Sounds like a very elaborate explanation for why every American is related to a Duke, Marquess, Earl, Viscount, or Baron in Merry Olde England. Now I understand why nobody seems to come from peasant stock...
"It is extremely well suited to numerical and scientific computation."
I wish it wasn't true. It's nasty, horrible, ugly to write and maintain. But it is still true, it's damn fast. I write high performance EM simulators in C++. They're quite fast. On a really good day they'll reach the speed of the equivalent Fortran code. At best. I prefer C++ because I spend at least as much time messing with code as running it. The astrophysics guys here almost all write in Fortran. The protein folders too. If you're trying to simulate the whole universe, every bit of speed makes a difference and Fortran still has the edge. Sigh. Maybe I'll be forced back into writing it some day.
Okay, so why did you switch? Presumably you already had an Apache server up, running and stable? Having already learnt to use one, you must have had a compelling reason to change to the other. What did IIS give you that Apache didn't?
Of the many complaints one might have about Linux, Xorg stability really shouldn't be one of them. I measure my uptime in months, with Xorg running the whole time. I'm not unusual in this. The last time I restarted was when our SysAdmin walked into my office for a chat and lent against the power switch - boy was he embarrassed!
Incompetent Microsoft admins are also what gives many Microsoft products a bad name.
I'm a Linux/Unix user, and no fan of Microsoft, but that's one of the smartest comments I've read around here for a while. Like all those journals that have stopped accepting Latex submissions, choosing to accept only Word files instead. It's not actually Microsoft's fault, but it feels like it is. The problem is really cheap admins who don' know better, and the companies that keep hiring them.
As you say, engineering depts are well placed to partner with industry, but the partnerships are usually quite focussed - the money goes to a particular research project, rather than undergrad projects. Of course, a rich group will replace equipment faster, which does trickle through to students.
Universities are well aware of the cash they can make from patents and spin-outs, and make good use of them. However, the money tends to go into 'university central' rather than the department making the money, and then used to subsidise the less profitable areas. I work in one of the more profitable groups in my university, and while it is frustrating sometimes, I basically agree with this policy. However, I would expect equal tuition fees across departments to be part of the same deal...
Well, like it or not, a humanities degree is cheap compared to engineering or science. All that lab equipment (and space) costs money, not to mention the people who set it up and keep it running. I'm not saying I agree with differential pricing, I'm just pointing out the costs.
I remember when I was first trying to make the switch - I was completely confused by all the distributions. We need more people saying "it doesn't really matter, pick one". I eventually chose Redhat, because I had a couple of friends running it, and figured they could help out if I got stuck.
I've just visited linux.org to see what they say. There is a distributions link on the front page, which is good. But then it gets harder. There is a distribution search facility to get you started. I can choose a language (good), category (what's the difference between mainstream and personal?) and platform (for your average Windows user, what the hell is a platform?). The results are slightly more worrying. Debian comes up first (good so far), Gentoo second (I'm a Gentoo user, but I wouldn't recommend it to noobs), and so on. There are a bunch of links to out-of-date books on Amazon, and information about distros I've never heard of.
If we're serious about having more people use Linux, we're not helping ourselves. Maybe the current situation is okay - Linux is largely used by power users, and we enjoy it's power and manage the rough edges. But if we really want Linux to be available to all, we could make it easier to get started. Ubuntu helps a lot in this respect, but you need to know something about Linux to have even heard of Ubuntu. It's easy to forget, having got used to the Linux world, how confusing it looks from the outside.
I've often wodered about that. I'm guilty of doing the 'fun parts' and terrible at documentation. I'm amazed by the polish some FOSS projects manage to achieve - inkscape springs to mind as a personal favourite, but there are lots of others. Maybe now's a good time to say thank you to all those folks running the forums, writing the docs, managing bug reports, and all the other stuff that makes an application really work. Thanks guys!
