... by writing in an assembly language for a mythical computer that I had invented. My father was learning about computers from books and I also read all about them because I was fascinated and knew thats what I wanted to do as a career.
In 1968, working for the British Coal's R&D department, my father bought a Honeywell 516 minicomputer to prove that you could use them to control coal mines.
It had 8k words (16bits) of memory and the only peripherals were paper tape reader and punch, a teletype and some sort of simple parallel IO device (my father was building a massive bunch of electronics - with mag tape loops to represent conveyor belts with coal on - to simulate the coal nine).
I taught myself its assembly language, and started to write a compiler for it (simple Fortran like language that I had also made up).
My school had sent some of us to the local college to learn Algol, and then I had a gap year (sort of - had taken A levels a year early and had some spare time) before University, where I went to the National Physical Laboratory to program for money (on a KDF 9).
Computer Science was not a recognised University Degree back then so I did Electrical Engineering, but joined a British Software Company (Logica) where I did lots of embedded minicomputer projects in those early days before the PC. I still work for the company - but no longer in a technical role - so I get my kicks with open source at home.
The thread started with an AC saying all was FOSS before Gates and he introduced "commercial software". I pointed out that I had paid for software (ie it WAS commercial) well before that.
I admit I had used an assumption that paying for software made it non FOSS, which you correctly pulled me up on. I was then merely trying to say that in actual fact much of this was definitely NOT FOSS, not only had I paid for it, but I did not have access to the source.
The other thing that is important to realise - although, for instance, RSX11-M gave you the source, you certainly could not copy it from machine to machine with out paying for additional licences. Digital had prices for the licence and separately for the media and documentation.
I think that would also make this in the "not Open Source" category.
I know some source was around, I used to use RSX-11M which did a compilation from source as part of the system generation, and used it to debug a memory leak as late as 1985.
But all (apart from possibly the unix packages - I can't remember) that I mention did NOT have source.
Your horizons don't extend far enough back in time. It wasn't Gates who started commercial software.
I did a number of projects on PDP 11s in the late 1970's and early 1980s and often had to buy software for it. In one instance I remember paying £3000 for a Pascal compiler. In another, I paid a similar amounts for both a Unix System V licence and for Emacs.
By the beginning of the 1980's independent database vendors such as Oracle were selling software (our company competed for a while with similar product, but the company pulled out when the business plan showed we would have to sell a 1000 copies to recover the investment - and we decided to concentrate on bespoke application system development)
It looks to me like the actually carefully analysed the code and came up with the 985 errors. What about the ones they missed?
It seems they decided on the Windows XP as the worse end of a generic commercial range.
The only way I know of getting a good handle on how many bugs exist in a body of code is to plot the occurrence of bugs during development and into production against time.
You also plot the fix progress, and try and find out how fixing bugs generates new ones (because you see an increase in new bugs after a release that fixed old ones).
You can then assume that where the number of bugs that exist in code is headed is to a point where statistically fixing one bug creates one.
I worked a lot in software development in the 1970's and 1980's, First as a programmer, but by the later 70's as a project manager and then later as a line manager for a unit built customer systems to a deadline based on a product and its enhancements. (although I am in the same company still, I now am doing business development in another area)
One of the most important lessons that I learnt during that period is that although you can raise productivity with overtime when you are reaching a deadline, trying to do so over a long period is just counterproductive and normally makes you miss deadlines AND budget and deliver crap code. In one mess I had to sort out, the previous project manager had driven the team so hard that they reported the project ready for client acceptence when it wasn't. This was an international project and the client came one quarter way round the world to witness acceptence tests only for them to fail dismally. It took us nearly a year to fix that mess. But proper re-planning (remembering every extra man day on this project was off the bottom line - since it was a fixed price contract) allowed us to beat the revised schedule by a month.
In the late '80s early '90s I was the "chief engineer" of our particular subsidiary and I spent some time examining the software development process in depth to understand how to get better at it. I looked at this overtime issue and why productivity goes down.
