Afaict you can't easilly blacklist things like WGA notifications and MSRT, you can blacklist the current version but as soon as there is a new version they will be back.
Actually MS does offer an IPV6 stack for 2K though they say it is experimental and some of the API calls are in a different dll from in later releases.
BTW it is really not terriblly difficult to use getaddrinfo/getnameinfo if availible and fall back to gethostbyname/gethostbyaddr if it isn't. Afaict all the other winsock APIs are the same regardless of whether IPV6 is supported or not.
Among other things win2k brought WDM to the NT line. WDM added support for plug and play and allowed hardware vendors to develop a single driver for both 98/ME and 2K (and XP and for the most part vista too).
And decent directx support (afaict NT 4 had some support for directx but it was pretty crappy).
And USB support (afaict there was some third party stuff for NT 4 but few devices worked with it)
2K combined many of the important features of the 9x line (plug and play, wide hardware support, directx) with the stability and ability to handle large numbers of apps open at once.
2K to XP was a fairly minor change and that means if you are supporting XP then unless you use some really exotic apis your app or driver will most likely work just as well under 2K.
I dunno about typical american practices but here in the UK all lightswitches are rated at least 6A and the standard lighting cable is rated at over 10A in most installation conditions (volt drop permitting of course)
So switching to more efficiant lighting will not significantly affect installation costs. Adding more lighting points on the other hand would probablly significantly increase installation costs (mostly in terms of labour to connect them all up)
The experiance of everyone I know who has tried LED light fixtures is they simply don't have the ability to decently light a room (whether this is due to thier being simply less light output or whether it is some other characteristic of the light I don't know) while CFLs do. They are also FAR more expensive than CFLs.
Afaict some programmers (especially of the freeware/shareware variety) think that thier users think that smaller executables mean less bloated apps. If a developer belives this then it is very tempting to use a packer to reduce the apparent size of thier executable, especially if others are doing it too.
Yeah, the big chain stores try to rope people in with low base prices or special offers and then rip people off on extended warranties and accessories. Very annoying.
BTW you might want to take a look if any of the online computer parts vendors you know of are based near you and if so whether they have a sales counter at thier warehouse. If there are they can be a good place to get computer parts and cables at reasonable prices without having to wait for delivery.
and they are probably scared to death of the possibility of being on the wrong side of one of these. The US (and afaict many other countries too) has set statutory damages for copyright infringement that will easilly bankrupt most filesharers if the court can be convinced that the evidence is in order etc.
It's kinda like the lottery in reverse, you are unlikely to be hit but if you are then you are most likely seriously screwed and your best bet is probablly to pay the settlement demand. Maybe you can get off by forcing them to prove the files were actually downloaded rather than just made availible but I wouldn't bet on it.
Lukilly for users the XFree86 project practically committed suicide with a license change that many linux distros belived made the license of the X libraries incompatible with the license of apps that used them. This left the distros with little choice but to switch to a fork
Otherwise it wouldn't surprise me if we were still stuck with the crappy leaders of the xfree86 project.
Say there are 100 configurations (and there are probablly more) of base hardware. 100 possible perhiperals (and there are surely more) each of which may have it's driver installed or not installed and 10 firmware revisions. that is about 10^33 combinations
Fact is while there are less possible mac configurations than PC configurations there are still many orders of magnitude more possible combinations of hardware firmware revision and installed drivers than there are machines in existance let alone machines in the manufacturers test labs.
Given that we don't have either a statistical analisis of the failed machines and what they do and don't have in common or any information on why the display fails to come up on those machines I don't think we can reasonablly blame apples testing procedures.
Have you ever tried putting a db server on an openmosix cluster? If so did you gain anything from doing so? (personally I suspect that doing so would give little if any more performance than a single box)
True, but normal people when their computer stops working don't even try to fix it. They contact someone who can be it a friend/family or a professional. That person will almost certainly have thier own means of accessing the internet.
And afaict the time consuming but obvious brute force soloution of "stick original disks in and reinstall everything" will get the machine back up in this scenario.
What def would you use? The general definition seems to be
In a state where the hardware is fine but the software is fucked to the point where the system cannot be recovered without extrodinary measures.