Users...do complex things, from presentations, to databases, to collaboration, to complex spreadsheets, etc etc
I can imagine there are complex things some large organisations may want to do, but you haven't listed any of them. Maybe I'm setting the bar too high, but I'd say everything you listed is basic use of an office suite. Since when has banging together a quick powerpoint presentation been challenging?
But isn't this the problem already? I'd feel more sorry for the tech magazines if they weren't so full of long-term-shilling themselves. Most articles are thinly-disguised paid advertisements. Advertisers are leaving the big magazines because we are - I don't bother reading them any more. Or if I do read them, I also want a feedback site to go with it. Slashdot might not be the place to get advice about life insurance, but if a tech writer is spreading FUD or throwing bad statistics around, the lovely folks here will warn me.
Real realtime is quite an odd requirement - not many of us have real use for it, but it puts some pretty heavy requirements on the kernel. Why do you want it?
More thing in userspace, I'll sign up to that. I think you're probably re-inventing the microkernel, but that's no bad thing.
To quote from the bottom of the page: [The mm-tree] can crash your machine, eat your data (unlikely but not impossible) or kidnap your family (just because it has never happened it doesn't mean you're safe)
I notice the patches being tested include Reiser 4...suddenly the above warning appears a bit more sinister.
There is no "fragmentation". Any software that runs on the latest version of RHEL will also run on the latest version of Ubuntu. Or Slackware. etc.
Oh, how I wish that was true. I'm a Linux user and supporter, but let's be honest here. Any open source software that can be built for one distro can be built on another, thanks to the hard work of package builders/maintainers for those distros. The problems come with proprietary code. They come as precompiled binaries, assuming a certain system layout. For example, I use a package called XFDTD. It claims to run under Linux, demands Redhat, and fails under Gentoo, SuSE, CentOS. Sadly, this isn't unusual. Trying to look at things from the other side, a small software house can't afford to maintain packages for four or five Linux distributions. The faster the different distributions converge on a common system layout, the better for all of us.
And before someone says 'screw proprietary, use open source', I'm trying to stick to the real world here. I'd love to use and/or develop an open source package, but really don't have the time. I'm paid to apply rather than write the tools. If proprietary doesn't have a place in the Linux world, I can't use Linux at work.
I'm sure I remember a story about the competition between steam and IC engines for trucks. The IC crowd managed to get a road tax system introduced that depended upon the weight of the vehicle. Since steam engines were far heavier at the time, it pretty much finished off steam trucks in one blow. I can't find a reference though - does anyone know, or did I make this up?
I'm a working academic, and the only point I don't score is unexplained affluence. To be successful in academia you need contacts, collaborations with other (gasp, foreign) groups, and branching out is often a good thing. Showing interest in colleagues work is as likely to be a sign that you've connected some pieces that the rest of us haven't. Maybe it'll be the next-cure-for-cancer, maybe it'll be a daft idea that goes nowhere.
The point is, the listed indicators of espionage are also indicators of a successful researcher doing there job. It's a bit like saying terrorists eat bread, lets investigate everybody who eats bread. It's probably true, it's just not very useful.
I know, I know, I've been reading/. for too long. But here goes. Maybe MS want to attack Redhat with a patent barrage. Divide and conquer, right? Splinter off the smaller businesses (I'd guess Novell would be the next biggest) and then make a really big kill, to put the fear of Balmer in the OS community. I'm not sure where to put IBM in this story, but it's a start.
What the heck are MS up to?
My father is partially sighted. He has enough trouble reading the actual page (try navigating around advertising with a very limited field of view). Captchas just lock him out of the site.
...at our office we created several copies, preserving the original. During the copy process we received several "Security Alerts!" from our antivirus program. Somehow, I'm not getting the feeling this guy's a professional. Or has ever heard of anything like dd.
Slashdot Mirror
Sounds like a very elaborate explanation for why every American is related to a Duke, Marquess, Earl, Viscount, or Baron in Merry Olde England. Now I understand why nobody seems to come from peasant stock...
Astronomers don't usually worry too much about the math, so long as the powers of 10 are good to +/- 10ish, so you're well within bounds.
