Linux gaining momentum in South Africa
on
Open Source Africa
·
· Score: 4
This article makes some good points. Certainly very few of us in Africa can afford the kind of hardware and pipes that we see reviewed in U.S. magazines all the time. Just to put some things in perspective:
* South Africa has more telephone lines than the rest of Africa put together yet our national telephone penetration is about 8 per 100 inhabitants. * A 64 kilobit permanent connection to the Net here costs roughly $1000 per month. * The telephone operator is a government-enforced monopoly which is partly owned by SBC and Telekom Malaysia * The exchange rate of the South African Rand to the U.S. $ is about 6:1
What has all this to do with software? Quite a bit actually. The more costs you can get rid of the better, and Linux has shown itself time and again in this market to be cheaper, less expensive to run and completely free from the local piracy issues we read about on a more or less permanent basis here. This is also an extremely price-sensitive market, almost but not quite as sensitive as India. Currently the most IT spend in the country is done by the top 100 companies but that is swinging quickly to the hundreds of thousands of SMEs. When they take the biggest piece of the IT spending pie, then Linux will become an even more popular choice for those businesses. Microsoft in South Africa is a pretty slick operation but they've made their cock-ups. I sat in on a seminar for teachers once when the local MS salesperson threatened them all with death unless they paid full price for NT, Office and some other products. There were quite a few Linux converts that day:) Check out this site for the local Linux umbrella body which IMHO is doing an excellent job of promoting the OS. As an example, South Africa's equivalent of Comdex attracted about 50 000 visitors last year. The local Linux lads managed to sell in the region of 15 000 RedHat and Suse distros during the show - which means nearly 1 in 3 visitors walked away with Linux on CD.
So what if the above were true? What if, for sake of argument, certain portions of the Windows source included code which was recognised by a developer to contain source which is under the GPL? Since under the terms of that license, you may not use GPL-licensed code in a non-free (GNU definition) product, it might cause Microsoft a few headaches.
I can see the headlines now:
*Open Source code Discovered in Windows 98: Linux Developer Alleges License Violation * $3 billion R&D from Microsoft devoted to Ripping Off Linux Source
This may seem unlikely until you consider a tool like Interix, which provides Unix scripting tools on top of NT and 98. It would have been tempting to say the least, for the MS developers to use some existing code and recompile for their platform without checking the license.
With all Microsoft's screaming about piracy and licensing, it would be ironic if it could be shown that they blatantly violated the most popular license on the planet.
P.S. This is pure speculation - I have absolutely no evidence that Microsoft has any GPL'ed code in their sources
Only OEMs that may not jump on the Crusoe bandwagon are the Dells, etc. that have a a HUGE Intel based desktop/server market. Intel may say "if you use those Crusoe chips in your laptops, there maybe a Xeon shortage next month. And how many Xeon systems you have to build next month? Be a shame to lose them" Being put in situations like that make it hard to choose Transmeta.
Correct me if I'm a little slow, but isn't this kind of attitude from Intel exactly what anti-trust laws are supposed to prevent? As far as I understand the Sherman Act, it is perfectly legal to be a monopoly just as long as you don't abuse that monopoly power. Surely if Intel even hinted something like that to a big OEM, anti-trust investigators would be all over them? I thought that one of the key findings of Judge Jackon in the Microsoft case was that it had effectively stomped on competition and innovation through a combination of methods that included threatening big OEMs in much the same way you hypothesise Intel might do.
If Dell or anyone else want to use Crusoe as a hardware platform, Intel's recourse should be limited to mundane stuff like making better, cheaper and faster mobile chips that use less power. IOW - fair competition.
True - but I was answering your specific question about RedHat, KDE, Koffice et al. AOL drilling other connections deliberately (as seems to be the case) is different from a GPL program dying on you and losing your data.
Can someone post the AOL EULA? I'd be interested to see if it includes the phrase "ABSOLUTELY NO WARRANTY" - heh.
Who do I get to sue if I install a newer version of glibc and the RPM breaks all my apps?
No-one. Here is the relevant portion of the GPL which applies to the examples you mentioned.
NO WARRANTY
11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.
Well, perhaps not the optical black hole but this idea from the article: In the meantime, physicists are also pursuing the idea of creating "acoustical black holes" (dumb holes), regions that capture and trap sound waves. could come in very handy indeed.
