With coal/nuclear etc, you know well in advance where your fuel is coming from, how much of it you will have, and its easy to keep a large reserve incase of delivery problems. Sources of power that rely on ever changing environmental conditions are far less predictable.
What if a large solar or windfarm used it's surplus power to extract hydrogen from water (a rather inefficient process), and store the resulting hydrogen to be burned when theres no sun or wind? Especially if you combined solar and wind power on the same site, the wind usually picks up at night when there's no sun.
The problem is one of location... A nuclear or coal plant can be built pretty much anywhere, whereas wind farms are only any use in areas that get a steady supply of wind, which tends to be remote and quite high locations.. A stack of wind turbines on top of a mountain will be visible for miles around, when before it was a pretty undisturbed area as it's otherwise fairly worthless for humans.
Traffic from users using msn, or traffic from msnbot? I noticed that the msn crawler causes a disproportionate amount of traffic on the sites i host.
As for your increased hits from msn search, could it be that your site changes have caused it to get a higher ranking in the searches? Or perhaps your site has content that's more applicable to msn users... A linux related site is unlikely to get many hits from msn for instance, because not many linux users would want to use msn.
Perhaps see what the users are searching for, and see how high you come in different search engines, i would imagine your ranking in different engines will have a huge impact on the number of referred hits you get.
Yes, google are half a diverse, having one viable product instead of 2... They're also a lot newer, and competing in a market which is still highly competitive and very active.
MS on the other hand, have 2 profitable products, in markets where they have kept competition stifled for years, and which would otherwise have become commoditized by now. The commoditization is coming, sooner or later, MS will fight it tooth and nail until the bitter end, but the market should correct itself eventually, especially with outside intervention from the EU etc.
Advertising is a different market, it's much harder to lock other players out of the market in the same way, and the leading search engine has already changed several times (who remembers altavista?) and it's trivial to start using another if they offer a better service than google.
Yes, everyone needs an OS, but it also makes no sense for everyone to pay ms's ridiculously high prices for such a common commodity that can be produced so cheaply.
Google are in the same boat tho, they need to diversify in order to remain relevant, the current economic conditions are likely to cause a decrease in advertising etc, and there's no guarantee that internet advertising will still be wanted in a few years time.
Google's share of instant messaging is quite small, but growing.. And remember it's jabber based, so they can syndicate with other parties, i believe livejournal supports jabber, not sure if any other significant sites do, but theres plenty of smaller jabber servers too.
Then there is still AOL, who's messaging service is much bigger than msn/yahoo in some markets.
What i dislike about yahoo/msn im protocols tho, is that they were late to the party, and yet still chose to create a proprietary protocol despite a standard one existing. AOL created their own when there was no alternative, which isn't quite so bad tho they could have opened it up more/sooner.
The problem comes back to support. What do you tell a newly converted Linux user when he/she asks why their software and hardware won't work? You tell them to call up the company from whom they bought linux support from... Just like you would if they were using windows.
The difference is that linux offers a free unsupported version, you wouldnt complain about being unable to call microsoft for help if you had a pirated copy of windows (the closest thing you can get).
And they certainly would have shown that, if they could, since it would greatly help their campaign. The fact they can't is actually a lot more telling.
Well, the P4 was pretty laughable.... Modern GPUs are just adding to the increased power, compared to older less power hungry units. I would say that the core2 quad i have here consumes somewhat more power than the p3/933 it replaced... As for a dual p3, multi cpu systems are more likely to be higher end, and thus have more hardware, and more powerful power supplies etc. I had a quad capable p2/xeon, and it used a huge amount of power, but probably not more than a modern equivalent.
But the point is, if you were to use the technology of the core2, but use it to produce a system performance equivalent to the P54c, you would be able to use a small fraction of the power. And most end users did the same things on their p54c systems as they do on today's quad core boxes, only now they use more bloated software to do it.