"It is extremely well suited to numerical and scientific computation."
I wish it wasn't true. It's nasty, horrible, ugly to write and maintain. But it is still true, it's damn fast. I write high performance EM simulators in C++. They're quite fast. On a really good day they'll reach the speed of the equivalent Fortran code. At best. I prefer C++ because I spend at least as much time messing with code as running it. The astrophysics guys here almost all write in Fortran. The protein folders too. If you're trying to simulate the whole universe, every bit of speed makes a difference and Fortran still has the edge. Sigh. Maybe I'll be forced back into writing it some day.
Okay, so why did you switch? Presumably you already had an Apache server up, running and stable? Having already learnt to use one, you must have had a compelling reason to change to the other. What did IIS give you that Apache didn't?
Of the many complaints one might have about Linux, Xorg stability really shouldn't be one of them. I measure my uptime in months, with Xorg running the whole time. I'm not unusual in this. The last time I restarted was when our SysAdmin walked into my office for a chat and lent against the power switch - boy was he embarrassed!
Incompetent Microsoft admins are also what gives many Microsoft products a bad name.
I'm a Linux/Unix user, and no fan of Microsoft, but that's one of the smartest comments I've read around here for a while. Like all those journals that have stopped accepting Latex submissions, choosing to accept only Word files instead. It's not actually Microsoft's fault, but it feels like it is. The problem is really cheap admins who don' know better, and the companies that keep hiring them.
As you say, engineering depts are well placed to partner with industry, but the partnerships are usually quite focussed - the money goes to a particular research project, rather than undergrad projects. Of course, a rich group will replace equipment faster, which does trickle through to students. Universities are well aware of the cash they can make from patents and spin-outs, and make good use of them. However, the money tends to go into 'university central' rather than the department making the money, and then used to subsidise the less profitable areas. I work in one of the more profitable groups in my university, and while it is frustrating sometimes, I basically agree with this policy. However, I would expect equal tuition fees across departments to be part of the same deal...
Well, like it or not, a humanities degree is cheap compared to engineering or science. All that lab equipment (and space) costs money, not to mention the people who set it up and keep it running. I'm not saying I agree with differential pricing, I'm just pointing out the costs.
Don't be silly. We all know the Zune's a mythical beast, and doesn't really exist.
I remember when I was first trying to make the switch - I was completely confused by all the distributions. We need more people saying "it doesn't really matter, pick one". I eventually chose Redhat, because I had a couple of friends running it, and figured they could help out if I got stuck. I've just visited linux.org to see what they say. There is a distributions link on the front page, which is good. But then it gets harder. There is a distribution search facility to get you started. I can choose a language (good), category (what's the difference between mainstream and personal?) and platform (for your average Windows user, what the hell is a platform?). The results are slightly more worrying. Debian comes up first (good so far), Gentoo second (I'm a Gentoo user, but I wouldn't recommend it to noobs), and so on. There are a bunch of links to out-of-date books on Amazon, and information about distros I've never heard of. If we're serious about having more people use Linux, we're not helping ourselves. Maybe the current situation is okay - Linux is largely used by power users, and we enjoy it's power and manage the rough edges. But if we really want Linux to be available to all, we could make it easier to get started. Ubuntu helps a lot in this respect, but you need to know something about Linux to have even heard of Ubuntu. It's easy to forget, having got used to the Linux world, how confusing it looks from the outside.
It'll help the police catch those serial-suicide-bombers that keep getting away...
I've often wodered about that. I'm guilty of doing the 'fun parts' and terrible at documentation. I'm amazed by the polish some FOSS projects manage to achieve - inkscape springs to mind as a personal favourite, but there are lots of others. Maybe now's a good time to say thank you to all those folks running the forums, writing the docs, managing bug reports, and all the other stuff that makes an application really work. Thanks guys!
A quick look at http://about.openoffice.org/index.html reveals
Sun continues to sponsor development on OpenOffice.org and is the primary contributor of code to OpenOffice.org.