There are several reasons for this
1) With an alert mind, programmers (or anyone doing a serious mental task) have to move into a concentration mode. If not tired, it takes about 15 minutes with out interruption to achieve. Firstly when you are tired it takes longer to get into concentration mode, and secondly as the time pressures grow the supervisor level spend more time interrupting you to get "progress reports"
2) when people rush around without taking time to stop and think its a lot easier to go in the wrong direction. To solve bugs, or just to plan what to code need the ability to release oneself from the pressure. [I vividly remember from my early days working in the middle of London and slaving all day over a bug or a design problem. Many a time, I remember realising what the problem was within a minute or two of getting on the train home]
3) A large part of a project is about communication and a common understand and vision of how everything fits together. Its the leader that is pulling the team forward and creating the environment for good communication that makes things happen more right the first time (nobody gets it completely right the first time do they). When the "workers" in the team get pressured into long hours then resentment sets in and that in itself kills productivity.
4) People use their lunch hours and the time after work to do essential domestic chores - such as visiting the bank or posting letters etc. If they skip lunch they just do them anyway in other times.
I would very rarely allow my project managers to authorise overtime - normally it was only when a deadline was close. In the later days I also used to teach on our internal project management courses and this point was always somewhere on the syllabus.
If you follow the links in the article, you will get to the original article from the guy who has written this software. Tutorial D is just one implementation of the 'D@ specification. The reason it is called Tutorial is that it is not an enterprise strength version, but rather one for learning and experimental purposes.
I have continually looked around at alternative e-mail clients to Kmail. Apart from Outlook, I have yet to find another mail client that has a key piece of functionality - the ability to clear out old messages from a mail folder automatically.
I read a lot of mailing lists - some such as Debian-User with several hundred messages a day. I filter each mailing list into its own folder, and then set purge dates on the folder to delete messages.
I tried evolution, thunderbird, balsa and a few others - none of them have this function. Why doesn't this lack of ability to clear unwanted mailing list messages worry anyone else?
Unfortunately its not that simple. My wife uses a machine that is adequately powered running Win98. I tried to upgrade to WinXP, only to find that all her favorite programs (old DOS games) would no longer run. I had to quickly downgrade her.
Your argument misses an important element - The Market
If I am a programmer whose business is producing software, then I have to be able to sell my software at a price that is comparable in the market place. With a "GPL" economy around me, I don't necessarily have the luxury of being able to be in the software business just selling what I have produced. I have to find another model - one of which is to add value to existing GPL code, and then sell services around it, using the value that you have added as a marketing tool for your services.
Just to emphasise this a bit more, to show that there is more to the software business than selling products, I work for a "Systems Integration" company with a turnover of about I billion pounds (UK pounds) - where most of our business is buying 3 party software in, and then adding services (such as writing interfaces, configuring it etc etc). At the moment, much of that is with proprietary licenced software, but increasingly it supports our competitive bids to be able to offer GPL'ed software without the additional (to our services) licence costs. With a company branding which provides the world with an image of technical excellence, most of our sales activities are centred around our knowledge of their business using references from previous projects to show capability, and getting a price that is lower than our competitors. It does not matter to us that we have to provide the source of any modifications or enhancements we make.
The other thing you miss, is the "you scratch my back, I'll scratch yours" element of the GPL. That is, together with others - jointly you produce software that is better than either one of you could singularly - and with the GPL protections no-one else can ride on your back without scratching it. Looking at this from a capitalist point of view, this jointly helping one another is making the business model I described above more competitive than the old one of proprietary companies producing closed source software.
Or get Open Source ones from http://dri.sourceforge.net
Debian includes these packages automatically - which is what I use - in otherwords binary drivers which an installation does exist. Just depends on which one.
There is an issue which took me quite a while to get working - an am not sure if the lastest hardware discovery software packages now solve the problem, but you have to load
agpgart, the agp modules for your motherboard (mine is sis-agp) and the hardware driver (mine is radeon for the radeon 9200 series chips) ahead of X starting. I put lines doing this in/etc/modules.
I don't understand why you slam the previous poster. He said it was a rhetorical question - and it was to illustrate that even a monopoly cannot just increase the price beyond where it matches the value people get from a product. Even with a monopoly there is a price ceiling.
