I was under the impression that while the interface between a redundant PSU cage and the PSUs that go inside it was usually proprietary they usually connected to the motherboard using standard cables and connectors so you should be able to hook up a standard PSU.
If the cabling is nonstandard there is always wire cutters and either terminal block, crimps or solder and heatshrink;)
Bittorrent is an awesome protocol. It's a leap technologically from the old server/client download model Not really, peer to peer in general and bittorrent in particular is a reaction to a few things
1: upside-down pricing models that make end users upstream essentially free but hosting bandwidth expensive. 2: trying to get central warez hubs off on the technicality that they aren't actually distributing the warez. 3: the lack of multicast on the regular internet.
Technically IMO pushing content unnessaceraly through users very limited upstreams is a majoor step backwards,
Which means you will need to pay for the Fiber-Copper converter meh you will nearly always (there may be a few cases where properties are close enough together to make twisted pair ethernet feasible for linking to a cabinet in the street) need end equipment that can convert from a system that can go at least hundreds of meters to 100BASE-TX or 1000BASE-T for distribution in the house.
IMO if installing a network from scratch it would be pretty crazy not to install fiber all the way to the customers premisis. You might not need the bandwidth now but infrastructure tends to have a long life.
Heck, until PCI-E v3.0 is ready, that would saturate a 16x slot! No it wouldn't, not by a long way! PCIe 1.0 has a raw bitrate of 2.5Gbps and a data byte rate of 250MBps (that's a capital B for bytes) PER LANE.
x4 would cover your uncompressed stream with room to spare (though in reality the netwok card carrying it would probablly be x8 since 10 gigabit is right on the theoretical max of 1.0 x4 and it's good to have some slack).
and any good torrent client Are there any torrent clients that don't do that?!
BTW a quick tip, sometimes you will see one chunk keep failing. Usually that is because something is messing with the stream so if you see that issue turn on "encryption" (I put encryption in quotes because if you think you have any privicy at all when torrenting you are dangerously ignorant).
in the end, a valid HTML 4 page will still be a valid HTML 4 page in the future. It will but the question is not can old pages that were valid by the standards still be viewed in modern browsers it's whether websites should struggle continue to support a browser with an old and broken css implementation.
I remember back when IE6 cane out table based layout was still the normal way of doing things and css was still considered pretty new. Compared to the main alternative at the time (netscape 6 came out about the same time as IE6 and afaict was nto well received) IE6 seemed reasonable. Malware was nowhere near as advanced and pervasive as it is today either. Does anyone know what mozillas css compliance was like back then?
No it didn't, that was just when it moved from mainstream support to extended support (which XP is also in now BTW). Win2K is still in extended support (though not for much longer).
Like HTML 1-4 then. They don't specify formats for images, for example. PNG, GIF, and JPEG are all outside of the HTML spec. Indeed it didn't, lukilly image formats were a more settled issue than video formats and had less patent issues (yes there was a patent on gif but it was widely belived not to apply to decoders and unisys never pressed that issue, there were also some somewhat dubious jpeg patents that came to light just before they expired but thats nothing compared to H.264 which has shitloads of patents surrounding it) so gif jpeg and png were all widely implemented in both open and closed source browsers and there was no problem.
There is such a thing as scope, when it comes to specs. There is, and when you define a spec it's up to you how you define it's scope. But if you define it too narrowly and don't tell people what to use to fill in the gaps then they will fill them in in the way that suits them and you can end up with multiple incompatible systems.
Which is what is happening with the html 5 video stuff. On the one hand you have mozilla who refuse to enter into a patent license agreement that they belive violates the principles of opensource. On the other hand you have nokia and apple and youtube refusing to implement threora (nokia claim it's a submarine patent risk) .
The HTML 5 guys were* stuck between a rock and a hard place. IF they mandate a format** they alienate one of the camps. If they don't then they produce an incomplete standard and websites who want to support all the clients will have to double encode thier content and potentially pay h.264 licensing fees in future.
*Note: afaict HTML5 isn't actually out yet so this could still change. I'm pretty sure there was at one point a draft HTML 5 spec that mandatated theora support.