Need some peace and quiet? Try our new Portable Model NahNahNahICan'tHearYou(tm). Used like a normal Walkman(tm), this portable version of the ABH frees you up from all those irritating sound waves you didn't want to hear! Relax and read the paper while standing next to a jet engine! Experience the complete peace of space without the millions of dollars in training needed to get there! Selectively eliminate the sounds of barking dogs, crying children, nagging spouses, co-worker PC startup WAVs and more!
If you live near an airport, why not ask for a free demonstration of our mid-size Dead Quiet(tm)- suitable for all single-story residences. Eliminate those pesky sound waves from passing jets and imagine you're at the quiet country house you couldn't afford in the first place!
* It is not advisable to operate the NahNahNahICan'tHearYou(tm) Portable Personal Accoustical Black Hole while operating machinery or driving. ABH Technologies Limited does not endorse standing next to jet engines during operation. Example used for illustrative purposes only.
I'm not so sure this is the case. Regions were a particular thorn in my side long before DeCSS came to prominence. I live in a country served by Region 2, and since I've spent flipping great wodges of cash buying CDs and videos from US e-tailers before, it annoyed me that Region 1 DVDs wouldn't work on the players available here.
Came the time to invest in an LG 8080B DVD and I realised that it was firmware region-locked. No problem, DeCSS doesn't care about those, right? Well it seems to. Try as I may, I can't seem to authenticate a disc under Linux using the LivID tools if the disc is the wrong region. Luckily this Web site fixed my region problems for me, but I've tried several times to read "wrong" region discs in my drive with no success under Linux.
I've had a quick squiz at tstdvd.c but it doesn't give much help - the real reason is probably buried in the ioctls patch.
I once typed in a whole listing of Space Invaders from some mag on a school PET - POKE'd straight into memory. I must have got a byte or two wrong because the program crashed. Luckily I had read in a BYTE magazine article that if you were very quick with the off-on switch, sometimes the "little RAM chips would be forgiving enough to keep most of your code."
I tried this and it actually worked! Does anyone with more knowledge of the PET know why this was possible? RAM refresh not as unforgiving in those days? I'd really like to know:)
If you can, find copies of R is for Rocket and S is for Space. Great collections of sci-fi short stories and in manageable chunks. I loved 'em aged 13:)
Per that article, the Chinese government has two overriding needs: to keep their tight control over China and to embrace the Internet for economic gain. IMHO, these goals are mutually exclusive.
This is dead right, but I think it goes further. The cyber-economy (better phrase anyone?) is arguably the fastest-growing economy on the planet. Sure, the relative peanuts spent online now pale in comparison to the trillions spent in meat space, but it won't be too long before the Internet economy dominates world trade. When that transition takes place - be it with a bang or a whimper - any nation state that has controls over its citizens economic activities will be even further behind that those that have not. I say further because China already lags the West in this regard. And economic ping times of 999 aren't going to help it - companies trading in China will struggle just to keep up with news as an example. Minor you might say - until you're a Web-based news service and then it becomes quite key.
China is going to be fascinating to watch if you believe - as I do - the case The Sovereign Individual makes: that cheap processing power + the Internet is subverting the nation state at a rate of knots by giving freedom to individuals that governments can't touch. Examples of this abound in the US right now but (differences of the US Federal Government and the Chinese government aside) I want to see what happens when the Net bumps heads with a totalitarian state...
What does how UNIX works have to do with whether people write viruses to attack it? Have you forgotten that the first viruses/Trojan horses to infect the Internet were hosted by UNIX systems? They attacked UNIX because that's what most of the systems on the Internet ran at that time.
I agree it doesn't have to do with whether people will write viruses, but how they go about such a task.
he just assumed that more malicious coders + more popularity = more viruses. This assumption is supported by the fact that most viruses attack Windows systems today. Why? Because most systems connected to the Internet run Windows.
To put it in a bit of context, he had just had a year where his work mail server fell over because of Melissa, all his documents had been trashed thanks to the zipfiles virus, and another variant of good old WordMacro had rendered his setup useless. I remarked on my immunity to these attacks which is when he came up with the comment that as Linux becomes more popular, so you'll see more viruses for it.
I took some time explaining that Linux was different because of a) availability of source code The availability of source code allows a virus writer to find weaknesses that a virus might exploit...period. You can argue that the availability of source code allows more people to look for security holes, but it does not itself prevent exploitation of security holes. On the contrary, existing security holes become vastly easier to exploit when the virus writer can see exactly how the source code works.
