CentOS is alive. Two days after a core group of developers posted an open letter to primary admin Lance Davis, threatening to fork the open source OS if he didn't discuss his apparent disappearance from the project, Davis has answered their call - and he seems to have quelled their complaints. "The CentOS Development team had a …
The damage is done. They're not going to attract any new users and this debacle will serve as an excellent stick with which to beat IT budget managers in order to move to something less politically unstable.
If I were Davis I'd be distancing myself from this financially now - rather than falling rapidly over the edge to its death like it might've done it's going to slide slowly into the footnotes of distro history. Dignified or what.
Another example of what ails Linux
Another example of how the Linux community eats itself from within. All these half-way always-in-doubt distros undermine Linux adoption and momentum in the marketplace. And now Google Chrome will further fragment the Linux market.
Despite Microsoft's security problems, the whole Vista debacle, the rise of the netbook, Open Office sucking less, Firefox doing great, and a global recession to make people cost-conscious, this constant internal fragmentation of the Linux landscape is what really holds Linux back. And even the old die-hard Linux advocates don't run Linux anymore, they all have Macs -- what message does that send?
If all the other distros were dropped and the Linux community focussed on making the current leader better (namely, Ubuntu), then Linux might actually win some market share.
But...Macs are selling strong, Windows 7 runs fine on netbooks, Google Chrome will shoot Ubuntu in the knee (as the entire IT press cheers "Hurray! Google is not Microsoft!") and the economy will recover -- but how will the Linux community organize itself in response? What is the strategy for the next five years? Maybe another couple dozen forked distros (Moblin-esque) and more in-fighting about Gnome vs. KDE or SysV vs. Upstart or whatever... Pick one distro and drop the others before it's too late.
And the reason for his disappearance was?
Or was this not disclosed?
Your comment doesn't make any sense. CentOs is a server OS, you compare it's use with desktop OSs, How many admins do you think ditched a server version of linux to install OSX server?
RE: Another Example of what ails linux
Why do you think that ONE distro is going to be good for the evolution of FOSS?
One of the great strenghts of FOSS is that there are many people working on the different distros who are looking at the same software from different angles. IMHO this is a strength.
And why Ubuntu? Does Canonical make any money and do they have a viable business model that will start making them money? If they become the only distro then I would expect an awful lot of developers would walk away from their projects.
AFAIK, RedHat (and possibly Novell) are the only Distros that are making money of which, an awful lot is ploughed back into development of Linux. What do you think is going to happen to these companies? Do you really want to force the world to become Brown/Orange. I switched to Linux for choice. IMHO, this is not choice.
I'm using Ubuntu as an example only. I wouldn't want every linux desktop to be Red or Green or blue (RH, SUSE,Mandriva) either.
Now that the CentOS project seems to be back on track, I'll continue to use it on projects where the economice can't justify a RHEL subscription.
And on the upside...
This episode is roughly on a par with a small company growing up and realising it needs a bit of corporate governance. There are two key differences though
- it was done in public. Most "conventional" business crises happen behind close doors so few outsiders ever realise anything was ever amiss.
- the source code was always in the public domain so if the CentOS project had disappeared its users could have made a quiet, orderly transition to an alternative (RedHat or Scientific to mention just 2).
Both of these factors work in the end users favour and are a good thing for Open Source.
Re: Another example of what ails Linux
As far as dropping other distros goes, I wouldn't trade further adoption for a lack of choice in distros. Personally I'm not that big a fan of Ubuntu, and I couldn't give a toss whether it is the lead or not.
I chose to use Linux because of the choice, I run a distro that suits me. There is a laptop in the house running Ubuntu because my girlfriend finds it easier, but the main PC is gentoo, one of the lower powered file servers runs DSL, and the web server runs Gentoo. My netbook runs Debian, and my laptop runs Gentoo.
Each of these was picked for it's suitability for the task in hand, and I wouldn't trade that for anything. Unless things have changed significantly recently, most Linux advocates feel the same.