**To avoid any ambiguity I consider a format to be the combination of everything that goes into defining the file. For a video format that would be the video codec, the audio codec and the container format.
The above makes as much sense as the summary and article. Paypal? Software? Makes perfect sense to me, afaict the main buisness for paypal is side incomes and very small buisnesses where the seller either can't be bothered to or doesn't do enough buisness to justify setting up a proper credit card merchant account.
One such small buisness is selling software downloads. Right now you either have to set up your own infrastructure for that (and remember much of the point of paypal is to avoid setting up infrastructure) or go to some third party operators and at least the one i've dealt with in the past (as a customer) seems to have a pretty shady reputation, the name "payloadz" doesn't exactly inspire confidence either.
Since paypal presumablly has plenty of infrastructure for offering stuff only to appropriately logged in users adding thier own software download service seems like a good way to increase revenue (presumablly they will charge more for this service than for a simple money transfer)
Afaict becoming a tier-1 provider (that is a provider who doesn't have to buy "transit" to get to the whole internet) requires more than just private datacenter to datacenter links. Afaict it also requires sneaky tactics and a gameplan for how to use those tactics from the start.
That is because becoming a tier-1 provider requires persuading all the existing tier-1 providers to peer with you. But of course a provider is likeky to be reluctant to peer with someone they are currently selling transit to....
If I were one of these folks, I'd register my domain in a neutral country. For example, you can register.com domains with Gandi.net in France or with NameForName in Russia, or... well, here's a list [icann.org] of ICANN-accredited registrars, most of which support the.com registry While I do see advantages to this I also see major disadvantages. If someone does try to go after your site with lawsuits they can probablly do it both in the US (where you are resident and where the registry is based) and in your "neutral country" (where the registrar you are using is based).
I could easilly see this ending with needing to fight two paralell lawsuits in different countries!
I agree with keeping them on different providers though, it won't save you if the domain host deliberately takes down your domain for some reason but it does reduce the chance of that happening and protect you against all manner of other things that can go wrong.
Am I supposed to sue? can I even sue given that ICANN policy is not an agreement i'm actually a party too? Not to mention that a lawsuit would cost far less than just paying up unless there are a huge number of domains involved! Am I supposed to as ICANN to sue to release my domain? if so is there a chance in hell of them actually doing so?
My guess is they use them for banning, if you ban a troublemakers user account they will just make another one. If you ban thier IP they will just grab another one from thier ISPs DHCP server.
MAC addresess aren't so obvious, especially if the error just says "your computer is banned" or similar.
Of course this also means that if the little guy does get a patent and someone infringes it often the best option is to either sell it to a patent troll or become one.
If you don't actually make anything then the big company wont be able to countersue for infringing thier patents.
Afaict the idea is every thread does it's work for a given frame, signals the monitor thread that it's done and waits for the monitor thread to signal the next frame.
When all the worker threads signal the monitor thread that they are done the monitor thread swaps the references to the source and destination states and signals the worker threads to begin working on the next frame.
This way you only change the source data structure while the monitor thread is the only thread running. The rest of the time the threads can read it without the need for any locking.
I see two possible outcomes of such a tax scenario.
One is we would end up with one entity handling everything from manufacture (and even manufacture of the tools/materials needed for said manufacture) or import through to final sale.
Another would be arrangements where rather than people paying thier immediate supplier they paid the people up the tree directly.
The former of these would be an environment where it was even harder for non-megacorps to operate. The latter would be an administrative mess.
And that was the real rub. Win 98 was a lot more functional and internet-enabled than Win 95. Really? that's not how I remember it.
Most if not all of the desktop functionality changes in 98 were also available for 95 through windows desktop update (included with internet explorer 4.x) and I don't remember internet explorer behaving any differently either.
As a legacy system 98 was easier to deal with than 95 due to usable USB support (yeah there was a variant of 95 with USB but I never could find any drivers for USB perhipherals for it) support for WDM drivers but neither of these things were that important when 98 was current.