True, but doesn't get rid of the bloat... With more efficient software, you could consolidate a lot more functions onto a single system using virtualization. Also, there's no need to run multiple complete operating systems, that's far from efficient, why not simply run all the apps on top of a single kernel on a single machine? Same result, but you can use shared mem between each instance (libc etc), and you don't have the overhead of a vm, io device emulation and multiple kernel instances running.
Yes, printers can churn out paper much faster than a typewriter... It's really quite ridiculous the amount of paper that people use these days, considering the technology we have available.
Each new architecture does use less power, but that's not the way things are heading... New CPUs may have much better performance/watt, but the overall performance is increased too, thus the amount of actual power used stays the same or even increases. There's also increasingly bloated software, all this managed high level language code etc, which uses far more energy to do the same work. And modern powerful servers which sit idle for the most part. You could easily make lower performing servers using modern techniques, and reduce power consumption hugely... Modern embedded processors are faster than high end server processors from a few years ago, and yet use a small fraction of the power, but they wouldn't be good running modern bloated apps in high level languages.
Making it more expensive won't punish the companies building datacenters, it will punish their customers, end users who want internet access, or to use online services. They should be ensuring the infrastructure is available and affordable in the places where cheap green power is available.
The problem is that commercial companies are providing broadband, and providing it to an outlying community with a small population and long cable distances isn't profitable.
If you have government or a non profit providing it, the cables are actually cheaper per mile to lay out there since there's less in the way, the cost of digging up city streets is very high because of the disruption it causes. If you just dig a trench alongside an empty highway, or alongside a railroad, you don't cause much disruption and can get the work done much quicker with a lot less red tape.
Well... What if the government owns the physical infrastructure, or a non profit body, and then providers rent the infrastructure from them... And force them to reinvest any and all profits in improvements of the underlying network.
Kinda like the UK system, but where the owner of the infrastructure isnt trying to compete with the same companies they're providing infrastructure to. In the UK, BT have to rent out lines wholesale to other ISPs as well as allowing the bigger isps to install kit in exchanges... Regulators actually keep the wholesale prices high, so that the companies installing kit in exchanges don't get priced out of the market by bt wholesale. The problem is that, installing kit in thousands of exchanges is still a high barrier to entry, and its not financially viable to do it everywhere, so some people are still stuck with no alternative to BT's service, which is kept priced high by the regulator.
If you instead do away with loop unbundling, and turn over the infrastructure to an independent non profit, then any isp and any end user will all be in the same boat.
As you've pointed out, privatised telecom provision will never be fair for everyone, and regulation can make things worse in some cases.
There is plenty of documentation of voting irregularities, which at the very least should be investigated before OOXML can be ratified.
The fast track process is for existing "defacto standards" that are widely used and implemented, and only really need a rubber stamp. OOXML is not widely implemented nor widely used at this point, it should have gone through the normal process. Perhaps the recent standardisation of PDF as ISO32000 was through the fast track, and would have deserved being fast tracked.
According to ISO guidelines, standards should reuse existing standards, preferably ISO ones... OOXML does not, it does mostly the same thing as ODF but in a completely different way, it also stores dates in a way conflicting with existing ISO standards, stores country codes in a different way, stores measurements in a different way and more. Thus it is in violation of ISO guidelines and should not have been approved.
There are other more specific issues, plenty documented out on the web... But the 3 above show where they have violated ISO rules, which should at the very least be enough to kick ooxml off the fast track and into the regular process.
As for ignoring it, unfortunately microsoft are large enough that they can force their inferior format on the market, so it will be impossible to ignore. If the market were free, and people were able to choose products based on technical merit microsoft wouldn't be anywhere near as big as they are.
The only valid point against OOXML is that it contains unclear and/or unimplementable aspects, thus denying others from the ability to create supporting implementations. However, if this is the case, and MS is unwilling to create OOXML implementations for non-MS/Apple platforms, how successful do you really expect the standard to be? Very. That's the problem, despite the inferiority of the "standard", microsoft have the resources to push it far and wide, thus harming the industry as a whole. That's the problem when the market becomes distorted by one party having too much power.