GNU/Linux is not commercial, mainstream adoption would be great, but it's not really the stated aim. A lot of the code is written by people purely for the fun of it, or to achieve a task that they want to do. As a rule, the capitalist aim of achieving market dominance does not apply. The only real benefit of such adoption to most OSS developers would be a warm fuzzy feeling, we certainly don't make much (if any) money from it.
And as far as fragmentation goes, MS is well on the way down that path with its numerous SKU's. Albeit without the headache of shared libraries being named different things or kept somewhere different.
If the whole world wants to adopt Ubuntu, that's fine, but I won't be changing unless it suits my needs. If the whole world wants to use Windows, as long as it is an informed choice, thats their own choice. Again, unless it suits my needs, I wont be.
Paris, because she more than suits my needs
Linux is not held back.
I don't think people understand what Linux is.
Firstly Linux is a kernel, that is important not from a pedantic viewpoint, but from understanding that hardware drivers are required, and that Linux is not an operating system or a distribution.
A distribution can be commercial or community based, community based doesn't have to return a profit or make market penetration, a commercial distribution does. Ubuntu and RedHat are commercial operating systems with a Linux kernel, and they care but no one else does, and the vast majority of people who use Linux for development don't use those two distributions.
Not many people have jumped ship to Mac, instead what you have seen is Linux people advocate Mac over Windows and Linux distributions, to general users.
Linux was not developed to somehow affect the Windows market, it was developed as a challenge for those who found minix boring.
No one likes to develop end user (numpty) applications for nothing, those you pay for, applications for numpties will never be gratis, what you think of as polish, developers see as wasted development time, and lost CPU cycles.
GNU was designed to take on commercial Unix and it is being doing a pretty good job of that.
Google Chrome is a Linux distribution, as is Splashtop and Asus gateway, that means that Linux will eventually have the largest market penetration, it may actually have the largest penetration on computer systems as whole already.
The fragmentation of the Linux distributions is what makes Linux great, probably not great for numpties, but great for developers and operating system builders, it is brilliant.
Choice means that you have many different viewpoints on how an operating system can be built and distributed, it also means that Linux has many heads, and that offers security, stability and extensive testing. Linux as a kernel ain't going anywhere, there are too many experts who can step in.
And the interoperability between Linux distributions is again excellent, you can quite easily jump between distributions if you know what you are doing.
You hear this bunkum and FUD from people who are just not in the field of Information Technology, they are at best neophytes struggling probably to use a decent editor. And somehow they think they can tell everyone else to follow their over simplified, simplistic, retarded line of consolidation of distributions. Not only is it just plain wrong from a technical viewpoint, too many people will just laugh in your face at an attempt to control their projects.
The Linux kernel is forked to the nines with patches, and bloody good job to, it allows you to select best of breed for the custom deployment of a system.
The main problem is people outside of IT want to make IT simple for numpties, but for those in IT there is no advantage for them to do that, as it is quite simple already, so it will cost.
If you want Linux Market Penetration put your money or your coding skills where your gob is and make a distribution that will get that market penetration, don't ask others for that, most of us are quite happy with what we have and how things are going. I suspect you neither have the money or the skills Jason Fossen, so for you I would advise you get a Mac :)
There is something that ails Linux?
"Pick one distro [I guess Ubuntu, probably running on Mac hardware] and drop the others before it's too late."
Note that since the last bubble burst about nine years ago, Linux has just chugged quietly along like a somewhat rickety steam locomotive with a few hundred polished bells and whistles and too many logos on the cab side. It worked so far, do people really need to change the approach and opt for a centralized Microsoft model with a master helmsman laying out five-year plans?
It wouldn't be possible.
And why should stagnation be a good idea?
A break from the interwebs
I fail to see what the whole hoo-har was about. So what if this guy had a small break from the internets. He works on a voluntary community project. Not like he actually gets anything from it.
This shows the strength of free software
The failure of one link does not break the chain with FOSS. The fact that others were prepared to fork the distro and work around the problem shows the robustness of the model. Linux is now really too big for it to be allowed to fail. I am sure that IBM would not let it nor would Google and nor would lots of others, great and small.
Compare this with the closed model. If your supplier lets you down, where do you turn?