Fat32 was undoubtablly helpful for those replacing thier hard drives but there was a version of 95 which supported that.
Internet connection sharing was also handy but that didn't come until 98 second edition.
Overall unless you wanted to fit a new hard drive to a machine that didn't have OSR2 98 was a meh at the time.
hmm, my experiance has been that hardware either doesn't have 64-bit windows drivers at all or it has drivers that work under XP proffessional x64 edtion (which isn't really XP at all, it's really a desktop edition of 2K3)
The only way to prevent this is to disable password caching completely. Mind you that doesn't actually help you that much either because the malware can simply alter your menus so that every terminal you open has a path containing a keylogging sudo wrapper supplied by the malware.
Yeah I really don't get why MS put the 64-bit stuff in system32 and then added a load of redirection kludge-code to make 32-bit apps work rather than just leaving the 32-bit libraries where they were and adding a system64 directory for 64-bit stuff.
It's worse than that, on the DS games drive the wifi hardware directly so while the DSi does support WPA you can only use it in games that specifically support it.
What makes traffic shaping tricky is that you have to do it at the pinch point(s). Worse depending on network design and loading the pinch point(s) can move arround (though in this situation they probablly won't).
Now in something like a conference network the pinch point is probabblly the connection from the conference network to the internet (assuming all internal backbones are faster than the route offsite). So this is where you have to do your traffic shaping. However this is a high bandwidth point AND is handling traffic from a lot of users, that means giving every user a fair go is actually quite tricky (effectively you need a seperate queue for each user and to monitor when a packet belonging to that user was last successfully transmitted)
Much simpler to simply detect problem users and either block them completely or put them in a different priority class. Then the pinchpoint router only needs a couple of queues.
My understanding is the reason phones moved from nimh to liion was higher energy density afaict laptops moved to liion for the same reason (laptops moved much earlier than phones though).
Slashdot Mirror
I was under the impression that while the interface between a redundant PSU cage and the PSUs that go inside it was usually proprietary they usually connected to the motherboard using standard cables and connectors so you should be able to hook up a standard PSU.
If the cabling is nonstandard there is always wire cutters and either terminal block, crimps or solder and heatshrink ;)
Bittorrent is an awesome protocol. It's a leap technologically from the old server/client download model
Not really, peer to peer in general and bittorrent in particular is a reaction to a few things
1: upside-down pricing models that make end users upstream essentially free but hosting bandwidth expensive.
2: trying to get central warez hubs off on the technicality that they aren't actually distributing the warez.
3: the lack of multicast on the regular internet.
Technically IMO pushing content unnessaceraly through users very limited upstreams is a majoor step backwards,
Which means you will need to pay for the Fiber-Copper converter
meh you will nearly always (there may be a few cases where properties are close enough together to make twisted pair ethernet feasible for linking to a cabinet in the street) need end equipment that can convert from a system that can go at least hundreds of meters to 100BASE-TX or 1000BASE-T for distribution in the house.
IMO if installing a network from scratch it would be pretty crazy not to install fiber all the way to the customers premisis. You might not need the bandwidth now but infrastructure tends to have a long life.
Heck, until PCI-E v3.0 is ready, that would saturate a 16x slot!
No it wouldn't, not by a long way! PCIe 1.0 has a raw bitrate of 2.5Gbps and a data byte rate of 250MBps (that's a capital B for bytes) PER LANE.
x4 would cover your uncompressed stream with room to spare (though in reality the netwok card carrying it would probablly be x8 since 10 gigabit is right on the theoretical max of 1.0 x4 and it's good to have some slack).
They seem a pretty good choice as long as you don't have any significant daytime use. If you do then idnet are probablly a better bet.
O2/Be and SKY are probablly the best services technically but afaict are let down by shitty support.
A&A also offer a Be based service which may be an interesting option.
and any good torrent client
Are there any torrent clients that don't do that?!
BTW a quick tip, sometimes you will see one chunk keep failing. Usually that is because something is messing with the stream so if you see that issue turn on "encryption" (I put encryption in quotes because if you think you have any privicy at all when torrenting you are dangerously ignorant).
in the end, a valid HTML 4 page will still be a valid HTML 4 page in the future.
