This is what most hosting companies who offer VPS's do, as Virtuozzo is one of the only virtualisation packages that doesn't suffer from massive performance hits. Granted this is more about what level things are virtualised at than better coding but the performance difference is very significant.
I've seen servers running over a hundred virtual servers (mostly all low usage) without any problems, something that is completely out of the question with VMWare (even ESX).
I'm sure there will be many who will point out that it has less isolation and therefore less secure, however so far I haven't heard of any compromises.
Probably not quite what you want, but I've used an Instreamer before to digitise audio and distribute it across a LAN. It's not exactly cheap but works well and simple to use.
Otherwise, you could write / find some software to suit (I ended up writing my own). All I did was write a simple app that fed raw PCM data across a UDP socket and some playback software at the other end. With less than 20ms overall latency it was pretty good for near realtime usage, and used a very small amount of bandwith (8kHz signal, 16 bit).
I'm assuming they were refferring to the signal bandwidth, not the actual bitrate. From what I understood they were trying to say that it's the same bandwith, but significantly better quality.
As a reference for efficiency, most CDMA phone services have around 1.25Mhz of raw bandwith and they can fit 32,000 channels into this!
Have you actually tried using Mono and C# at all? It's actually a lot more mature than you think, in fact I was shocked how mature it already is. The other thing is the development rate of it, a lot of work is being done to finish the.NET 2.0 stuff off and a lot of work is being completed on the developer. I'd consider it's current state as "production ready" on both Windows and Linux systems (and probably Sun, but I haven't tested it).
I've already found it's ability to use libraries from multiple sources very helpful, I've used a DLL written in VS.NET 2003 that had no consideration for Linux yet it works perfectly. Mono also allows you to run Java code in the Mono environment but also allows you to use the.NET stuff (and other code you write in C# etc) in Java! That flexibility to me is very impressive, it's well worth checking out.
As for the main topic, both C# and Java are just languages. Programming is a lot about knowing how to think about a problem rather than the actual syntax. If you can conceptualise how to tackle a problem in a programmatic way that's half the job of a programmer. Learning the language itself is a secondary task and you can pickup a working knowledge of most languages within a few days.
Don't limit yourself to just one language to be an expert in, it'll limit your ability to be employed down the track. Looking at C# and Java however are good choices, if you know how to program in C you'll wonder why it took you so long!
I like C# but I've been doing it all using Mono which does allow C# and.NET to be used across multiple platforms just the same as Java. Don't just limit yourself to Microsoft's implementation of C#, using Mono allows you to build some pretty powerful applications quite quickly (using Glade and GTK for the GUI side of it).
The old saying of "if it's not broke, don't fix it" may be true, but what about if you can make it better? What if it's working currently but you have the ability to halve either the resources needed to support it, or the time taken to use it.
To me the best time to start the migration to a different product / platform is before too many problems occur or before it's type of life is complete.
For example, if the old system was broken then the rollout schedule of the new one is usually rushed and leads to more mistakes and problems. I've unfortunately seen this over and over, and even cases where they've rolled back to the old system.
It never hurts to start the planning early, even if it means mistakes and changes like this.
Maybe this is a good lesson we can all learn from and try to improve. How much OSS/FS are you using that you haven't contributed anything to? You don't have to be a programmer to help out, there's plenty of other tasks. You could:
Contribute to the documentation
Help by answering questions to mailing lists / forums
Write a tutorial
Offer to update the website design / content
Write / add to the FAQ
Don't get me wrong, I'm not trying to judge people. I'm guilty myself of not contributing much back to the OSS community but it's something I want to change this year. It doesn't have to be much, many hands make light work as they say.
Solutions exist to implement secure WiFi, but it comes with a cost.
Harris makes an encrypted PCMCIA 802.11b based card that has high grade encryption built in. It certainly makes the system impossible to get into, but they're far from cheap ($2k+).
Most satellites have a limited footprint/ . By this I mean they have a limited area in which the satellite dish is pointed (in general terms), so buying a service from from outside the footprint of that satellite means it won't work.
I have a HP nx7010 laptop, and surprisingly Ubuntu 5.04 works perfectly on it. A base install has the screen driver and resolution fine (ATI 9200 mobile + 15.4" 1680 x 1050 screen), power management working (Pentium-M drops to 600Mhz at idle and ramps up as required), wifi card working (although I am using an alternate to the one supplied) sound and ethernet all detected.