Re: Alan Barnard
>Linux is now really too big for it to be allowed to fail. I am sure that IBM would not let it nor would Google and nor would lots of others, great and small.
Not a great choice of words as too big to fail is a horrible term that points out how corrupt and bought by the highest political contributor the USA government has become. Google and IBM are very nice to have as proponents but Linux won't fail because it is truly owned by all of humanity. Having a small "elite" group of people control the software necessary to operate one of mans greatest invention always was a terrible idea and perhaps one of the few true great contributions the hippy generation made was to fix this (Stallman and others).
Jason, I think you have summed the (Linux) issue up almost perfectly.
Alas, it will fall on both deaf and (mostly) obtuse ears, as the comments about whether it is a server OS or whether Linux is really a Kernal arguments are offered up as excuse fodder just as they have been for years.
But then the real irony is not how much Apple have embarrassed the market leader Windows with OSX, but how much they have embarrassed the Linux community.
Because Apple has proved you can create a fast, efficient, slick as mustard operating system with an interface that (mostly) looks like the dogs doo doos, and run it on an x86 flavour of some archaic 1970s OS called Unix.
Something alas, the Linux community has singularly failed to do in the last 18 years.
@Linux is not held back
You managed to say exactly what I wanted to say, whilst managing to describe Joe Public as numpties without causing too much offense. I can think of some much worse terms for users!
I wouldn't necessarily say that 'polish' was wasted time, it depends on the level you aim for. For example, I've been writing a Database backend (Amazing what you start on a bored weekend with a bottle of wine!) and it does exactly what I want it to. BUT using it for what I wanted was a pain, so I added a little bit of polish so I could act like a numpty and not have to learn all of the protocols.
It's all about getting balance, if creating a nice experience for the user impacts upon the systems capabilities, then there's too much polish. Let the commercials worry about that, or write a nice front end as a second project. So in essence you're right, but we all like a little bit of polish (think autocompletion in BASH for example)
All it would take is to fork the project, and you could continue as normal, just with a different label.
It does seem odd to let one person have so much control though.
@Jason Fossen et ali
"Pick one distro and drop the others before it's too late."
Nah. Not one. Three. Customized Slackware for desktops, either Slack or customized BSD for servers (routers, firewalls, what have you), and SLES on the rare occasion when my clients need a UN*X on big iron (I'd run Slack on z, but Slack S/390 isn't ready for primetime yet, and may never be. I have it running on a couple-three LPARs, though, just to keep up with Mark's progress). Some people require RHEL on big iron (large corporations being what they are ...), but thankfully I don't have any as clients. I don't like RedHat. Don't get me wrong, I don't like SuSE either (I'm a UN*X purist), but big corporations like to have long-term stability guarantees without IBM prices.
Before flaming me, try to remember that I'm talking about work machines, not silly little PCs that people use at home, mostly for playing games. At home, use whatever floats your boat ... as long as it works for you, nobody on the planet can say you are wrong.
One other comment ... Big companies can go away in a hurry. Remember Enron? FOSS is, by its very nature, here to stay. FOSS will ALWAYS be with us, and will probably always be improving. Either you can learn the ins and outs of FOSS, or you can be left behind in technological oblivion when your corporation of choice evaporates. Just food for thought.
There's two (at least) opposing views here, and both are sort of right.
On the one hand, you've got the folks who say "Linux will never get to the level of ubiquity that Windows and OSX has because each additional distro confuses things, and ultimately puts people off. Get behind one distro, and the world will come to you"
On the other, you've got people who say "The strength of Linux is the variety: I can put exactly what I want on each machine and get the best from each bit of hardware, why change that to go for a single distro that makes things work less well?"
The problem with that second viewpoint is that until there is some consolidation within Linux and it's thousands of distributions- corporates, schools and universities (etc) won't migrate away from Windows until they can be sure that the migration costs (for example) won't be wasted by them opting for the "wrong" distro. Or the choice of OS wrecks some other strategic goal. Or the Chief exec finds his own copy of Word at home doesn;t like ODF. Or whatever.