True - although that sounds a little like a variant of security through obscurity. What I was getting at was the mindset difference between downloading a.tar.gz file and compiling it yourself vs. running some binary a friend has sent you in the e-mail. He had this idea that Linux users also spend all day at work swapping silly little.EXE files via e-mail:)
b) permissions Viruses defeat permissions by exploiting weaknesses in system software or application code...that is their function. To cite "permissions" as protection from viruses is like saying that burglaries can't occur because people have locks on their doors.
I think that's a good analogy:) Permissions do prevent Joe User from wiping out the whole system with a rogue app and affecting everyone else.
c) the extreme wariness of the average Linux user of running untrusted binaries. The average Linux user will be someone who just wants to get their job done without having to learn to be a system administrator. By then hopefully, admin and security admin will be a lot easier than it is now.
This reminds me of a conversation I had with a Linux-clueless colleague in the media industry. He expressed the opinion that as Linux gets more popular, so "you'll see more and more viruses for it." I laughed, because he was the same person who told me he loved NT for its stability.
But then I stopped and thought for a second. Given his complete ignorance of how Unix-like operating systems work, he just assumed that more malicious coders + more popularity = more viruses. I took some time explaining that Linux was different because of a) availability of source code b) permissions and c) the extreme wariness of the average Linux user of running untrusted binaries. I said my attitude is that if I can't get the source for it, then I won't run it - and I certainly won't run it as root.
Result: he's now running RedHat as his OS of choice. Yes these stories are funny to any halfway experienced user of Linux. But take some time to explain to a Windows-using friend why they are, and you're well on the way to more effective advocacy.
This article is paying lip-service to the community. Given ZD Net's track record I wouldn't be suprised if it was written specifically to get posted to slashdot (like many other articles like this which seem to make it to the front page). Don't buy it.
Not sure what you're getting at here. If you mean that the article was on ZDNet, it wasn't - it's on InfoWorld which is an IDG publication (and a bitter rival of all things ZD). If you mean that InfoWorld is just imitating ZDNet by posting a pro-Linux column, then I have a rebuttal or two. Apart from the boneheaded Bob Metcalfe column of last year, InfoWorld and IDG have a very good track record on Linux support. Nick Petreley writes the back page column in InfoWorld "Down To The Wire". It's the most consistently pro-Linux column I know, except for perhaps his editorials in LinuxWorld, another IDG publication, of which he is the editor. InfoWorld gave Redhat "Best Supported Product of 1998 Award", something which raised serious eyebrows when it was given. ComputerWorld has also had some excellent reviews of various distros and applications like StarOffice. About the only bump in the road from the average/.er's point of view has been the idiotic column by Metcalfe. But that's not too much out of the ordinary - columnists are expected to be opinionated and stirring is part of their job. The bottom line: IDG has supported Linux since long before it was fashionable to do so, and they've put their money where their mouths are.
You're spot on about the uses of autoconf you mention but it has other advantages as well. It encourages the use of very deep packages without having to remember the (nightmarish) rules about recursive Makefiles. Instead of dumping all my source in one directory, I can organise it in a logical fashion and get the flexibility, dependency stuff checking and modularity that goes with it. That's just too much work to do every time with vanilla make I find. Autoconf also simplifies creating a distribution (make dist), putting in checks into your code (make check), installing and uninstalling correctly no matter where the prefix is, and checking for the correct libraries before compile time. Not that I don't understand make - in fact I think that using autoconf and scanning its output has helped me get a handle on how make does its magic but doing it by hand? Eeek. A standardised programming language would work. I think Perl makefiles are already a step towards that idea. But a new one would take time to catch on. Nearly all GNU stuff comes with configure scripts. I get nervous when I download source and it doesn't have a configure because 90% of the time, something is wrong and it needs a bit of makefile hacking. A front-end for autoconf would be a good idea as proposed by some previous posters. Those who want to learn it before a nice GUI comes out or a replacement is found can check Havoc Pennington's online book which has a chapter on the package.
because some of the codebase belongs to other companies. I got this from James Gosling who I chatted to last week (he was visiting this country). He said to me that Scott McNealy is pro OSS but at the moment his hands are tied up a bit by companies like SCO who are being sticky about the issue since they have some code in there. It made sense to me given the history of commercial *nix source - I think that Sun hasn't gone all the way with their source licensing but couldn't figure out why. Maybe it's because they can't just yet...
After reading comments from you over the years about your own so-called ignorance (eg. stencil buffer defaults, polar co-ordinates in Doom, compiler technology for Quake 1), I would like to know the following: what is your attitude to learning? Is it mainly grab-the-laptop-and-retire-to-a-hotel-to-experimen t or do you actually read books on graphics algorithms? I would be interested because you have always seemed to be a programmer who is aware of what you don't know, but don't care since what you do know is more than good enough...