Although Linux may not be for everyone at the relative price difference between Windows and Linux, as the price difference increases more individuals will take the plunge. As that happens more companies will be incentivised to install it pre-configured, to develop device drivers for their hardware and do all the things for linux that currently make people find windows not to challenging techncially. This will lead to more people switching to Linux, driving the incentive for more companies to invest in the services market around it.
I would say that we are close to a point where this might happen anyway (viz the statements made by IBM, Novell and Sun all indicate they think this point is close). A price increase in Windows will just speed the process up.
Of course, people may just stick with the older versions of software for now on their current hardware. But even those will eventually need to get new hardware and will be forced into an upgrade. It will be interesting to see at what point hardware vendors include (and eventually default to) a linux option.
Simple really - receives mail and decides what to do with it. Could deliver it locally, or forward it on to another MTA.
Of course its a bit more complex than that. These days you probably want to do some validation of the mail that you are receiving (checking that the sender is actually who he says he is, filtering out spam and/or viruses), and you also need to do a level of optimisation when mail is addressed to several people. The other main function is to rewrite addresses (for instance, inside an organisation you may be fred@mymachine.company.com - to the outside world you may be fred.bloggs@company.com)
Except that the printers are designed with part of the electronics in the ink cartridge, making it expensive and difficult to copy. This is a game played by the printer manufacturers to control this market.
The problem with that argument is that we are talking about very small market shares anyway, and so these assumptions don't necessarily hold.
My household has 6 computers (including daughters at University Laptops and my work owned laptop) where two of the machines run linux. However 5 windows licences were purchased and zero linux licences (debian - all completely downloaded).
Whilst linux desktops in corporate life are still in the "pilot" stage, I can also see that its going to be much more by a single copy of a distribution and in house distribution on existing hardware.
So whilst the linux market share is small, these sort of situations where the figures are completely screwed will be in the majority.
Of course once a market share does start to become significant, then the "install it yourself" installations DO become insignificant and then the logic holds true.
You will find even more to like when 3.2 is out on Monday. I am running a pre-release here, and it is faster, Konqueror seems less dodgy on rendering sites, and there are a range of new facilities,
I have tried unsuccessfully to find a better mail program than kmail
Things that it does that I haven't found another mail client to do for me
1) allow me to purge old mail messages automatically from a folder after they get to a certain age (essential if you are reading high volume mailing lists)
2) Single key (space bar) navigation down each message, unread messages in folder, over all folders (except trash)
Those two simple things have kept me back from switching from KDE to Gnome.
Not having done it from knoppix, I may well be wrong about this, but...
Can't you just run "debootstrap stable targetdir" (where targetdir is the mounted root partition that you want to eventually use for your on disk setup)
mount proc on targetdir/proc
then chroot into it edit/etc/apt/sources.list and use aptitude to set precisely the apps that you want.
Ensure/etc/lilo.conf is how you want it for the final system and then run lilo. Voila a ready to boot stable environment.
I've done almost that in to build a chroot environment inside my existing one when I was trying to test something.
Thats how I feel about it too. I have been running a server, initially on stable, but then about a year ago upgraded (using aptitude after changing my sources.list to testing) on the fly and it just never goes wrong. The only slight hickup was when _I_ decided to switch from exim to exim4 and it took me a few days of planning to make sure I understood the new configuration file before making the move.
This server is also running a web site, and I've managed to switch between hand crafted web pages, PHPNUKE, PostNUKE and Wiki and back all withough problems too.
I run unstable on the desktop - occassional glitches, but nothing that stops me finding a way round and continuing to work.
I also disagree with his proposal that we should shun proprietary software for the sake of encouraging the development of free software
I argue on my my website that I could shun proprietary software in order to encourage more people to produce free software. In other words I am using my customer "buying" pressure to favour the supplier who is providing me the best offer to improving that offer.
Totally for my own benefit, but economically providing the same effect
Remember DMCA is only an American law. And although I don't believe it prohibits reverse engineering for detecting copyright enfringement, even if it did, the community could do it outside the US.
We run Windows 98 SE on our family machine at home where my wife plays lots of old DOS games in full screen mode. I think there is a dynamic resolution change from 1024*768 to 640*480 as she starts the game up.