Not even Windows got this close:)
I have yet to test the 56k modem (or even look for that matter), the SD card reader (as I don't use it) and I've yet to try to suspend the laptop. But, for the rest to work without any hassles impressed me. I started playing with Linux back in the RedHat 4 days, and it was a different story then!
One good way to test Linux compatability would be to take a Live CD into the store and run it first. This way you can see within a few minutes if you will have any dramas or not!
Not really, many government organisations have been using Linux and F/OSS for may years now, only difference is there's now an "official" guide to help people out. It's not going to mean that every government department will switch 100% of their systems over, but help provide information which will allow them to make the best choice. A donation from Mr Gates won't change a thing.
For those who haven't worked in an Australian government IT section before, it's a different world to the private industry. The budgets are generally higher and paperwork / justification required is usually higher than the "real" world.
This is where this document will help, there's now an "official" reference point that can be used to justify the use of F/OSS solutions.
Where the promotion of most F/OSS systems fail is that they focus on the cost as the most important factor. This isn't the biggest gain, especially for departments where the budget makes it an insignificant factor. Freedom from proprietary file formats and the ease of customisation I think are the biggest benefits and this needs to be the focus.
One thing that I can't see addressed properly is a government scheme to donate funding to a lot of the smaller open source development projects. For the government to pay anyone (in most instances) they want an invoice and a product to be delivered.
It would be great to be able to submit a recommendation to a government organisation that can then distribute small amounts of funding to the developers. This would allow the smaller projects (eg a simple php calendar app used on a government intraweb) to get some recognition and possibly allow them to develop their software further.
It will be interesting to see if this model actually works.
The subscription (IMHO) is a different to things like the SA forums, this monthly access isn't there to cover the costs of hosting. Atomic PC is run by a large company (AJB Publishing, which is a part of Haymarket Media) and the way I see it the printed magazine is their core business.
Is this the only magazine that will be charging for access to their supporting forums? Is this to try and convince users to buy the magazine or is it to try and reward buyers with extra benefits? I'm just not too sure why they are charging money.
I can understand hobbyist type sites needing to cover costs, they normally don't have other products to offset the hosting costs. I run a small Australian PC website with a few others (here) and can appreciate the fact that the owners don't want to be left out of pocket for hosting costs.
We have always remained free to use and read, offsetting the costs either with sponsorship or a single banner ad on each page. I'm hoping that I can always keep it that way, as I know a subscription based model would never work for us.
I've used a few different 2 way satellite links, latency is generally about 500-600ms for a round trip in the systems I used. Using the voice port in a Cisco router was quite useable, in fact I was surprised at how well the TCP/IP stack functioned with the high latency.
Slashdot Mirror
This is what most hosting companies who offer VPS's do, as Virtuozzo is one of the only virtualisation packages that doesn't suffer from massive performance hits. Granted this is more about what level things are virtualised at than better coding but the performance difference is very significant.
I've seen servers running over a hundred virtual servers (mostly all low usage) without any problems, something that is completely out of the question with VMWare (even ESX).
I'm sure there will be many who will point out that it has less isolation and therefore less secure, however so far I haven't heard of any compromises.
Otherwise, you could write / find some software to suit (I ended up writing my own). All I did was write a simple app that fed raw PCM data across a UDP socket and some playback software at the other end. With less than 20ms overall latency it was pretty good for near realtime usage, and used a very small amount of bandwith (8kHz signal, 16 bit).
I'm assuming they were refferring to the signal bandwidth, not the actual bitrate. From what I understood they were trying to say that it's the same bandwith, but significantly better quality. As a reference for efficiency, most CDMA phone services have around 1.25Mhz of raw bandwith and they can fit 32,000 channels into this!
I've already found it's ability to use libraries from multiple sources very helpful, I've used a DLL written in VS .NET 2003 that had no consideration for Linux yet it works perfectly. Mono also allows you to run Java code in the Mono environment but also allows you to use the .NET stuff (and other code you write in C# etc) in Java! That flexibility to me is very impressive, it's well worth checking out.