Some measures that I would consider extraordinary: * opening the case and soldering extra wires/connectors to the board * buying special programming hardware * removing components that aren't intended to be user removable * creating a media shitstorm and hence bullying the manufacturer into releasing previously hidden recovery information
That was afaict not the case here, the firmware screen was still coming up so a reinstall should have been possible and the initial software issue could be directly installed if the user had an external keyboard monitor and mouse (hardly specialist hardware)
I wouldn't consider either a normal OS reinstall or using an external keyboard monitor and mouse to control the machine to be extraordinary measures.
Well phones (regardless of whether they are landline or cellular) have ALWAYS been an ongoing cost since communication over a shared infrastructure is naturally a service.
and much as people may bitch about thier (cell)phone bills the fact is you get far more for your money than you used to say 10 years ago. This patents proposal seems to imply that the service will be far more expensive than the conventional service model for all but the lightest uses.
SAAS is taking something that is naturally a product and turning it into a service. Like rental it has it's place but forcing people into it (I don't think MS is suicidal enough to try and force people into it but I could be wrong) for software they could previously purchase outright is going to seriously piss them off and make them far more serious about considering migration to alternatives.
IMO trying to force people into subscriptions and/or pay by the hour is likely to cause many people to like you say screw it and either take the pain of moving to alternatives or stick with old versions (many are doing that anyway) and pirate extra copies if they can't get them legally.
And if too many people say screw it then the network effects advantage that keeps ms office alive will disapear (while ooo is a bloated pig that can be made up for with extra hardware)
Despite this patent I don't think MS is suicidal enough to make subscriptions/pay by the hour the only option.
Though IIRC MS is trying to use the carrot of lower prices and other side benifits to tempt corporations and academic institutions into subscriptions deals that they then become basically stuck with.
The big thing IMO is using a consistant (and at least in the case of MS office consistant means the same version, I dunno if openoffice is better about keeping thier layout engine consistent between versions) office suite both among machines you use and between the machines you use and the machines people you work with use.
Sure for simple documents conversions are possible but for complex documents wysiwyg word processing basically relies on everyone having a layout engine with the exact same behaviour (pdf gets arround this by doing a lot of the layout in advance but this loses editability).
So if thier lecturers all use office 2003 and the uni machines all have office 2003 then the path of least resistance is to use office 2003 on thier own machine(s). Whether they buy it at the academic discount price or pirate it depends on thier circumstances beliefs (some universities even have a subscription which allows students to install it on thier own machines without paying)
plus at least here in the uk they will probablly have used at least one of office 2K, office XP or office 2K3 at school or "6th form college" before they went to university.
plus at least in my experiance openoffice is a bloated pig compared to office 2K to 2K3.
I have not yet used office 2K7 on a serious enough basis to comment on whether it is more or less shit than openoffice. It is certainly very different from both openoffice and older versions of MS office.
mmm, pretty high actually. Remember that while the resoloution was low by modern standards there was absoloutely no use made of data compression techniques. The composite video signal was more or less (there were some complications due to the recording gaps created by the helical scan system) recorded directly onto the tape.
most home recordable VHS tapes can store either three or four hours of PAL video at thier standard speed (tapes used for prerecorded releases are usually shorter and I have seen five hour tapes availible for sale before), according to wikipedia NTSC vhs tapes use a higher tape speed and therefore store less hours) most later VCRs also support "long play" and "extended play" modes which cram more onto the tape (at the cost of lower quality).
Wikipedia reckons VHS tape has about 3mhz of bandwith (they don't say if that is at PAL or NTSC tape speeds, i'm assuming NTSC since I suspect the wikipedia article was written from american sources) with a SNR of 43db.
If we assume the noise is white guasian noise then per the shannon-hartly theorem.
43db expressed as a power ratio is approximate 20000 the base 2 logarithm of this is approximately 14
per the shannon-hartly theorem (ignoring the +1 since it is negligable) with such a large SNR) this would give a maximum achivable reliable data rate of 14*3=42Mbps
multiply that by 10000 seconds (about 3 hours) and you get a theoretical capacity of about 52GB.
Of course considerations of maintaining reliability with poor quality equipment and the fact that perfect codecs don't exist mean the real capacity given by products that used vhs tape for data storage was much lower. Typically a few gigabyte per tape IIRC.
thumb drives is the popular name for usb sticks Maybe it is in your circle of aquaintences, arround here (EEE deparment, university of manchester) most people seem to call them pendrives with flash stick, memory stick and USB stick all being fairly common. I don't think i've ever met anyone in person who uses the name thumdrive.