Simple reason for all this is: Senior people lose their jobs when they guess the wrong way on the big decisions. Or to put it another way: if you were tasked with saying exactly when to switch away from Windows, *and* which of an infinite number of distros will "win"( and that's the corporate mentality: that there must always be a winner and that there can only be one right answer), would you be brave? Or would you allow the ship to carry on sailing on it's Redmond-bound course, pocketing your exec salary every month?
Without Linux (whether you think of it as a kernel or an OS, or whatever) getting a foothold in the schools, universities and offices of the world, it's not going to break onto the home desktop and it's userbase will be small. No more money will drop into the distro food chain (support companies, devco's etc.) and linux will stay niche.
This set of circumstances is a shortcoming in the mindset of the institutions and not the fault of the product or the community, but it's there and pretending it's not is a daft thing to do. At least, it is if you want ubiquity. If you don't, then carry on as you were (presumably compiling gentoo from the command line).
@Linux is not held back
Like so many things in life, one size never fits all.
Linux is in an awkward place at the moment, being the only real alternative to Microsoft's domination from top to bottom of the computing world. It is the only single OS that goes from embedded devices, PDAs and phones, desktop, server all the way to mainframes and supercomputer (I know OSX fits in many of these categories, but I have yet to see a supercomputer running it!). And before anybody shouts that it is not a single OS, I would suggest that it is more a single OS than Windows Mobile, Embedded Windows, Server and 7 will ever be.
But because of this, it needs diversity. The requirements of low power for portable devices, prettyness for the desktop and maximum instructions per second for HPC do not fit together in a single distro.
So, for the masses using Intel and AMD PC's, Ubuntu is desperately needed (support, ease of install and use, good HW coverage). For smaller devices Chrome and Android work. For servers, Redhat Enterprise, Debian, CentOS and SuSE. For bleeding edge development, Fedora. For HPC, any of the many custom distros used by IBM or SGI or Cray or NEC. And if you have a preference for another distro not mentioned here, please use it, with my blessing.
Where's the problem? There is no 'tearing apart' of the Linux developer community to support these, and while the publicised events at CentOS ripple the water, they will never really damage the perception of the people who use CentOS. And even if it does, I'm sure that most places using it would prefer to switch to RedHat Enterprise or a Centos fork rather than Windows.
The only problem I can see is making sure that the people behind the distros remain committed and engaged. This is what has happened at CentOS, and even if another path had been followed (CentOS forking), I don't believe any users would have suffered.
Unfortunately, it is not possible for the community supported model to offer all that established commercial OS providers can. We will see some distros fall out of favour (Slackware springs to mind). So we need players like RedHat, Novell, IBM et. al. as well to generate revenue that pays for at least some of the people who contribute to the core development.
All that happens in forums like this is really noise, albeit interesting in parts.
RE: Another Example of what ails linux
> even the old die-hard Linux advocates don't run Linux anymore, they all have Macs
Assuming that you mean "on the desktop", this old die hard hasn't -- I've used OS X in various freelance jobs, but still prefer Linux on the desktop for server admin, software development, and general office-type things.
> comments about whether it is a server OS or whether Linux is really a Kernal arguments are offered up as excuse fodder just as they have been for years
Clearly, Neil Stansbury has no idea of the issues -- it's "kernel", by the way. His statement is akin to saying that arguments about comparing a sportscar to an articulated lorry to a fuel-injected engine are invalid. Don't make me laugh!
@Jason Fossen, Neil Stansbury
I'm glad that at least one of you shows his credentials by make kernel a proper noun and misspelling it :-)
The point someone else made about FOSS is that it is very difficult to leave users of a distro seriously embarrassed when some organisation goes under. More recently we're even seeing the entire build process opened up (Fedora seems to have led the way here).
Let's take a couple of hypothetical, not to mention highly unlikely, examples. First, what would happen if Red Hat disappeared off the face of the earth. Red Hat is still the number one distro supplier (Ubuntu is a growing second and undoubtedly the leader in some areas) and they make quite a lot of money off support. When Red Hat vanishes, the people who provide support for clones suddenly find the market open for new customers, they get more cash, they employ the ex-Red Hat folks and before too many months we find that Red Hat has been replaced with a new organisation. It won't be a Red Hat clone, it'll have different marketing goals, a different vision, etc. It may be that after a little while longer one of the other Red Hat replacements comes along to get the number one position. Sure, there'll be a fair amount of turmoil and shouting until the dust settles, but the vast majority of users (customers) won't have problems. And development will only suffer a slight hiccup: Red Hat is a big contributor to Linux development, but it's not the only one: IBM, Oracle and Novell are just three big names.