It will but the question is not can old pages that were valid by the standards still be viewed in modern browsers it's whether websites should struggle continue to support a browser with an old and broken css implementation.
I remember back when IE6 cane out table based layout was still the normal way of doing things and css was still considered pretty new. Compared to the main alternative at the time (netscape 6 came out about the same time as IE6 and afaict was nto well received) IE6 seemed reasonable. Malware was nowhere near as advanced and pervasive as it is today either. Does anyone know what mozillas css compliance was like back then?
No it didn't, that was just when it moved from mainstream support to extended support (which XP is also in now BTW). Win2K is still in extended support (though not for much longer).
Like HTML 1-4 then. They don't specify formats for images, for example. PNG, GIF, and JPEG are all outside of the HTML spec.
Indeed it didn't, lukilly image formats were a more settled issue than video formats and had less patent issues (yes there was a patent on gif but it was widely belived not to apply to decoders and unisys never pressed that issue, there were also some somewhat dubious jpeg patents that came to light just before they expired but thats nothing compared to H.264 which has shitloads of patents surrounding it) so gif jpeg and png were all widely implemented in both open and closed source browsers and there was no problem.
There is such a thing as scope, when it comes to specs.
There is, and when you define a spec it's up to you how you define it's scope. But if you define it too narrowly and don't tell people what to use to fill in the gaps then they will fill them in in the way that suits them and you can end up with multiple incompatible systems.
Which is what is happening with the html 5 video stuff. On the one hand you have mozilla who refuse to enter into a patent license agreement that they belive violates the principles of opensource. On the other hand you have nokia and apple and youtube refusing to implement threora (nokia claim it's a submarine patent risk) .
The HTML 5 guys were* stuck between a rock and a hard place. IF they mandate a format** they alienate one of the camps. If they don't then they produce an incomplete standard and websites who want to support all the clients will have to double encode thier content and potentially pay h.264 licensing fees in future.
*Note: afaict HTML5 isn't actually out yet so this could still change. I'm pretty sure there was at one point a draft HTML 5 spec that mandatated theora support.
**To avoid any ambiguity I consider a format to be the combination of everything that goes into defining the file. For a video format that would be the video codec, the audio codec and the container format.
The above makes as much sense as the summary and article. Paypal? Software?
Makes perfect sense to me, afaict the main buisness for paypal is side incomes and very small buisnesses where the seller either can't be bothered to or doesn't do enough buisness to justify setting up a proper credit card merchant account.
One such small buisness is selling software downloads. Right now you either have to set up your own infrastructure for that (and remember much of the point of paypal is to avoid setting up infrastructure) or go to some third party operators and at least the one i've dealt with in the past (as a customer) seems to have a pretty shady reputation, the name "payloadz" doesn't exactly inspire confidence either.
Since paypal presumablly has plenty of infrastructure for offering stuff only to appropriately logged in users adding thier own software download service seems like a good way to increase revenue (presumablly they will charge more for this service than for a simple money transfer)
Afaict becoming a tier-1 provider (that is a provider who doesn't have to buy "transit" to get to the whole internet) requires more than just private datacenter to datacenter links. Afaict it also requires sneaky tactics and a gameplan for how to use those tactics from the start.
That is because becoming a tier-1 provider requires persuading all the existing tier-1 providers to peer with you. But of course a provider is likeky to be reluctant to peer with someone they are currently selling transit to....
If I were one of these folks, I'd register my domain in a neutral country. For example, you can register .com domains with Gandi.net in France or with NameForName in Russia, or... well, here's a list [icann.org] of ICANN-accredited registrars, most of which support the .com registry
While I do see advantages to this I also see major disadvantages. If someone does try to go after your site with lawsuits they can probablly do it both in the US (where you are resident and where the registry is based) and in your "neutral country" (where the registrar you are using is based).
I could easilly see this ending with needing to fight two paralell lawsuits in different countries!