The issue is that OOXML is an inferior "standard", which has been forced through the process using bribery, corruption and ignorance, and all while a superior standard for the same thing already exists. If microsoft were truly interested in standards and interoperability at all, they would have implemented the existing standard, and joined it's steering committee when they were first invited several years ago.
We don't want to be forced to use that inferior format, simply because microsoft bought and paid for enough people at standards boards, as this will hurt the industry as a whole. It's only microsoft who stands to benefit from OOXML, everyone else loses in one way or another.
And some people like the power and flexibility of having such options, the trick is to have sensible defaults, and then file off less common options under an advanced option.
I don't understand why they think that, broadcom hardware isn't generally very good, and there are much better options available. Their BCM4400 10/100 ethernet cards are particularly bad in my experience (their 1gb chipsets seem a lot better, but i'd still rather have an intel card).
Their wifi cards, even with the supported windows drivers, generally under perform compared to other chipsets. They also make linux drivers, but only for specific embedded systems (access points) that use their chipsets.
I have an HP PSC 1315, it worked perfectly with OSX 10.4, I have 10.5 now and the scanner drivers don't work on it, the printer component works but only with apple's generic deskjet driver, not hp's driver. By contrast, HP makes open source linux drivers for this printer/scanner combo, they still work on the latest ubuntu release which is actually newer than OSX 10.5. The scanner doesn't work on Vista either...
I also have a WinTV USB2 tv tuner card, there are no vista drivers for it, it still works perfectly with an up to date linux system using the open source in-kernel drivers.
There's also issues with older nvidia cards, nvidia no longer maintains drivers for older cards, so they don't work with newer kernels. They also stopped producing drivers for the IA64 architecture, leaving those few of us who own IA64 hardware stuck with newer kernels or newer nvidia cards, and they don't support PPC or other architectures at all. By contrast, i have open source drivers for some truly ancient video cards, so they all still work perfectly.
I would suggest his computer, the data on it and his online accounts etc, now belong to his family along with the rest of his belongings, in the absence of any will saying differently, so if the family want you to look at it that's their right. It's no different to someone who maintains a paper diary, it falls to their family after death who may or may not read it at their discretion.
Slashdot Mirror
How practical would it be to extract hydrogen from water using excess energy, and then burn that as a backup plan?
With coal/nuclear etc, you know well in advance where your fuel is coming from, how much of it you will have, and its easy to keep a large reserve incase of delivery problems.
Sources of power that rely on ever changing environmental conditions are far less predictable.
What if a large solar or windfarm used it's surplus power to extract hydrogen from water (a rather inefficient process), and store the resulting hydrogen to be burned when theres no sun or wind?
Especially if you combined solar and wind power on the same site, the wind usually picks up at night when there's no sun.
The problem is one of location...
A nuclear or coal plant can be built pretty much anywhere, whereas wind farms are only any use in areas that get a steady supply of wind, which tends to be remote and quite high locations.. A stack of wind turbines on top of a mountain will be visible for miles around, when before it was a pretty undisturbed area as it's otherwise fairly worthless for humans.
Traffic from users using msn, or traffic from msnbot?
I noticed that the msn crawler causes a disproportionate amount of traffic on the sites i host.
As for your increased hits from msn search, could it be that your site changes have caused it to get a higher ranking in the searches? Or perhaps your site has content that's more applicable to msn users...
A linux related site is unlikely to get many hits from msn for instance, because not many linux users would want to use msn.
Perhaps see what the users are searching for, and see how high you come in different search engines, i would imagine your ranking in different engines will have a huge impact on the number of referred hits you get.
Yes, google are half a diverse, having one viable product instead of 2...
They're also a lot newer, and competing in a market which is still highly competitive and very active.
MS on the other hand, have 2 profitable products, in markets where they have kept competition stifled for years, and which would otherwise have become commoditized by now. The commoditization is coming, sooner or later, MS will fight it tooth and nail until the bitter end, but the market should correct itself eventually, especially with outside intervention from the EU etc.