Let's look at the second unlikely example. The number one Windows supplier vanishes off the face of the earth. While this might be considered to be a good thing by some people, it's going to have far more serious consequences. There aren't a bunch of Windows clones suppliers who can step in and provide support (there are people who can provide support for Windows, but, in large part, they can't provide bug fixes not having the source code). What would happen now, of course, is that the customers would have to look elsewhere for the stuff they need to work. They'd need to replace Exchange, Office, Active Directory, IIS, oh, and the operating system. Turmoil doesn't even begin to describe what happens and the people on the receiving end would be the users (customers).
Is it any wonder that Linux is as popular as it is? I'm surprised that it's not more popular, but I stand in awe of the Microsoft Marketing Machine (hmmm, "www" upside down).
"corporates, schools and universities (etc) won't migrate away from <current OS> until they can be sure that the migration costs (for example) won't be wasted by them opting for the "wrong" distro."
Seems entirely reasonable, subject to the signifcant correction that these same folks are now realising that an enforced migration from one Windows/Office (say XP) to a later one (say Vista, and now Win7) may also have significant costs, whereas with Linux/FOSS they are not forced to move, they can (if they choose to do so) stay largely where they are and upgrade in their own time not their MS-centric IT vendor's time.
> And even the old die-hard Linux advocates don't run Linux anymore, they all have Mac
Probably running Linux in a VM.
> CentOs is a server OS
What's this running as my desktop then?
>Don't make me laugh
I wasn't attempting to - but you see then I really don't care how I get from London to Birmingham - in a shiny new A8 or a 4 tonner - the motorway only lets me do 70mph and as long as I'm dry and on time - I just don't care. You might enjoy the waves of admiration about how clever/rich/handsome you are - but I have a hot date in Birmingham - so I just don't care.
Neither incidently do I care that I compile a "Linux" kernEl with a GNU toolchain etc etc - it is excuse fodder, and it misses the point and it has done for years.
Oh and not that it really matters, but I DO have the skills, I AM a coder and I HAVE created my own distro from source.
Just too many
The comments here illustrate nicely why Linux kernel plus GNU presentation/application layer, server or "desktop", have such problems getting accepted in the wider world of amateur and professional use. I read complaints about the diversity of MS Windows offerings. Linux based distributions put this in the shade.
This is why Windows, far from my taste as it is, is successful: At least until XP, users can recognise it and work the same way, for the most part, from Windows 95 until today. It may be clunky; it my be broken or even dangerous; but anyone familiar with Windows can use it and its applications without further training at home, at work, on laptops, on servers, on desktops with little thought for the "distribution". And that, dear reader, for a user who just wants to get a job done or a game played or a CV written or a DVD watched, that is the key.
I have used, at home and at different jobs (as admin. and developer) Redhat, SuSE, Ubuntu, Slackware, Knoppix (great for emergencies) and others. I am actually more of a mainstream UNIX engineer (Solaris, AIX, HP UX, BSD, ULTRIX, Primix (almost UNIX) and others) and now, at home, OS/X. Heavens, it sometimes seems as if SysV and BSD (in all its forms) have more in common than any two forms of Linux.
I more or less gave up on Linux: life is just too short to learn so many ways of installing/maintaining or writing software packaging, hunting for that odd Codex, driver or whatever, finding that the latest version breaks all and has to be reinstalled from scratch .... Why should even a home enthusiast think he needs N different versions for his girlfriend/wife/boyfriend/husband's laptop, his own, his desktop, his little server? This is nonsense and no amount of bleating about free choice, competition, lorries versus cars or whatever washes, not if you know UNIX and real computing and have got a life away from the device used mainly to write your CV, read the internet and occasionally test some idea you had.