I bought Windows XP (Home Edition) to upgrade her. However, all her DOS games started running like a pig. I concluded that something in the XP driver was trying to do some form of autoscaling on the fly. I had to uninstall and go back to Win98SE.
So either there is a massive functionality loss - from my point of view - or, there is some setting I have yet to find which allows these DOS games to get direct access to the hardware.
Anyone out there got the same problem, or know how to solve it?
Although this blackout occured in the US, it has not been the only one. Italy, Scandinavia and the UK have also suffered similar problems.
This has been sending political shockwaves all over the place as the realisation that the pressure on costs on the network owners has meant that contigency margins have been reduced. In the UK (where I live) the public face is one of calm, but behind the scenes there is considerable concern.
What it shouldn't mean though, is a step back to the old days when networks were considerably over engineered. Care should be taken that those who advocate this don't put us back 15 to 20 years and hike the price of energy as a result.
What I think is needed is much more sophisticated control systems that recognise the domino effect built in as a safety measure and actually proactively shutdown parts of the networks in much the same way as firemen make firebreaks to stop a rampant fire. This would keep control of a situation that currently just runs out of hand, and would prevent both the widespread loss of power, but also a much faster return to normality.
Groklaw spends a lot of time analyzing every SCO press release down to the tiniest detail, as if they really mattered. It's also populated by a group that makes Slashdot posters look like disinterested observers.
Really, the basic interpretation of IBM's UNIX contract is unclear, which means that SCO still has a real shot at this.
Groklaw are not only looking at the press releases, much more importantly they are looking through and analysing the legal documents
Now of course SCO may have a chance but the more you analyse the legal documents the more unlikely that seems
Now you may be trolling, but I can't see how either side of the argument can be won by only making assertions. You have to follow through with evidence. As I see it the pro SCO camp haven't produced evidence, the anti SCO camp have produced arguments in abundance.
I know where I will be putting my money, and it won't be in SCOX stock.
Slashdot Mirror
... by writing in an assembly language for a mythical computer that I had invented. My father was learning about computers from books and I also read all about them because I was fascinated and knew thats what I wanted to do as a career.
In 1968, working for the British Coal's R&D department, my father bought a Honeywell 516 minicomputer to prove that you could use them to control coal mines.
It had 8k words (16bits) of memory and the only peripherals were paper tape reader and punch, a teletype and some sort of simple parallel IO device (my father was building a massive bunch of electronics - with mag tape loops to represent conveyor belts with coal on - to simulate the coal nine).
I taught myself its assembly language, and started to write a compiler for it (simple Fortran like language that I had also made up).
My school had sent some of us to the local college to learn Algol, and then I had a gap year (sort of - had taken A levels a year early and had some spare time) before University, where I went to the National Physical Laboratory to program for money (on a KDF 9).
Computer Science was not a recognised University Degree back then so I did Electrical Engineering, but joined a British Software Company (Logica) where I did lots of embedded minicomputer projects in those early days before the PC. I still work for the company - but no longer in a technical role - so I get my kicks with open source at home.
No point in having a pissing match over this.
The thread started with an AC saying all was FOSS before Gates and he introduced "commercial software". I pointed out that I had paid for software (ie it WAS commercial) well before that.
I admit I had used an assumption that paying for software made it non FOSS, which you correctly pulled me up on. I was then merely trying to say that in actual fact much of this was definitely NOT FOSS, not only had I paid for it, but I did not have access to the source.
The other thing that is important to realise - although, for instance, RSX11-M gave you the source, you certainly could not copy it from machine to machine with out paying for additional licences. Digital had prices for the licence and separately for the media and documentation.
I think that would also make this in the "not Open Source" category.
I know some source was around, I used to use RSX-11M which did a compilation from source as part of the system generation, and used it to debug a memory leak as late as 1985.
But all (apart from possibly the unix packages - I can't remember) that I mention did NOT have source.
Your horizons don't extend far enough back in time. It wasn't Gates who started commercial software.
I did a number of projects on PDP 11s in the late 1970's and early 1980s and often had to buy software for it. In one instance I remember paying £3000 for a Pascal compiler. In another, I paid a similar amounts for both a Unix System V licence and for Emacs.