Advertising is a different market, it's much harder to lock other players out of the market in the same way, and the leading search engine has already changed several times (who remembers altavista?) and it's trivial to start using another if they offer a better service than google.
Yes, everyone needs an OS, but it also makes no sense for everyone to pay ms's ridiculously high prices for such a common commodity that can be produced so cheaply.
Google are in the same boat tho, they need to diversify in order to remain relevant, the current economic conditions are likely to cause a decrease in advertising etc, and there's no guarantee that internet advertising will still be wanted in a few years time.
Google's share of instant messaging is quite small, but growing..
And remember it's jabber based, so they can syndicate with other parties, i believe livejournal supports jabber, not sure if any other significant sites do, but theres plenty of smaller jabber servers too.
Then there is still AOL, who's messaging service is much bigger than msn/yahoo in some markets.
What i dislike about yahoo/msn im protocols tho, is that they were late to the party, and yet still chose to create a proprietary protocol despite a standard one existing. AOL created their own when there was no alternative, which isn't quite so bad tho they could have opened it up more/sooner.
Just like you would if they were using windows.
The difference is that linux offers a free unsupported version, you wouldnt complain about being unable to call microsoft for help if you had a pirated copy of windows (the closest thing you can get).
And they certainly would have shown that, if they could, since it would greatly help their campaign.
The fact they can't is actually a lot more telling.
Well, the P4 was pretty laughable....
Modern GPUs are just adding to the increased power, compared to older less power hungry units.
I would say that the core2 quad i have here consumes somewhat more power than the p3/933 it replaced...
As for a dual p3, multi cpu systems are more likely to be higher end, and thus have more hardware, and more powerful power supplies etc. I had a quad capable p2/xeon, and it used a huge amount of power, but probably not more than a modern equivalent.
But the point is, if you were to use the technology of the core2, but use it to produce a system performance equivalent to the P54c, you would be able to use a small fraction of the power. And most end users did the same things on their p54c systems as they do on today's quad core boxes, only now they use more bloated software to do it.
True, but doesn't get rid of the bloat...
With more efficient software, you could consolidate a lot more functions onto a single system using virtualization.
Also, there's no need to run multiple complete operating systems, that's far from efficient, why not simply run all the apps on top of a single kernel on a single machine? Same result, but you can use shared mem between each instance (libc etc), and you don't have the overhead of a vm, io device emulation and multiple kernel instances running.
Yes, printers can churn out paper much faster than a typewriter...
It's really quite ridiculous the amount of paper that people use these days, considering the technology we have available.
They are quite expensive, especially considering they aren't paying for their power...
Mars would be good for long term off-site backups...
Each new architecture does use less power, but that's not the way things are heading...
New CPUs may have much better performance/watt, but the overall performance is increased too, thus the amount of actual power used stays the same or even increases.
There's also increasingly bloated software, all this managed high level language code etc, which uses far more energy to do the same work. And modern powerful servers which sit idle for the most part.
You could easily make lower performing servers using modern techniques, and reduce power consumption hugely... Modern embedded processors are faster than high end server processors from a few years ago, and yet use a small fraction of the power, but they wouldn't be good running modern bloated apps in high level languages.
Making it more expensive won't punish the companies building datacenters, it will punish their customers, end users who want internet access, or to use online services.
They should be ensuring the infrastructure is available and affordable in the places where cheap green power is available.
The problem is that commercial companies are providing broadband, and providing it to an outlying community with a small population and long cable distances isn't profitable.
If you have government or a non profit providing it, the cables are actually cheaper per mile to lay out there since there's less in the way, the cost of digging up city streets is very high because of the disruption it causes. If you just dig a trench alongside an empty highway, or alongside a railroad, you don't cause much disruption and can get the work done much quicker with a lot less red tape.
Well...
What if the government owns the physical infrastructure, or a non profit body, and then providers rent the infrastructure from them... And force them to reinvest any and all profits in improvements of the underlying network.
Kinda like the UK system, but where the owner of the infrastructure isnt trying to compete with the same companies they're providing infrastructure to.