Sony's memory sticks lost out when the thumbdrives hit 1GB and got real cheap and were further undermined by the widespread adoption of flash by cell and camera manufacturers -plus I'm not sure they were adopted by many 3rd parties I was always under the impression that the only reason people bought sony memory sticks was because they had other sony hardware (camera, music player, aibo etc)that relied on them. IIRC thier price per megabyte was always higher than other flash formats (CF, SM, SD, XD, USB sticks and so on)
For the netbook though I would point out that there were both wince machines (e.g. the hp journada) and ultraportable laptops (e.g. the toshiba libretto) arround at the same time in similar form factors.
IMO the modern netbook fulfills two niches 1: those who want an ultraportable laptop to run thier standard software on the move (the psion netbook couldn't do this as it used an arm processor and epoc) 2: those who want a machine for mobile internet (which didn't really exist in any decent form when the psion netbook came out).
This leaves me wondering who exactly the psion netbook was aimed at.
The re-routing over the hours/days following the event is pretty much what would be expected given that theese are commercial links that companies have to pay to send thier data over.
Presumablly someone at each ISP has to do (perhaps only approximately and mentally) a cost-benifit analysis and decide that the cost of sending the data via an alternative route is worth the reputation saved by getting thier customers back online quickly.
Also afaict the middle east isn't the most politically stable of areas and therefore I wonder if a lack of land based cabling makes the problem worse.
Slashdot Mirror
IIRC WGA tends to err on the side of caution if it sees something it can't understand. Probabblly to keep the false positive rate down.
By the way, did you know that the Xbox 360 is a PowerPC box too?
As is the wii..............
Afaict you can't easilly blacklist things like WGA notifications and MSRT, you can blacklist the current version but as soon as there is a new version they will be back.
Actually MS does offer an IPV6 stack for 2K though they say it is experimental and some of the API calls are in a different dll from in later releases.
BTW it is really not terriblly difficult to use getaddrinfo/getnameinfo if availible and fall back to gethostbyname/gethostbyaddr if it isn't. Afaict all the other winsock APIs are the same regardless of whether IPV6 is supported or not.
Among other things win2k brought WDM to the NT line. WDM added support for plug and play and allowed hardware vendors to develop a single driver for both 98/ME and 2K (and XP and for the most part vista too).
And decent directx support (afaict NT 4 had some support for directx but it was pretty crappy).
And USB support (afaict there was some third party stuff for NT 4 but few devices worked with it)
2K combined many of the important features of the 9x line (plug and play, wide hardware support, directx) with the stability and ability to handle large numbers of apps open at once.
2K to XP was a fairly minor change and that means if you are supporting XP then unless you use some really exotic apis your app or driver will most likely work just as well under 2K.
I dunno about typical american practices but here in the UK all lightswitches are rated at least 6A and the standard lighting cable is rated at over 10A in most installation conditions (volt drop permitting of course)
So switching to more efficiant lighting will not significantly affect installation costs. Adding more lighting points on the other hand would probablly significantly increase installation costs (mostly in terms of labour to connect them all up)
The experiance of everyone I know who has tried LED light fixtures is they simply don't have the ability to decently light a room (whether this is due to thier being simply less light output or whether it is some other characteristic of the light I don't know) while CFLs do. They are also FAR more expensive than CFLs.
Afaict some programmers (especially of the freeware/shareware variety) think that thier users think that smaller executables mean less bloated apps. If a developer belives this then it is very tempting to use a packer to reduce the apparent size of thier executable, especially if others are doing it too.
Yeah, the big chain stores try to rope people in with low base prices or special offers and then rip people off on extended warranties and accessories. Very annoying.
BTW you might want to take a look if any of the online computer parts vendors you know of are based near you and if so whether they have a sales counter at thier warehouse. If there are they can be a good place to get computer parts and cables at reasonable prices without having to wait for delivery.
and they are probably scared to death of the possibility of being on the wrong side of one of these.
The US (and afaict many other countries too) has set statutory damages for copyright infringement that will easilly bankrupt most filesharers if the court can be convinced that the evidence is in order etc.
It's kinda like the lottery in reverse, you are unlikely to be hit but if you are then you are most likely seriously screwed and your best bet is probablly to pay the settlement demand. Maybe you can get off by forcing them to prove the files were actually downloaded rather than just made availible but I wouldn't bet on it.