By the beginning of the 1980's independent database vendors such as Oracle were selling software (our company competed for a while with similar product, but the company pulled out when the business plan showed we would have to sell a 1000 copies to recover the investment - and we decided to concentrate on bespoke application system development)
It looks to me like the actually carefully analysed the code and came up with the 985 errors. What about the ones they missed?
It seems they decided on the Windows XP as the worse end of a generic commercial range.
The only way I know of getting a good handle on how many bugs exist in a body of code is to plot the occurrence of bugs during development and into production against time.
You also plot the fix progress, and try and find out how fixing bugs generates new ones (because you see an increase in new bugs after a release that fixed old ones).
You can then assume that where the number of bugs that exist in code is headed is to a point where statistically fixing one bug creates one.
So I agree this is a flawed comparison.
I worked a lot in software development in the 1970's and 1980's, First as a programmer, but by the later 70's as a project manager and then later as a line manager for a unit built customer systems to a deadline based on a product and its enhancements. (although I am in the same company still, I now am doing business development in another area)
One of the most important lessons that I learnt during that period is that although you can raise productivity with overtime when you are reaching a deadline, trying to do so over a long period is just counterproductive and normally makes you miss deadlines AND budget and deliver crap code. In one mess I had to sort out, the previous project manager had driven the team so hard that they reported the project ready for client acceptence when it wasn't. This was an international project and the client came one quarter way round the world to witness acceptence tests only for them to fail dismally. It took us nearly a year to fix that mess. But proper re-planning (remembering every extra man day on this project was off the bottom line - since it was a fixed price contract) allowed us to beat the revised schedule by a month.
In the late '80s early '90s I was the "chief engineer" of our particular subsidiary and I spent some time examining the software development process in depth to understand how to get better at it. I looked at this overtime issue and why productivity goes down.
There are several reasons for this
1) With an alert mind, programmers (or anyone doing a serious mental task) have to move into a concentration mode. If not tired, it takes about 15 minutes with out interruption to achieve. Firstly when you are tired it takes longer to get into concentration mode, and secondly as the time pressures grow the supervisor level spend more time interrupting you to get "progress reports"
2) when people rush around without taking time to stop and think its a lot easier to go in the wrong direction. To solve bugs, or just to plan what to code need the ability to release oneself from the pressure.
[I vividly remember from my early days working in the middle of London and slaving all day over a bug or a design problem. Many a time, I remember realising what the problem was within a minute or two of getting on the train home]
3) A large part of a project is about communication and a common understand and vision of how everything fits together. Its the leader that is pulling the team forward and creating the environment for good communication that makes things happen more right the first time (nobody gets it completely right the first time do they). When the "workers" in the team get pressured into long hours then resentment sets in and that in itself kills productivity.
4) People use their lunch hours and the time after work to do essential domestic chores - such as visiting the bank or posting letters etc. If they skip lunch they just do them anyway in other times.
I would very rarely allow my project managers to authorise overtime - normally it was only when a deadline was close. In the later days I also used to teach on our internal project management courses and this point was always somewhere on the syllabus.
If you follow the links in the article, you will get to the original article from the guy who has written this software. Tutorial D is just one implementation of the 'D@ specification. The reason it is called Tutorial is that it is not an enterprise strength version, but rather one for learning and experimental purposes.
I have continually looked around at alternative e-mail clients to Kmail. Apart from Outlook, I have yet to find another mail client that has a key piece of functionality - the ability to clear out old messages from a mail folder automatically.
I read a lot of mailing lists - some such as Debian-User with several hundred messages a day. I filter each mailing list into its own folder, and then set purge dates on the folder to delete messages.
I tried evolution, thunderbird, balsa and a few others - none of them have this function. Why doesn't this lack of ability to clear unwanted mailing list messages worry anyone else?
Unfortunately its not that simple. My wife uses a machine that is adequately powered running Win98. I tried to upgrade to WinXP, only to find that all her favorite programs (old DOS games) would no longer run. I had to quickly downgrade her.