I agree with keeping them on different providers though, it won't save you if the domain host deliberately takes down your domain for some reason but it does reduce the chance of that happening and protect you against all manner of other things that can go wrong.
Am I supposed to sue? can I even sue given that ICANN policy is not an agreement i'm actually a party too? Not to mention that a lawsuit would cost far less than just paying up unless there are a huge number of domains involved!
Am I supposed to as ICANN to sue to release my domain? if so is there a chance in hell of them actually doing so?
My guess is they use them for banning, if you ban a troublemakers user account they will just make another one. If you ban thier IP they will just grab another one from thier ISPs DHCP server.
MAC addresess aren't so obvious, especially if the error just says "your computer is banned" or similar.
Of course this also means that if the little guy does get a patent and someone infringes it often the best option is to either sell it to a patent troll or become one.
If you don't actually make anything then the big company wont be able to countersue for infringing thier patents.
Afaict the idea is every thread does it's work for a given frame, signals the monitor thread that it's done and waits for the monitor thread to signal the next frame.
When all the worker threads signal the monitor thread that they are done the monitor thread swaps the references to the source and destination states and signals the worker threads to begin working on the next frame.
This way you only change the source data structure while the monitor thread is the only thread running. The rest of the time the threads can read it without the need for any locking.
I see two possible outcomes of such a tax scenario.
One is we would end up with one entity handling everything from manufacture (and even manufacture of the tools/materials needed for said manufacture) or import through to final sale.
Another would be arrangements where rather than people paying thier immediate supplier they paid the people up the tree directly.
The former of these would be an environment where it was even harder for non-megacorps to operate. The latter would be an administrative mess.
And that was the real rub. Win 98 was a lot more functional and internet-enabled than Win 95.
Really? that's not how I remember it.
Most if not all of the desktop functionality changes in 98 were also available for 95 through windows desktop update (included with internet explorer 4.x) and I don't remember internet explorer behaving any differently either.
As a legacy system 98 was easier to deal with than 95 due to usable USB support (yeah there was a variant of 95 with USB but I never could find any drivers for USB perhipherals for it) support for WDM drivers but neither of these things were that important when 98 was current.
Fat32 was undoubtablly helpful for those replacing thier hard drives but there was a version of 95 which supported that.
Internet connection sharing was also handy but that didn't come until 98 second edition.
Overall unless you wanted to fit a new hard drive to a machine that didn't have OSR2 98 was a meh at the time.
hmm, my experiance has been that hardware either doesn't have 64-bit windows drivers at all or it has drivers that work under XP proffessional x64 edtion (which isn't really XP at all, it's really a desktop edition of 2K3)
The only way to prevent this is to disable password caching completely.
Mind you that doesn't actually help you that much either because the malware can simply alter your menus so that every terminal you open has a path containing a keylogging sudo wrapper supplied by the malware.
Yeah I really don't get why MS put the 64-bit stuff in system32 and then added a load of redirection kludge-code to make 32-bit apps work rather than just leaving the 32-bit libraries where they were and adding a system64 directory for 64-bit stuff.
It's worse than that, on the DS games drive the wifi hardware directly so while the DSi does support WPA you can only use it in games that specifically support it.
Nice idea for smaller networks but I doubt having a seperate bucket for every IP scales very well.
What makes traffic shaping tricky is that you have to do it at the pinch point(s). Worse depending on network design and loading the pinch point(s) can move arround (though in this situation they probablly won't).
Now in something like a conference network the pinch point is probabblly the connection from the conference network to the internet (assuming all internal backbones are faster than the route offsite). So this is where you have to do your traffic shaping. However this is a high bandwidth point AND is handling traffic from a lot of users, that means giving every user a fair go is actually quite tricky (effectively you need a seperate queue for each user and to monitor when a packet belonging to that user was last successfully transmitted)
Much simpler to simply detect problem users and either block them completely or put them in a different priority class. Then the pinchpoint router only needs a couple of queues.
My understanding is the reason phones moved from nimh to liion was higher energy density afaict laptops moved to liion for the same reason (laptops moved much earlier than phones though).