So why MAC OS/X? simple. I appreciate a nice, graphical interface (though my favourite for work is twm to display my xterms!); but what I really like is having a more or less straight BSD UNIX underneath and fully available to me that actually works like a mainstream UNIX. So I can do my usual Solaris development etc. at work and work on the same things, whether java via Eclipse, Perl, Python, Shell or Awk on my Mac (as when I had to stay home through a back injury). Best of all, I get all this without endless downloads, finding libraries etc. and can still install GNU packages. Also, I get a good, basic back-up/restoral system as standard and the ability to encrypt my home directory by default (no arguments about how good or bad; at least something is there).
Linux is great if you are the sort who thinks life is only worthwhile if you build your own car or bicycle or you are just too tightfisted or hard up for anything else. But the very fact one has to worry about which distribution and how to get the software to play your DVD (and how to install it ...) is an indicator of why, despite the enormous promise, it is unlikely to succeed really until a buyer/downloader can be sure that, whatever he gets, it meets, by default, common standards such as POSIX, X/Open etc.; that the latest, current (as opposed to experiemental) wireless protocols and so on are there and working by default (so I do not have to run my wireless router in a degraded mode for friends with netbooks for instance) and that whatever he did on the last x versions, will still work on the latest. I can still use the same X, twm, hundreds of scripts in various languages etc. on the latest Solaris (and Cygwin, great software) as I used fifteen years ago, as well as some of the excellent, new software..
Why all the fuss ?
All of this doesn't make much sense to me. The business world already has standardized on a distribution. Most places run RedHat because a proper contract makes the PHBs feel good, whether it's useful or not. Other distributions are only used when the admins have their say (doesn't happen that often in non tech places) or in smaller shops.
And even if CentOS vanished tomorrow, DimeOS would take over in a couple weeks. Its users probably wouldn't even have time to be impacted.
And re. the Mac, an awful lot of the people who have tried OS X (myself included) quickly came back to Linux / Unix. That horrible mess with its mangled filesystem, numerous quirks and poor i18n was just as bad as Windows to me. Unless you needed an application that ran in it, few people I know stuck with Apple's system, especially given their tendency to not be very open (and the pain of dealing with their users).
I know my iBook has been a doorstop for two years now, replaced it with a random Samsung PC that works just as well, doesn't always require an external mouse (X11 with 1 button is just too painful) and runs a KDE desktop just fine.
And now I can easily type all of the Latin-9 charset without looking up a table, in a logical way, and I have a lazy mouse focus which does *not* pop windows to the front. All of which are among the minimum essentials I expect from any current machine (Windows doesn't deliver either in that area : you *still* can't reliably type in French in Windows).
MacOS certainly isn't the best thing out there unless you set your requirements pretty low. It's better than Windows XP but that's not saying much.
most of these comments are a kneejerk reaction to a misleading story.
centos as a distro was not in danger, just that one guy held the keys to the domain and donations. the worst case scenario was that a different domain name would have to be used, make an rpm to change the repo url, and continue.
So why isn't Linux dead yet?
For something with a chronic ailment, it seems remarkably resilient.
@AC "Just too many"
I can certainly appreciate many of the things you have said, indeed when I was reading it I wondered whether I had written it in my sleep until I got to the point about OSX.
You are, however, taking the Luddite view that I strenuously try to avoid. Yes, UNIX has been a good operating system (and my bread-and-butter) for the whole quarter century plus of my working life, but that does not mean that it will remain a good operating system forever. Like it or not (and I don't), genetic UNIX is now a dead end. Novell, SCO or whoever owns the AT&T code base now have no interest in reviving UNIXWare, HP/UX and Tru64 are legacy (thanks HP!) and the future of OpenSolaris is questionable, with the sands rapidly running out on Solaris for SPARC. This leaves AIX as the last actively developed AT&T derived UNIX (I'm ignoring the smaller companies, most of which are gone or going anyway).
OpenBSD, by the very nature of the court battle between Berkeley and AT&T that made it AT&T code free can only nominally be called a genetic UNIX (yes, I know about the V7 code base, I was around then), and I do not remember whether OpenBSD, FreeBSD, or NetBSD actually got SVID or XOpen accredited.