In the UK, BT have to rent out lines wholesale to other ISPs as well as allowing the bigger isps to install kit in exchanges... Regulators actually keep the wholesale prices high, so that the companies installing kit in exchanges don't get priced out of the market by bt wholesale. The problem is that, installing kit in thousands of exchanges is still a high barrier to entry, and its not financially viable to do it everywhere, so some people are still stuck with no alternative to BT's service, which is kept priced high by the regulator.
If you instead do away with loop unbundling, and turn over the infrastructure to an independent non profit, then any isp and any end user will all be in the same boat.
As you've pointed out, privatised telecom provision will never be fair for everyone, and regulation can make things worse in some cases.
There is plenty of documentation of voting irregularities, which at the very least should be investigated before OOXML can be ratified.
The fast track process is for existing "defacto standards" that are widely used and implemented, and only really need a rubber stamp. OOXML is not widely implemented nor widely used at this point, it should have gone through the normal process. Perhaps the recent standardisation of PDF as ISO32000 was through the fast track, and would have deserved being fast tracked.
According to ISO guidelines, standards should reuse existing standards, preferably ISO ones... OOXML does not, it does mostly the same thing as ODF but in a completely different way, it also stores dates in a way conflicting with existing ISO standards, stores country codes in a different way, stores measurements in a different way and more. Thus it is in violation of ISO guidelines and should not have been approved.
There are other more specific issues, plenty documented out on the web... But the 3 above show where they have violated ISO rules, which should at the very least be enough to kick ooxml off the fast track and into the regular process.
As for ignoring it, unfortunately microsoft are large enough that they can force their inferior format on the market, so it will be impossible to ignore. If the market were free, and people were able to choose products based on technical merit microsoft wouldn't be anywhere near as big as they are.
The issue is that OOXML is an inferior "standard", which has been forced through the process using bribery, corruption and ignorance, and all while a superior standard for the same thing already exists. If microsoft were truly interested in standards and interoperability at all, they would have implemented the existing standard, and joined it's steering committee when they were first invited several years ago.
We don't want to be forced to use that inferior format, simply because microsoft bought and paid for enough people at standards boards, as this will hurt the industry as a whole. It's only microsoft who stands to benefit from OOXML, everyone else loses in one way or another.
And some people like the power and flexibility of having such options, the trick is to have sensible defaults, and then file off less common options under an advanced option.
I don't understand why they think that, broadcom hardware isn't generally very good, and there are much better options available. Their BCM4400 10/100 ethernet cards are particularly bad in my experience (their 1gb chipsets seem a lot better, but i'd still rather have an intel card).
Their wifi cards, even with the supported windows drivers, generally under perform compared to other chipsets. They also make linux drivers, but only for specific embedded systems (access points) that use their chipsets.
I've had serious problems with binary drivers...
I have an HP PSC 1315, it worked perfectly with OSX 10.4, I have 10.5 now and the scanner drivers don't work on it, the printer component works but only with apple's generic deskjet driver, not hp's driver.
By contrast, HP makes open source linux drivers for this printer/scanner combo, they still work on the latest ubuntu release which is actually newer than OSX 10.5.
The scanner doesn't work on Vista either...
I also have a WinTV USB2 tv tuner card, there are no vista drivers for it, it still works perfectly with an up to date linux system using the open source in-kernel drivers.
There's also issues with older nvidia cards, nvidia no longer maintains drivers for older cards, so they don't work with newer kernels. They also stopped producing drivers for the IA64 architecture, leaving those few of us who own IA64 hardware stuck with newer kernels or newer nvidia cards, and they don't support PPC or other architectures at all. By contrast, i have open source drivers for some truly ancient video cards, so they all still work perfectly.
As for the ethical aspect...
I would suggest his computer, the data on it and his online accounts etc, now belong to his family along with the rest of his belongings, in the absence of any will saying differently, so if the family want you to look at it that's their right.
It's no different to someone who maintains a paper diary, it falls to their family after death who may or may not read it at their discretion.