In particular the 8500 chip. They have a reference DVD player design, which just needs an IDE DVD transport, front panel and PSU to be up and running. And it's tiny. I like reference designs; they make it much easier for small companies to put together interesting kit that the big guys don't see a big enough market for.
Let's consider another fact. The source to the Linux kernel, along with the mountain of GNU code and others licensed under the GPL, is available without prejudice. It won't disappear if RedHat, SuSE, Mandrake, Caldera et. al. all die tomorrow in a freak accident.
Compare that to what would happen to MS software if MS were to disappear up their own behinds in a flash of financial lucidity (make of that what you will) - I seriously doubt Microsoft would ever open their source base in any meaningful fashion, even if ordered to by the courts.
I agree - forget arguing about tools; changing CM tools will not solve the problem, since the problem is not with the tool. The problem lies in lack of analysis and design.
If you don't control your interfaces properly you're sunk. Unstable interfaces mean:
You don't know what your code should do. If you did, the interface would have been pretty much correct at the start.
Other people in the team don't know what your code should do - how can they when you don't know - so how can they use it?
If anyone does use your code, they are going to have to change their code every time you change yours.
If no-one knows what your code should do from one minute to the next, how can it be tested?
So interfaces are key. If you want to control interfaces efficiently, you need to minimise change to those interfaces. To minimise change in the interfaces you need to know and understand what it is that your software should do. Good design and, especially, good architecture attempt to address exactly that problem.
One last point; the function of a CM tool is to provide a quality record; essentially so that one can always re-create a previous build etc. That's what they're designed for. They are not there to patch up sloppy development practices.
Notepad? Real MS Windows Users use MS Word. Then wonder why the system doesn't understand their 12-point Arial doc file that looks on screen exactly the same as the stuff their geek friend had in Notepad...
Incidentally, I have colleagues who use Word as their tool of preference for editing source code. Mainly so that they can use the spellchecker. What is the world coming to...
This announcement from Sputnik struck me as interesting; it perhaps indicates an answer to "how do you make money in an open source environment".
Specifically, from what I can see, Sputnik does not make money from its source code as such, it makes money from the network and its partners in the delivery of whatever they deliver over the network. Note that entry to the network is authenticated - you can't just walk in without an account so ultimately users will pay for access to the network.
For Sputnik, they have presumably decided, by and large, that they won't be making money specifically from the software. They intend to make their money actually delivering service and perhaps content. Making the underlying software technology open source drastically reduces the cost of the software (i.e. maintenance, development etc). From a business perspective this is a Good Thing - don't pour money into something that isn't going to make you a profit back.
Further, much is made of the Sputnik team having avoided publicity until they have a working product. This also makes good business sense when you consider that building their partners and network is where the value lies, and if they had released early betas left right and centre they would probably have delayed their own launch whilst at the same time encouraging other people to, er, borrow the idea and corner network partners before Sputnik could.
Ultimately, if the idea takes off, there will be more than enough space for competitors. However early in the game, it will only really be effective in certain small areas - cities that already have a lot of spare/obsolete IT capacity in company offices that they might as well get some value out of by becoming service nodes on the radio network. Potential competitors early in the game could mess up the whole idea, fragmenting the market before it has become established, leading at best to long delays until the behemoths come in and clean up from the pieces of failed start-ups.
Any potential competitor wishing to enter at this stage, either has to use the open source software (hence maintaining standards & visibility), or alternatively drop a wedge of cash developing a proprietary alternative, thereby delay their own entry - undesirable at the start of something like this.
Overall I think the open source concept makes exceptionally good sense in this case.
Re:Algorithm patented ? Not in Europe...
on
More on MPEG4
·
· Score: 1
See the FFII software patents working group for a large collection of material on patents in Europe. In particular their horror gallery of active European software patents shows that software patents are alive and well in Europe.
Re:And how are they supposed to measure this?
on
More on MPEG4
·
· Score: 1
In the UK, perhaps the most expensive place to by fuel in Europe, about 80% of the pump price of unleaded petrol is tax, according to the Automobile Association
(note; 'gallons' on the site refer to UK gallons at 4.454litres/gallon as compared to US gallons at, IIRC, 3.8litres/gallon). They also have comparison chart of worldwide fuel prices - compare 21.33 pence/litre in the US against 71 pence/litre in the UK.