As for the main topic, both C# and Java are just languages. Programming is a lot about knowing how to think about a problem rather than the actual syntax. If you can conceptualise how to tackle a problem in a programmatic way that's half the job of a programmer. Learning the language itself is a secondary task and you can pickup a working knowledge of most languages within a few days.
Don't limit yourself to just one language to be an expert in, it'll limit your ability to be employed down the track. Looking at C# and Java however are good choices, if you know how to program in C you'll wonder why it took you so long!
I like C# but I've been doing it all using Mono which does allow C# and .NET to be used across multiple platforms just the same as Java. Don't just limit yourself to Microsoft's implementation of C#, using Mono allows you to build some pretty powerful applications quite quickly (using Glade and GTK for the GUI side of it).
To me the best time to start the migration to a different product / platform is before too many problems occur or before it's type of life is complete.
For example, if the old system was broken then the rollout schedule of the new one is usually rushed and leads to more mistakes and problems. I've unfortunately seen this over and over, and even cases where they've rolled back to the old system.
It never hurts to start the planning early, even if it means mistakes and changes like this.
Don't get me wrong, I'm not trying to judge people. I'm guilty myself of not contributing much back to the OSS community but it's something I want to change this year. It doesn't have to be much, many hands make light work as they say.
Solutions exist to implement secure WiFi, but it comes with a cost.
Harris makes an encrypted PCMCIA 802.11b based card that has high grade encryption built in. It certainly makes the system impossible to get into, but they're far from cheap ($2k+).
Product: SecNet11
In the end, a lot of the exploitable networks comes from either poor management, lack of information or lack of control within government areas.
Most satellites have a limited footprint/ . By this I mean they have a limited area in which the satellite dish is pointed (in general terms), so buying a service from from outside the footprint of that satellite means it won't work.
Not even Windows got this close :)
I have yet to test the 56k modem (or even look for that matter), the SD card reader (as I don't use it) and I've yet to try to suspend the laptop. But, for the rest to work without any hassles impressed me. I started playing with Linux back in the RedHat 4 days, and it was a different story then!
One good way to test Linux compatability would be to take a Live CD into the store and run it first. This way you can see within a few minutes if you will have any dramas or not!
For those who haven't worked in an Australian government IT section before, it's a different world to the private industry. The budgets are generally higher and paperwork / justification required is usually higher than the "real" world.
This is where this document will help, there's now an "official" reference point that can be used to justify the use of F/OSS solutions.
Where the promotion of most F/OSS systems fail is that they focus on the cost as the most important factor. This isn't the biggest gain, especially for departments where the budget makes it an insignificant factor. Freedom from proprietary file formats and the ease of customisation I think are the biggest benefits and this needs to be the focus.
One thing that I can't see addressed properly is a government scheme to donate funding to a lot of the smaller open source development projects. For the government to pay anyone (in most instances) they want an invoice and a product to be delivered.
It would be great to be able to submit a recommendation to a government organisation that can then distribute small amounts of funding to the developers. This would allow the smaller projects (eg a simple php calendar app used on a government intraweb) to get some recognition and possibly allow them to develop their software further.
Does such a scheme exist?
The subscription (IMHO) is a different to things like the SA forums, this monthly access isn't there to cover the costs of hosting. Atomic PC is run by a large company (AJB Publishing, which is a part of Haymarket Media) and the way I see it the printed magazine is their core business.
Is this the only magazine that will be charging for access to their supporting forums? Is this to try and convince users to buy the magazine or is it to try and reward buyers with extra benefits? I'm just not too sure why they are charging money. I can understand hobbyist type sites needing to cover costs, they normally don't have other products to offset the hosting costs. I run a small Australian PC website with a few others (here) and can appreciate the fact that the owners don't want to be left out of pocket for hosting costs.
We have always remained free to use and read, offsetting the costs either with sponsorship or a single banner ad on each page. I'm hoping that I can always keep it that way, as I know a subscription based model would never work for us.
I've used a few different 2 way satellite links, latency is generally about 500-600ms for a round trip in the systems I used. Using the voice port in a Cisco router was quite useable, in fact I was surprised at how well the TCP/IP stack functioned with the high latency.
Aah, so this is why the game is late, they're waiting on WinXP 64! I knew there would be a valid reason :)