So what you now have is a diminishing number of marginally incompatible UNIX systems which adhere to a set-in-stone standard which is becoming increasingly unimportant, and Linux. If you look at where the technological change is coming from, it is certainly not from the UNIX community. Where have the latest X11 and graphics driver changes come from. How about the virtualisation technologies (and, yes, IBM use Linux as an enabler for their hypervisor). Web browsing, Multimedia, printer support, User interface. This work is all happening in Linux space and being backported on occasion to the UNIX base. This includes Perl, Python, Ruby, Apache and any number of other Open Source packages. And often, it is very difficult to compile these on AIX, at least, because of the number of additional libraries needed. This is a much more difficult problem than it would be on *ANY* Linux distro.
The number of people I now work with in UNIX space who EXPECT the GNU variants of the command set by default is now considerable. I keep having to bite my tongue to not remind them that GNU's Not UNIX, and they should not think that they are the same.
I work mainly with AIX, and I am finding that the number of pure AIX people I deal with is minuscule. Everybody who has an interest in computers outside of work is at least dabbling in Linux, if only to give them another career strand if and when AIX falls out of favour with the banks and government agencies.
So by all means immerse yourself in OSX as the closest thing to a genetic UNIX available on the desktop, but please do not regard yourself as a typical UNIX person. You're not any more. (Do you really use TWM as your window manager? I'll admit it's fast, but the word basic does not even start to describe it! If you do, you would probably feel very happy with fvwm on most Linux distros).
BTW. I'm currently playing with V7/x86. Now that is a true genetic UNIX, although not much use for watching DVD's. In case you are interested, it's running inside VirtualBox on Ubuntu 8.04 LTS, which is very suitable as a low maintenance Linux distribution.
Mine coat is the faded corduroy jacket with the leather elbow patches, and has the Lyons annotated UNIX V6 source in the inside pocket. Careful, it's like me, old and a bit fragile.
Kernel vs. Distribution
OK, got my flak jacket on for this one...
So, the difference in the distribution model between Linux the "kernel" and a "distribution" is that Linux is the DOS (the disk operating system, by definition, really) and the GUI and all the "system" apps that run on top of it?
Of course, it isn't politically correct to just come out and say that, because it would seem prehistoric to have a command line operating system and then run a GUI OS on top of it. I mean, that is like, so 80's, and this is the OS of the future, because we based this on an OS from the 70's. Wait a minute...
I actually think we _need_ to promote this distinction: The kernel _is_ the OS. Everything else is the interpreter that parses commands (whether they be from a GUI or command line), or is an app that performs a collection of related tasks. But, we probably won't be able to advocate this distinction to "the public" until someone (else) gets off their duff and writes KDE for Windows, eh?
"obtuse ears," "slick as mustard," "looks like the dogs doo doos."
Come clean, you're really Kevin Keegan giving it 101% there, aren't you?
'Brink of death' is nonsense
The 'brink of death' in the title is nonsense. They had one key contributor gone AWOL. The rest of the team was and is willing and capable of doing the job. They could have relatively easily changed servers and domain names if it had come to that. The donations were already put on hold. CentOS was not close to 'brink of death' by any means.
To the people who are saying that this demonstrates why various distros are unreliable: CentOS is a recompilation of RHEL, which is backed by the largest open source company in the world. Red Hat employs thousand of people and the largest number of Linux developers of any company. A CentOS installation can be kept alive by recompiling RHEL source packages if necessary, and RHEL is likely the most reliable distro there is as far as continued support is concerned. There are flaky distros out there, but CentOS is not one of them.
As the case of Lance whatshisname demonstrates, the important thing about the diversity of distros is not that every geek can find his favorite for his favorite gadget, but that if the leadership of one distro goes nuts, it does not affect Linux as a whole much. Think Mark Shuttleworth suddenly deciding tht he'd rather fly to Mars that develop software. Even if Red Hat fumbled things in a big way, there are other who would be able pick up their customers sooner or later. In contrast, if Apple had not got Jobs back, the whole company might have disappeared by now.
