLT provides "un-XMLing" features when in HTML mode. If W3C announced a deal with Microsoft and Adobe to phase out bad HTML "features" from their website creation tools, there might be a chance of changing something...
Surely the bigger loser here is Novell. Oracle is competing with Red Hat for support of RHEL, but then Red Hat never had the monopoly in that market anyway: plenty of people used Red Hat without paying for AS-quality support. What this move does do is make the Red Hat flavour of Linux even more clearly the mainstream enterprise distribution. How well Red Hat will cope with competition from Oracle in offering support for that product remains to be seen, but I'd have thought that selling a non-Red Hat flavour of Linux to an IT department suddenly got a whole lot harder.
Only after Intel surpasses the best offerings from AMD, does Dell finally open up to AMD.
I suspect that a company the size of Dell takes a while to change direction, and, more importantly, that the decision to use AMD processors in some machines was not primarily about who had the most MIPS on the day of the announcement. AMD is now respectable, so Dell can sell AMD, and having both Intel and AMD know that Dell can choose from either processor range can't do any harm when it comes to negotiating prices.
Well, I would have said that making filling in a tax return over a secure connection platform-specific was pretty tricky, before trying to fill in my French tax return. (And with Windows you have to be an administrator for their java app to work properly.) As I said, just providing a certificate as a .exe is a good way to make using that certificate without Windows relatively difficult.
See here for the .exe certificate example (link 2).
This is the government whose websites offer secure certificates as .exe archives, and whose tax website advises the user to ignore any security warnings that come up while using their system. If there's a way to render ODF unreliable, insecure and/or platform-specific, I'd say that the French are the people to do it.
It's part of the same issue insofar as the ISP needs to balance its books. Their solution was effectively to tax the providers, starting with the biggest one, and it sounds like it would have boiled down to an indirect government subsidy. Getting money from the providers means they need less money from the end users. And the consumer organisation is out of touch with the realities of broadband costs, which sounds par for the course.
No, it isn't what they are selling. In the small-print of every ISP I've looked at, they say that the peak throughput is not guaranteed. If you want a 1:1 contention rate, there are ISPs that will sell it to you - at a price. My non-guaranteed ADSL package costs me a few dollars a month. The same ISP will sell me ADSL with a 1:1 contention rate for several hundred dollars a month...
The same goes for hosting packages. A lot of the "fabulous monthly bandwidth for 35 cents" offers work by throttling per-second bandwidth to make exceeding their monthly bandwidth impossible. In practice, this means that no-one will ever get close to using the headline bandwidth figure unless they set out to generate traffic with exactly the right profile, and there will be a bottleneck as soon as there is a peak in requests, even if the total monthly bandwidth is way below the headline figure. You can get fully-burstable bandwidth, and you pay a lot of money for it.
I agree that there's a problem in how broadband is sold, but I really don't know how an ISP would translate, say, a 50:1 contention rate into terms that my grandmother would understand or care about.
Surely giving customers what they pay for is not only reasonable but the only sane way to run a business? The basic problem here is the expectation that domestic broadband should be able to run at peak throughput, 24/7, for an attractive flat-rate price. That just isn't viable: the only reason it works is because most broadband users don't generate much traffic, yet.
The two obvious solutions to this are to bill bandwidth-hogging consumers for the bandwidth they consume, or to increase the flat-rate cost to a point where it covers a much lower contention rate than at present. Both those options are desperately unpopular, so it looks like the ISP tried making the figures add up another way.
And there won't always be a cheaper solution around the corner, not if you care about reliability and actual (rather than published peak) bandwidth. One of my broadband connections is astoundingly cheap 10MBit cable, and I think I've managed to get it close to 10MBits about once in the past year. (This is connecting to my own remote servers that clock ten times that speed when transferring data to other remote data parks, so the bottleneck is definitely nearer the broadband end of the connection.)
I know a lot more about perl and about python, but I've just spent half an hour looking around, and I can't find anything remotely resembling CPAN, which seems like one extremely good reason to stick with Perl. The nearest I can find is a static list of modules on the python.org site that, in total, is about a tenth the length of the CPAN list of XML-related modules alone.
CPAN is another reason why your "10,000 lines of code" thing doesn't work. Larry Wall's "laziness" principle says that you use other people's modules whenever you can, and the result is often a script consisting of a list of includes of well-documented third-party modules followed by a very short piece of code that uses those modules.