Your argument misses an important element - The Market
If I am a programmer whose business is producing software, then I have to be able to sell my software at a price that is comparable in the market place. With a "GPL" economy around me, I don't necessarily have the luxury of being able to be in the software business just selling what I have produced. I have to find another model - one of which is to add value to existing GPL code, and then sell services around it, using the value that you have added as a marketing tool for your services.
Just to emphasise this a bit more, to show that there is more to the software business than selling products, I work for a "Systems Integration" company with a turnover of about I billion pounds (UK pounds) - where most of our business is buying 3 party software in, and then adding services (such as writing interfaces, configuring it etc etc). At the moment, much of that is with proprietary licenced software, but increasingly it supports our competitive bids to be able to offer GPL'ed software without the additional (to our services) licence costs. With a company branding which provides the world with an image of technical excellence, most of our sales activities are centred around our knowledge of their business using references from previous projects to show capability, and getting a price that is lower than our competitors. It does not matter to us that we have to provide the source of any modifications or enhancements we make.
The other thing you miss, is the "you scratch my back, I'll scratch yours" element of the GPL. That is, together with others - jointly you produce software that is better than either one of you could singularly - and with the GPL protections no-one else can ride on your back without scratching it. Looking at this from a capitalist point of view, this jointly helping one another is making the business model I described above more competitive than the old one of proprietary companies producing closed source software.
Or get Open Source ones from http://dri.sourceforge.net
/etc/modules.
Debian includes these packages automatically - which is what I use - in otherwords binary drivers which an installation does exist. Just depends on which one.
There is an issue which took me quite a while to get working - an am not sure if the lastest hardware discovery software packages now solve the problem, but you have to load
agpgart, the agp modules for your motherboard (mine is sis-agp) and the hardware driver (mine is radeon for the radeon 9200 series chips) ahead of X starting. I put lines doing this in
I don't understand why you slam the previous poster. He said it was a rhetorical question - and it was to illustrate that even a monopoly cannot just increase the price beyond where it matches the value people get from a product. Even with a monopoly there is a price ceiling.
Although Linux may not be for everyone at the relative price difference between Windows and Linux, as the price difference increases more individuals will take the plunge. As that happens more companies will be incentivised to install it pre-configured, to develop device drivers for their hardware and do all the things for linux that currently make people find windows not to challenging techncially. This will lead to more people switching to Linux, driving the incentive for more companies to invest in the services market around it.
I would say that we are close to a point where this might happen anyway (viz the statements made by IBM, Novell and Sun all indicate they think this point is close). A price increase in Windows will just speed the process up.
Of course, people may just stick with the older versions of software for now on their current hardware. But even those will eventually need to get new hardware and will be forced into an upgrade. It will be interesting to see at what point hardware vendors include (and eventually default to) a linux option.
Simple really - receives mail and decides what to do with it. Could deliver it locally, or forward it on to another MTA.
Of course its a bit more complex than that. These days you probably want to do some validation of the mail that you are receiving (checking that the sender is actually who he says he is, filtering out spam and/or viruses), and you also need to do a level of optimisation when mail is addressed to several people. The other main function is to rewrite addresses (for instance, inside an organisation you may be fred@mymachine.company.com - to the outside world you may be fred.bloggs@company.com)
Except that the printers are designed with part of the electronics in the ink cartridge, making it expensive and difficult to copy. This is a game played by the printer manufacturers to control this market.
Advanced course in Capitalism,
The problem with that argument is that we are talking about very small market shares anyway, and so these assumptions don't necessarily hold.
My household has 6 computers (including daughters at University Laptops and my work owned laptop) where two of the machines run linux. However 5 windows licences were purchased and zero linux licences (debian - all completely downloaded).
Whilst linux desktops in corporate life are still in the "pilot" stage, I can also see that its going to be much more by a single copy of a distribution and in house distribution on existing hardware.
So whilst the linux market share is small, these sort of situations where the figures are completely screwed will be in the majority.
Of course once a market share does start to become significant, then the "install it yourself" installations DO become insignificant and then the logic holds true.