Bsd, Linux, OSX
I recently moved my websites over to a 64bit centos image running on an 8 AMD quad core CPU after 4 years serving off BSD. Seems pretty solid so far.
Yum is much more convenient than the ports system albeit I used to compile most stuff from source on BSD. The size / memory footprint of stuff tends to be bigger with the binaries (eg apache prefork using 20mb ram vs 9mb compiled on BSD) but the updating will be easier.
Full virtualization of my VPS slice is also nice as I get top rather than making do with ps and awk in a bsd jail. I find Redhat / centos /init.d is easier than BSD too.
Jeez, typing In the comment box here sucks on iPhone, I can't scroll...!
OSX / Iphone are great for stuff just working. WPA-2 encrypted wifi took me several days to get working in Ubuntu and Wpa_supplicant / dhclient still don't work properly. iPhone connected up first time. Mac hardwares too pricey though for me - 2k for a MacBook or £300 for a Dell inspiron plus
ubuntu - no contest and besides I hate that stupid single button mouse and apples closed mentality. I ran an OSX server once for a h264 video encode/decode cluster but as someone else said - apples wierd twist on things (odd filesystem etc) made me crave for real BSD...
Good luck to Centos project and thanks for your efforts in making a solid reliable distro available to all.
Occasionally on BSD shared memory corruption (following out of diskspace postgres crash) would necessitate a hard reboot. Haven't seen this yet on Centos.
The other distro I considered before opting for centos was debian 64bit
@I didn't do IT.
"So, the difference in the distribution model between Linux the "kernel" and a "distribution" is that Linux is the DOS (the disk operating system, by definition, really) and the GUI and all the "system" apps that run on top of it?"
Kinda. The Linux kernel provides similar services (in a multiuser, multitasking sort of way) to the old single user, single tasking PC/MS-DOS+drivers. A "Linux distribution" or "BSD distribution" includes a kernel and whatever added software the distro authors deem useful for whatever task they are trying to accomplish. Apple's OSX is a BSD distribution, to all intents and purposes.
"Of course, it isn't politically correct to just come out and say that, because it would seem prehistoric to have a command line operating system and then run a GUI OS on top of it. I mean, that is like, so 80's, and this is the OS of the future, because we based this on an OS from the 70's. Wait a minute..."
Political correctness has nothing to do with it. It's how it works. Even Apple's OSX is, at it's core, a command-line driven system with a pretty face. And OSX is, like, totally based on an OS that is SO '70s, which in turn were based on various OSes that like totally originated in the '60s (some would say '50s, ohmygawd, that's like totally ANCIENT! Gag me!).
"I actually think we _need_ to promote this distinction: The kernel _is_ the OS. Everything else is the interpreter that parses commands (whether they be from a GUI or command line), or is an app that performs a collection of related tasks."
Most techies already know this. Most techies also know that most of those fancy GUI pointy-clicky "system control" widgets are largish chunks of code that manipulate text files (except for Microsoft's 'orrible, 'orrible Registry). Those of us with Clue manipulate the text-based configuration files directly, it's easier. Those of us with MoreClue acknowledge that TheGreatUnwashed seem to need pointy-clicky bits & bobs for configuration. Those of us with LotsOfClues (and good marketing) come up with Apple's OSX ... And those of us who are fed up with hand-holding the local userbase pay attention to how they are using General Purpose Computers and provide a desk-top solution that doesn't require much by the way of support.
"But, we probably won't be able to advocate this distinction to "the public" until someone (else) gets off their duff and writes KDE for Windows, eh?"
 That bit of Val-speak gave me a headache. I'm going to bed.
Sooo someone got on world of warcraft and found him? Bout time...
- Updated Microsoft Azure goes TITSUP (Total Inability To Support Usual Performance)
- The Return of BSOD: Does ANYONE trust Microsoft patches?
- Review Apple takes blade to 13-inch MacBook Pro with Retina display
- Munich considers dumping Linux for ... GULP ... Windows!
- Pic iPhone 6 flip tip slips in Aussie's clip: Apple's 'reversible USB' leaks