Since nearly all compilers that produce executable code from source first generate assembly, it's a really good idea to understand assembly in the first place.
The next step (which really applies to C programs but includes some other languages as well) would be to run 'gcc -S yourprog.c' on some C code that contains common constructs such as do {} while, for loops, function calls and library calls. You can then have a look at the output and see how a compiler produces assembly from C source.
Compiler methods differ of course, but many are similar. After enough study, you'll find yourself almost "seeing" the C code that an assembly dump was made from.
This is a very simplistic method but works well if you're prepared to put in the time and effort.
This story is funny given the comments in "Charlie and The Great Glass Elevator" about "Mr Hilton" wanting to put a space hotel up in space to rival the US's official program. IIRC, Lancelot R. Gillgrass, President at the time was not happy:)
It was to me anyway. I read it aged 16 and understood maybe 2% of it (if that). But I can look back and say it sparked my interest in programming, classical music, Escher and mathematics - all of which I retain at age 32. Nowdays each yearly read yields maybe 1% more understanding:)
Slashdot Mirror
Just to put some things in perspective:
* South Africa has more telephone lines than the rest of Africa put together yet our national telephone penetration is about 8 per 100 inhabitants.
* A 64 kilobit permanent connection to the Net here costs roughly $1000 per month.
* The telephone operator is a government-enforced monopoly which is partly owned by SBC and Telekom Malaysia
* The exchange rate of the South African Rand to the U.S. $ is about 6:1
What has all this to do with software? Quite a bit actually. The more costs you can get rid of the better, and Linux has shown itself time and again in this market to be cheaper, less expensive to run and completely free from the local piracy issues we read about on a more or less permanent basis here. This is also an extremely price-sensitive market, almost but not quite as sensitive as India. :)
Currently the most IT spend in the country is done by the top 100 companies but that is swinging quickly to the hundreds of thousands of SMEs. When they take the biggest piece of the IT spending pie, then Linux will become an even more popular choice for those businesses.
Microsoft in South Africa is a pretty slick operation but they've made their cock-ups. I sat in on a seminar for teachers once when the local MS salesperson threatened them all with death unless they paid full price for NT, Office and some other products. There were quite a few Linux converts that day
Check out this site for the local Linux umbrella body which IMHO is doing an excellent job of promoting the OS. As an example, South Africa's equivalent of Comdex attracted about 50 000 visitors last year. The local Linux lads managed to sell in the region of 15 000 RedHat and Suse distros during the show - which means nearly 1 in 3 visitors walked away with Linux on CD.
I can see the headlines now:
*Open Source code Discovered in Windows 98: Linux Developer Alleges License Violation
* $3 billion R&D from Microsoft devoted to Ripping Off Linux Source
This may seem unlikely until you consider a tool like Interix, which provides Unix scripting tools on top of NT and 98. It would have been tempting to say the least, for the MS developers to use some existing code and recompile for their platform without checking the license.
With all Microsoft's screaming about piracy and licensing, it would be ironic if it could be shown that they blatantly violated the most popular license on the planet.
P.S. This is pure speculation - I have absolutely no evidence that Microsoft has any GPL'ed code in their sources
P.P.S. s/license/licence/ where appropriate
Correct me if I'm a little slow, but isn't this kind of attitude from Intel exactly what anti-trust laws are supposed to prevent? As far as I understand the Sherman Act, it is perfectly legal to be a monopoly just as long as you don't abuse that monopoly power. Surely if Intel even hinted something like that to a big OEM, anti-trust investigators would be all over them? I thought that one of the key findings of Judge Jackon in the Microsoft case was that it had effectively stomped on competition and innovation through a combination of methods that included threatening big OEMs in much the same way you hypothesise Intel might do.
If Dell or anyone else want to use Crusoe as a hardware platform, Intel's recourse should be limited to mundane stuff like making better, cheaper and faster mobile chips that use less power. IOW - fair competition.
Can someone post the AOL EULA? I'd be interested to see if it includes the phrase "ABSOLUTELY NO WARRANTY" - heh.
No-one. Here is the relevant portion of the GPL which applies to the examples you mentioned.
NO WARRANTY
11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.