You will find even more to like when 3.2 is out on Monday. I am running a pre-release here, and it is faster, Konqueror seems less dodgy on rendering sites, and there are a range of new facilities,
I have tried unsuccessfully to find a better mail program than kmail
Things that it does that I haven't found another mail client to do for me
1) allow me to purge old mail messages automatically from a folder after they get to a certain age (essential if you are reading high volume mailing lists)
2) Single key (space bar) navigation down each message, unread messages in folder, over all folders (except trash)
Those two simple things have kept me back from switching from KDE to Gnome.
Not having done it from knoppix, I may well be wrong about this, but ...
/etc/apt/sources.list and use aptitude to set precisely the apps that you want.
/etc/lilo.conf is how you want it for the final system and then run lilo. Voila a ready to boot stable environment.
Can't you just run "debootstrap stable targetdir" (where targetdir is the mounted root partition that you want to eventually use for your on disk setup)
mount proc on targetdir/proc
then chroot into it edit
Ensure
I've done almost that in to build a chroot environment inside my existing one when I was trying to test something.
Thats how I feel about it too. I have been running a server, initially on stable, but then about a year ago upgraded (using aptitude after changing my sources.list to testing) on the fly and it just never goes wrong. The only slight hickup was when _I_ decided to switch from exim to exim4 and it took me a few days of planning to make sure I understood the new configuration file before making the move.
This server is also running a web site, and I've managed to switch between hand crafted web pages, PHPNUKE, PostNUKE and Wiki and back all withough problems too.
I run unstable on the desktop - occassional glitches, but nothing that stops me finding a way round and continuing to work.
I argue on my my website that I could shun proprietary software in order to encourage more people to produce free software. In other words I am using my customer "buying" pressure to favour the supplier who is providing me the best offer to improving that offer.
Totally for my own benefit, but economically providing the same effect
Remember DMCA is only an American law. And although I don't believe it prohibits reverse engineering for detecting copyright enfringement, even if it did, the community could do it outside the US.
Mplayer folks are hungarian aren't they?
There already is a whole raft of standards for home interconnection, and then home to outside world.
How does this new standard add to that?
Look at OSGi, uPnP and LonWorks just to show a few of them
So
We run Windows 98 SE on our family machine at home where my wife plays lots of old DOS games in full screen mode. I think there is a dynamic resolution change from 1024*768 to 640*480 as she starts the game up.
I bought Windows XP (Home Edition) to upgrade her. However, all her DOS games started running like a pig. I concluded that something in the XP driver was trying to do some form of autoscaling on the fly. I had to uninstall and go back to Win98SE.
So either there is a massive functionality loss - from my point of view - or, there is some setting I have yet to find which allows these DOS games to get direct access to the hardware.
Anyone out there got the same problem, or know how to solve it?
Although this blackout occured in the US, it has not been the only one. Italy, Scandinavia and the UK have also suffered similar problems.
This has been sending political shockwaves all over the place as the realisation that the pressure on costs on the network owners has meant that contigency margins have been reduced. In the UK (where I live) the public face is one of calm, but behind the scenes there is considerable concern.
What it shouldn't mean though, is a step back to the old days when networks were considerably over engineered. Care should be taken that those who advocate this don't put us back 15 to 20 years and hike the price of energy as a result.
What I think is needed is much more sophisticated control systems that recognise the domino effect built in as a safety measure and actually proactively shutdown parts of the networks in much the same way as firemen make firebreaks to stop a rampant fire. This would keep control of a situation that currently just runs out of hand, and would prevent both the widespread loss of power, but also a much faster return to normality.
Groklaw spends a lot of time analyzing every SCO press release down to the tiniest detail, as if they really mattered. It's also populated by a group that makes Slashdot posters look like disinterested observers.
Really, the basic interpretation of IBM's UNIX contract is unclear, which means that SCO still has a real shot at this.
Groklaw are not only looking at the press releases, much more importantly they are looking through and analysing the legal documents
Now of course SCO may have a chance but the more you analyse the legal documents the more unlikely that seems
Now you may be trolling, but I can't see how either side of the argument can be won by only making assertions. You have to follow through with evidence. As I see it the pro SCO camp haven't produced evidence, the anti SCO camp have produced arguments in abundance.
I know where I will be putting my money, and it won't be in SCOX stock.