So, as MightyMait so eloquently put it, you're saying Sun Microsystems is a bunch of amateurs and/or hobbyists?
Users...do complex things, from presentations, to databases, to collaboration, to complex spreadsheets, etc etc
I can imagine there are complex things some large organisations may want to do, but you haven't listed any of them. Maybe I'm setting the bar too high, but I'd say everything you listed is basic use of an office suite. Since when has banging together a quick powerpoint presentation been challenging?
But isn't this the problem already? I'd feel more sorry for the tech magazines if they weren't so full of long-term-shilling themselves. Most articles are thinly-disguised paid advertisements. Advertisers are leaving the big magazines because we are - I don't bother reading them any more. Or if I do read them, I also want a feedback site to go with it. Slashdot might not be the place to get advice about life insurance, but if a tech writer is spreading FUD or throwing bad statistics around, the lovely folks here will warn me.
Real realtime is quite an odd requirement - not many of us have real use for it, but it puts some pretty heavy requirements on the kernel. Why do you want it?
More thing in userspace, I'll sign up to that. I think you're probably re-inventing the microkernel, but that's no bad thing.
You're not a QNX fan by any chance?
To quote from the bottom of the page: [The mm-tree] can crash your machine, eat your data (unlikely but not impossible) or kidnap your family (just because it has never happened it doesn't mean you're safe)
I notice the patches being tested include Reiser 4...suddenly the above warning appears a bit more sinister.
iPhone support?
There is no "fragmentation". Any software that runs on the latest version of RHEL will also run on the latest version of Ubuntu. Or Slackware. etc.
Oh, how I wish that was true. I'm a Linux user and supporter, but let's be honest here. Any open source software that can be built for one distro can be built on another, thanks to the hard work of package builders/maintainers for those distros. The problems come with proprietary code. They come as precompiled binaries, assuming a certain system layout. For example, I use a package called XFDTD. It claims to run under Linux, demands Redhat, and fails under Gentoo, SuSE, CentOS. Sadly, this isn't unusual. Trying to look at things from the other side, a small software house can't afford to maintain packages for four or five Linux distributions. The faster the different distributions converge on a common system layout, the better for all of us.
And before someone says 'screw proprietary, use open source', I'm trying to stick to the real world here. I'd love to use and/or develop an open source package, but really don't have the time. I'm paid to apply rather than write the tools. If proprietary doesn't have a place in the Linux world, I can't use Linux at work.
Shouldn't that be 'tarred with the same command'?
I'm sure I remember a story about the competition between steam and IC engines for trucks. The IC crowd managed to get a road tax system introduced that depended upon the weight of the vehicle. Since steam engines were far heavier at the time, it pretty much finished off steam trucks in one blow. I can't find a reference though - does anyone know, or did I make this up?
I'm a working academic, and the only point I don't score is unexplained affluence. To be successful in academia you need contacts, collaborations with other (gasp, foreign) groups, and branching out is often a good thing. Showing interest in colleagues work is as likely to be a sign that you've connected some pieces that the rest of us haven't. Maybe it'll be the next-cure-for-cancer, maybe it'll be a daft idea that goes nowhere.
The point is, the listed indicators of espionage are also indicators of a successful researcher doing there job. It's a bit like saying terrorists eat bread, lets investigate everybody who eats bread. It's probably true, it's just not very useful.
I know, I know, I've been reading /. for too long. But here goes. Maybe MS want to attack Redhat with a patent barrage. Divide and conquer, right? Splinter off the smaller businesses (I'd guess Novell would be the next biggest) and then make a really big kill, to put the fear of Balmer in the OS community. I'm not sure where to put IBM in this story, but it's a start.
What the heck are MS up to?
My father is partially sighted. He has enough trouble reading the actual page (try navigating around advertising with a very limited field of view). Captchas just lock him out of the site.
...at our office we created several copies, preserving the original. During the copy process we received several "Security Alerts!" from our antivirus program. Somehow, I'm not getting the feeling this guy's a professional. Or has ever heard of anything like dd.