For legalese it's actually pretty clear :)
Need some peace and quiet? Try our new Portable Model NahNahNahICan'tHearYou(tm). Used like a normal Walkman(tm), this portable version of the ABH frees you up from all those irritating sound waves you didn't want to hear! Relax and read the paper while standing next to a jet engine! Experience the complete peace of space without the millions of dollars in training needed to get there! Selectively eliminate the sounds of barking dogs, crying children, nagging spouses, co-worker PC startup WAVs and more!
If you live near an airport, why not ask for a free demonstration of our mid-size Dead Quiet(tm)- suitable for all single-story residences. Eliminate those pesky sound waves from passing jets and imagine you're at the quiet country house you couldn't afford in the first place!
* It is not advisable to operate the NahNahNahICan'tHearYou(tm) Portable Personal Accoustical Black Hole while operating machinery or driving. ABH Technologies Limited does not endorse standing next to jet engines during operation. Example used for illustrative purposes only.
Came the time to invest in an LG 8080B DVD and I realised that it was firmware region-locked. No problem, DeCSS doesn't care about those, right? Well it seems to. Try as I may, I can't seem to authenticate a disc under Linux using the LivID tools if the disc is the wrong region. Luckily this Web site fixed my region problems for me, but I've tried several times to read "wrong" region discs in my drive with no success under Linux.
I've had a quick squiz at tstdvd.c but it doesn't give much help - the real reason is probably buried in the ioctls patch.
I tried this and it actually worked! Does anyone with more knowledge of the PET know why this was possible? RAM refresh not as unforgiving in those days? I'd really like to know :)
This is dead right, but I think it goes further. The cyber-economy (better phrase anyone?) is arguably the fastest-growing economy on the planet. Sure, the relative peanuts spent online now pale in comparison to the trillions spent in meat space, but it won't be too long before the Internet economy dominates world trade. When that transition takes place - be it with a bang or a whimper - any nation state that has controls over its citizens economic activities will be even further behind that those that have not. I say further because China already lags the West in this regard. And economic ping times of 999 aren't going to help it - companies trading in China will struggle just to keep up with news as an example. Minor you might say - until you're a Web-based news service and then it becomes quite key.
China is going to be fascinating to watch if you believe - as I do - the case The Sovereign Individual makes: that cheap processing power + the Internet is subverting the nation state at a rate of knots by giving freedom to individuals that governments can't touch. Examples of this abound in the US right now but (differences of the US Federal Government and the Chinese government aside) I want to see what happens when the Net bumps heads with a totalitarian state...
I agree it doesn't have to do with whether people will write viruses, but how they go about such a task.
he just assumed that more malicious coders + more popularity = more viruses. This assumption is supported by the fact that most viruses attack Windows systems today. Why? Because most systems connected to the Internet run Windows.
To put it in a bit of context, he had just had a year where his work mail server fell over because of Melissa, all his documents had been trashed thanks to the zipfiles virus, and another variant of good old WordMacro had rendered his setup useless. I remarked on my immunity to these attacks which is when he came up with the comment that as Linux becomes more popular, so you'll see more viruses for it.
I took some time explaining that Linux was different because of a) availability of source code The availability of source code allows a virus writer to find weaknesses that a virus might exploit...period. You can argue that the availability of source code allows more people to look for security holes, but it does not itself prevent exploitation of security holes. On the contrary, existing security holes become vastly easier to exploit when the virus writer can see exactly how the source code works.
True - although that sounds a little like a variant of security through obscurity. What I was getting at was the mindset difference between downloading a .tar.gz file and compiling it yourself vs. running some binary a friend has sent you in the e-mail. He had this idea that Linux users also spend all day at work swapping silly little .EXE files via e-mail :)
b) permissions Viruses defeat permissions by exploiting weaknesses in system software or application code...that is their function. To cite "permissions" as protection from viruses is like saying that burglaries can't occur because people have locks on their doors.
I think that's a good analogy :) Permissions do prevent Joe User from wiping out the whole system with a rogue app and affecting everyone else.
c) the extreme wariness of the average Linux user of running untrusted binaries. The average Linux user will be someone who just wants to get their job done without having to learn to be a system administrator. By then hopefully, admin and security admin will be a lot easier than it is now.
But then I stopped and thought for a second. Given his complete ignorance of how Unix-like operating systems work, he just assumed that more malicious coders + more popularity = more viruses. I took some time explaining that Linux was different because of a) availability of source code b) permissions and c) the extreme wariness of the average Linux user of running untrusted binaries. I said my attitude is that if I can't get the source for it, then I won't run it - and I certainly won't run it as root.