A quick rummage on Intel's site for the '4004' that Ivan found reveals a tech specs sheet which shows that the 4004 had 640 bytes of addressable memory and a 4-bit bus. Since the number of bugs in code is broadly proportional to the size of the code, it is not unreasonable to achieve bug-free in 640 bytes (or at least, free of bugs that actually cause a problem that cannot be resolved without modifying the code).
Slashdot Mirror
In particular the 8500 chip. They have a reference DVD player design, which just needs an IDE DVD transport, front panel and PSU to be up and running. And it's tiny. I like reference designs; they make it much easier for small companies to put together interesting kit that the big guys don't see a big enough market for.
Compare that to what would happen to MS software if MS were to disappear up their own behinds in a flash of financial lucidity (make of that what you will) - I seriously doubt Microsoft would ever open their source base in any meaningful fashion, even if ordered to by the courts.
Out of interest, how much corporation tax does Microsoft actually pay?
If you don't control your interfaces properly you're sunk. Unstable interfaces mean:
So interfaces are key. If you want to control interfaces efficiently, you need to minimise change to those interfaces. To minimise change in the interfaces you need to know and understand what it is that your software should do. Good design and, especially, good architecture attempt to address exactly that problem.
One last point; the function of a CM tool is to provide a quality record; essentially so that one can always re-create a previous build etc. That's what they're designed for. They are not there to patch up sloppy development practices.
Incidentally, I have colleagues who use Word as their tool of preference for editing source code. Mainly so that they can use the spellchecker. What is the world coming to...
This announcement from Sputnik struck me as interesting; it perhaps indicates an answer to "how do you make money in an open source environment". Specifically, from what I can see, Sputnik does not make money from its source code as such, it makes money from the network and its partners in the delivery of whatever they deliver over the network. Note that entry to the network is authenticated - you can't just walk in without an account so ultimately users will pay for access to the network. For Sputnik, they have presumably decided, by and large, that they won't be making money specifically from the software. They intend to make their money actually delivering service and perhaps content. Making the underlying software technology open source drastically reduces the cost of the software (i.e. maintenance, development etc). From a business perspective this is a Good Thing - don't pour money into something that isn't going to make you a profit back. Further, much is made of the Sputnik team having avoided publicity until they have a working product. This also makes good business sense when you consider that building their partners and network is where the value lies, and if they had released early betas left right and centre they would probably have delayed their own launch whilst at the same time encouraging other people to, er, borrow the idea and corner network partners before Sputnik could. Ultimately, if the idea takes off, there will be more than enough space for competitors. However early in the game, it will only really be effective in certain small areas - cities that already have a lot of spare/obsolete IT capacity in company offices that they might as well get some value out of by becoming service nodes on the radio network. Potential competitors early in the game could mess up the whole idea, fragmenting the market before it has become established, leading at best to long delays until the behemoths come in and clean up from the pieces of failed start-ups. Any potential competitor wishing to enter at this stage, either has to use the open source software (hence maintaining standards & visibility), or alternatively drop a wedge of cash developing a proprietary alternative, thereby delay their own entry - undesirable at the start of something like this. Overall I think the open source concept makes exceptionally good sense in this case.
See the FFII software patents working group for a large collection of material on patents in Europe. In particular their horror gallery of active European software patents shows that software patents are alive and well in Europe.
In the UK, perhaps the most expensive place to by fuel in Europe, about 80% of the pump price of unleaded petrol is tax, according to the Automobile Association (note; 'gallons' on the site refer to UK gallons at 4.454litres/gallon as compared to US gallons at, IIRC, 3.8litres/gallon). They also have comparison chart of worldwide fuel prices - compare 21.33 pence/litre in the US against 71 pence/litre in the UK.
On EU software patents - see here from the FFII Swpat Workgroup (as pointed out by the FSF). The situation could soon become worse than the US.
A quick rummage on Intel's site for the '4004' that Ivan found reveals a tech specs sheet which shows that the 4004 had 640 bytes of addressable memory and a 4-bit bus. Since the number of bugs in code is broadly proportional to the size of the code, it is not unreasonable to achieve bug-free in 640 bytes (or at least, free of bugs that actually cause a problem that cannot be resolved without modifying the code).