If $_ is behaving in an unexpected way, I use variable names. K&R is almost completely useless IMO because it hardly mentions any of the optional extras like, um, input and output. If you want to implement encryption, K&R gives you a totally generic and unbelievably low-level set of tools to do it with. The Camel book points you to CPAN. The difference is best measured in man-months...
The large square panels are presumably slides from his live presentation. They aren't supposed to stand alone, in fact I suspect that the whole presentation would make (as much) sense without any of them.
Personally, I find Wall's prose simply wonderful: I've been known to read entire chapters of the Camel book just for the asides. I think that to judge this presentation you have to imagine the equivalent speech about "the future of typing in C++" or "the evolution of the object model in java" and ask yourself if anyone would be awake by the end of page one. Wall is one of the few people I've come across who stands far enough back to get beyond "Side Effects Good/Bad", "Parentheses Good/Bad", "Strict Typing Good/Bad" and so on.
And at least the man does seem to have a nodding acquaintance with other languages. Try Learning Java for the other approach, where the opening chapters compare Java with other languages, and point to several weaknesses or omissions in perl. Unfortunately, perl does all the things the author thinks it doesn't, and it did some of them before Java had crawled out of the C.
I rewired our flat last summer and put CFL's in almost every socket. All of them were known brands. I've replaced several of them, and, in the lights that get the most use, I've replaced them twice. They last longer than filament bulbs, but they don't last anything like the ten years in TFA.
Most of my bulbs are in enclosed fittings, which I gather from discussion here has some effect (why?), but I suspect the main problem is that they don't like power surges (of which we get quite a lot).
In the last month I've been experimenting with LED lights, which are cute, but sourcing replacements for 100w bulbs with LEDs is a problem: the ones that arrived today feel like they are equivalent to 15w tungstens. LED spots are better.
is the fact that it's hosted on Windows (which we know from the .NET error page it was showing this morning).
I ran a non-gaming cybercafe for 3 years, and, around the same time, someone else started a cybercafe that was mainly about gaming. I never got rich, but we lasted longer than the gaming cafe.
One problem is simply per-hour income. For the vision to work, you need people to spend hours per week in front of one of your computers, and the hours they have available will be the same as other people's. Their total spend per month on doing this is limited. So you are going to be lucky to be very get a couple of dollars an hour out of people long term. What hourly salary do you want, and what are the overheads of the building, taxes, electricity, heating, insurance, breakages... If you count your time, I'd say you need a dozen customers all the time you are open, and that means a much larger capacity (because if regulars can't get in when they want to they won't be regular for very long).
One huge overhead is the cost of games licences. The people who looked for games in our cybercafe didn't say "Do you have games?" they said "Do you have version Y of game X?". I'm no expert on the gaming market, but you are going to need several different games, and you are going to have to buy new licences pretty often.
Another is the hardware. Our cybercafe used neolithic PCs and diskless LTSP, which was fine for browsing and WP. Even then, we found that people expected decent screens. If you want to provide a good gaming experience, you are going to need expensive kit and regular hardware upgrades.
And your competitor is what the kids could do themselves with their laptops and $20 of networking kit, or their desktop machines and ADSL. AFAICT, the cybercafe down the road folded when the relatively small client base worked out they could network their own computers and cut out the middle man.
Yes to everything that has been said about decent coffee, decor, customer relations, but I don't think that will pay the bills. I'd say it's a hygiene factor, ie lousy environment = no customers but spend on better environment does not give a linear increase in revenue.
Finally, don't count on mixing serious Internet use with serious gaming unless you have two rooms, because the clientele is different and the atmosphere people want is totally different. We picked up lots of Internet customers from our competitors because of the noise level generated by games and gamers.
We just pay for the basic package which gives us 30 days installation support and up2date. I failed to find the email I mentioned on my first sweep - maybe it was a support ticket. The context was that RHN got into a knot, so I was actually asking them to make 2 subsciptions work with 2 computers, and they suggested swapping the 1 working subscription between machines before they grasped the nature of the problem. I'll have another look for it later. We only have 2 or 3 licences at any one time (excluding our leased servers), so maybe the story is different for large corporate users.