Result: he's now running RedHat as his OS of choice. Yes these stories are funny to any halfway experienced user of Linux. But take some time to explain to a Windows-using friend why they are, and you're well on the way to more effective advocacy.
This article is paying lip-service to the community. Given ZD Net's track record I wouldn't be suprised if it was written specifically to get posted to slashdot (like many other articles like
/.er's point of view has been the idiotic column by Metcalfe. But that's not too much out of the ordinary - columnists are expected to be opinionated and stirring is part of their job.
this which seem to make it to the front page). Don't buy it.
Not sure what you're getting at here. If you mean that the article was on ZDNet, it wasn't - it's on InfoWorld which is an IDG publication (and a bitter rival of all things ZD). If you mean that InfoWorld is just imitating ZDNet by posting a pro-Linux column, then I have a rebuttal or two. Apart from the boneheaded Bob Metcalfe column of last year, InfoWorld and IDG have a very good track record on Linux support. Nick Petreley writes the back page column in InfoWorld "Down To The Wire". It's the most consistently pro-Linux column I know, except for perhaps his editorials in LinuxWorld, another IDG publication, of which he is the editor. InfoWorld gave Redhat "Best Supported Product of 1998 Award", something which raised serious eyebrows when it was given.
ComputerWorld has also had some excellent reviews of various distros and applications like StarOffice.
About the only bump in the road from the average
The bottom line: IDG has supported Linux since long before it was fashionable to do so, and they've put their money where their mouths are.
You're spot on about the uses of autoconf you mention but it has other advantages as well. It encourages the use of very deep packages without having to remember the (nightmarish) rules about recursive Makefiles. Instead of dumping all my source in one directory, I can organise it in a logical fashion and get the flexibility, dependency stuff checking and modularity that goes with it. That's just too much work to do every time with vanilla make I find.
Autoconf also simplifies creating a distribution (make dist), putting in checks into your code (make check), installing and uninstalling correctly no matter where the prefix is, and checking for the correct libraries before compile time.
Not that I don't understand make - in fact I think that using autoconf and scanning its output has helped me get a handle on how make does its magic but doing it by hand? Eeek.
A standardised programming language would work. I think Perl makefiles are already a step towards that idea. But a new one would take time to catch on. Nearly all GNU stuff comes with configure scripts. I get nervous when I download source and it doesn't have a configure because 90% of the time, something is wrong and it needs a bit of makefile hacking.
A front-end for autoconf would be a good idea as proposed by some previous posters. Those who want to learn it before a nice GUI comes out or a replacement is found can check Havoc Pennington's online book which has a chapter on the package.
because some of the codebase belongs to other companies. I got this from James Gosling who I chatted to last week (he was visiting this country).
He said to me that Scott McNealy is pro OSS but at the moment his hands are tied up a bit by companies like SCO who are being sticky about the issue since they have some code in there. It made sense to me given the history of commercial *nix source - I think that Sun hasn't gone all the way with their source licensing but couldn't figure out why. Maybe it's because they can't just yet...
After reading comments from you over the years about your own so-called ignorance (eg. stencil buffer defaults, polar co-ordinates in Doom, compiler technology for Quake 1), I would like to know the following: what is your attitude to learning? Is it mainly grab-the-laptop-and-retire-to-a-hotel-to-experimen t or do you actually read books on graphics algorithms?
I would be interested because you have always seemed to be a programmer who is aware of what you don't know, but don't care since what you do know is more than good enough...
Since nearly all compilers that produce executable code from source first generate assembly, it's a really good idea to understand assembly in the first place.
The next step (which really applies to C programs but includes some other languages as well) would be to run 'gcc -S yourprog.c' on some C code that contains common constructs such as do {} while, for loops, function calls and library calls.
You can then have a look at the output and see how a compiler produces assembly from C source.
Compiler methods differ of course, but many are similar. After enough study, you'll find yourself almost "seeing" the C code that an assembly dump was made from.
This is a very simplistic method but works well if you're prepared to put in the time and effort.
This story is funny given the comments in "Charlie and The Great Glass Elevator" about "Mr Hilton" wanting to put a space hotel up in space to rival the US's official program. :)
IIRC, Lancelot R. Gillgrass, President at the time was not happy
It was to me anyway. I read it aged 16 and understood maybe 2% of it (if that). But I can look back and say it sparked my interest in programming, classical music, Escher and mathematics - all of which I retain at age 32. :)
Nowdays each yearly read yields maybe 1% more understanding