I simply don't wish to spend $800+ per server per year
I pay less than half that per server
No, you can't install it 10 times and only pay support on 1 copy to get updates)
Yes you can: I have an email from RH somewhere advising me to use one RH account to update several machines. I have one account per working machine because life is just too short, but it's perfectly permissible.
Alternatively, has anyone ever actually used RedHat support? *I* wasn't impressed...
I've used it a couple of time, for installation issues, and found it to be as good as any phone support I have used (except that I ended up phoning the States). In each case, the first person I talked to seemed to know what a computer is, which is a step up on, say, the AOL people I spoke to yesterday (it was for a friend, honest).
But that's not why I used Redhat. It's because
That last reason is why I ditched SUSE a few years ago, even though I liked their distro. I'm an applications programmer. I don't want to spend my time fiddling with make files, I want to grab the modules I need to do the job and then do the job, and RHEL generally lets me do that. And I'm a lot more Linux-savvy than your average beancounter.
I'd say that the main thing going for Ubantu is that it isn't Gentoo.
Does anyone think that security is an issue? Some of the OSS CMS out there are truly scary in this respect.
Most of your points are actually good reasons for most corporations not to touch ReactOS with a bargepole:
MS stuff is expensive, late and sometimes buggy, but MS is much better at judging what combination of late, expensive and buggy the corporate market will bear than the OSS community. Corporate IT departments will moan about the cost of Vista, think through the implications of changing, and buy Vista. Or, maybe, they'll wait a couple of years and then buy Vista. Or, possibly, they'll go for a radical move to, say, Linux thin clients. But moving to a DIY clone of an obsolete version of Windows just doesn't work, mainly because of support issues.
I bet you said the same thing about linux (or samba, or bsd, or whatever) back in the day hey? :-)
If he did, he has been proved right. None of those OSs has made a serious dent in Microsoft's core market. (Linux is strong in the server market where Windows was initially not present at all.)
When corporations don't like the new Windows release, they tend to hang onto their existing one until their concerns are addressed. So, depending on the issues related to Vista, corporations might decide to stick with XP until 2008 or 2009. The longer they wait, the more the pressure grows to be compatible with Vista, and that's why they are likely to get Vista eventually (as well as the fact it turns up on all their new PCs whether they like it or not). Moving to a system that does some of what XP does when they have XP already really doesn't make a lot of sense, unless they want to get out of the whole MS system (in which case the existence or otherwise of Vista is largely irrelevant, and an XP clone would be an odd way to go).
Real coders use Scheme or XCPE for web apps
Real coders? So there is more than one person doing this?
And both amazingly powerful.
I don't doubt that Scheme at least is a wonderful language (although personally I prefer zetalisp). But how many libraries are there for, say, manipulating PDF, driving graphics packages or doing cryptography? Any coder who is reinventing every single wheel from scratch isn't a real coder. And how well does Scheme scale for web apps? How much native support does it have from, say, Apache? Can you preload? Can you compile at all for that matter, and, if so, do you have to do that offline and then fiddle around with binaries on the live server? Can you cluster? Can you serve a non-trivial page fast enough to cope with a busy website? Or are these real coders producing web apps for themselves and three dorm-room friends?
PHP is a lost cause... please google for "Ruby on Rails". You'll find no bullshit coders there.
Ok, I'll bite...
PHP is not a lost cause precisely because of your last sentence - what people love about PHP is that you can do whatever you like, armed with nothing more than a few HTML tags, whether or not you have a clue what the implications are, whether or not you can spell "security hole", whether or not there is any prospect of your code being maintained, even by the original coder, a week after it has been written. The stated aim of version 1 of Personal Home Page was to circumvent the unix user-based security system - the whole language is actually an exploit!
It's the rails of Ruby on Rails that puts off PHP programmers addicted to "I started typing HTML and accidentally created a banking website, but someone hacked it before I could show it to you" (and Ruby is a lovely name for dumbed-down perl without a few gigabytes of CPAN modules).
Cayman Islands is a nice place to go for diving and sun, but not for internet based business.
Surely it depends rather a lot on what sort of Internet-based business we are talking about. Running a spam empire only means sending one relatively short bit of text once - the machines doing the spamming could be anywhere in the world, and, indeed, if I was planning a semi-legal or illegal business, I'd be keen to keep the servers as far away from me (both physically and in terms of hops and audit trails) as possible.