Lukilly for users the XFree86 project practically committed suicide with a license change that many linux distros belived made the license of the X libraries incompatible with the license of apps that used them. This left the distros with little choice but to switch to a fork
Otherwise it wouldn't surprise me if we were still stuck with the crappy leaders of the xfree86 project.
Say there are 100 configurations (and there are probablly more) of base hardware. 100 possible perhiperals (and there are surely more) each of which may have it's driver installed or not installed and 10 firmware revisions.
that is about 10^33 combinations
Fact is while there are less possible mac configurations than PC configurations there are still many orders of magnitude more possible combinations of hardware firmware revision and installed drivers than there are machines in existance let alone machines in the manufacturers test labs.
Given that we don't have either a statistical analisis of the failed machines and what they do and don't have in common or any information on why the display fails to come up on those machines I don't think we can reasonablly blame apples testing procedures.
Have you ever tried putting a db server on an openmosix cluster? If so did you gain anything from doing so? (personally I suspect that doing so would give little if any more performance than a single box)
True, but normal people when their computer stops working don't even try to fix it. They contact someone who can be it a friend/family or a professional. That person will almost certainly have thier own means of accessing the internet.
And afaict the time consuming but obvious brute force soloution of "stick original disks in and reinstall everything" will get the machine back up in this scenario.
Still this is definately a major fuckup.
What def would you use?
The general definition seems to be
In a state where the hardware is fine but the software is fucked to the point where the system cannot be recovered without extrodinary measures.
Some measures that I would consider extraordinary:
* opening the case and soldering extra wires/connectors to the board
* buying special programming hardware
* removing components that aren't intended to be user removable
* creating a media shitstorm and hence bullying the manufacturer into releasing previously hidden recovery information
That was afaict not the case here, the firmware screen was still coming up so a reinstall should have been possible and the initial software issue could be directly installed if the user had an external keyboard monitor and mouse (hardly specialist hardware)
I wouldn't consider either a normal OS reinstall or using an external keyboard monitor and mouse to control the machine to be extraordinary measures.
MS has wanted to push people to SAAS and/or monthly subscriptions for years but i'm sure they know that pushing too hard would be suicidal.
I bet if this does see the light of day it will be as an additional option and few people will buy it.
Well phones (regardless of whether they are landline or cellular) have ALWAYS been an ongoing cost since communication over a shared infrastructure is naturally a service.
and much as people may bitch about thier (cell)phone bills the fact is you get far more for your money than you used to say 10 years ago. This patents proposal seems to imply that the service will be far more expensive than the conventional service model for all but the lightest uses.
SAAS is taking something that is naturally a product and turning it into a service. Like rental it has it's place but forcing people into it (I don't think MS is suicidal enough to try and force people into it but I could be wrong) for software they could previously purchase outright is going to seriously piss them off and make them far more serious about considering migration to alternatives.
IMO trying to force people into subscriptions and/or pay by the hour is likely to cause many people to like you say screw it and either take the pain of moving to alternatives or stick with old versions (many are doing that anyway) and pirate extra copies if they can't get them legally.
And if too many people say screw it then the network effects advantage that keeps ms office alive will disapear (while ooo is a bloated pig that can be made up for with extra hardware)
Despite this patent I don't think MS is suicidal enough to make subscriptions/pay by the hour the only option.
Though IIRC MS is trying to use the carrot of lower prices and other side benifits to tempt corporations and academic institutions into subscriptions deals that they then become basically stuck with.
The big thing IMO is using a consistant (and at least in the case of MS office consistant means the same version, I dunno if openoffice is better about keeping thier layout engine consistent between versions) office suite both among machines you use and between the machines you use and the machines people you work with use.
Sure for simple documents conversions are possible but for complex documents wysiwyg word processing basically relies on everyone having a layout engine with the exact same behaviour (pdf gets arround this by doing a lot of the layout in advance but this loses editability).
So if thier lecturers all use office 2003 and the uni machines all have office 2003 then the path of least resistance is to use office 2003 on thier own machine(s). Whether they buy it at the academic discount price or pirate it depends on thier circumstances beliefs (some universities even have a subscription which allows students to install it on thier own machines without paying)
plus at least here in the uk they will probablly have used at least one of office 2K, office XP or office 2K3 at school or "6th form college" before they went to university.
plus at least in my experiance openoffice is a bloated pig compared to office 2K to 2K3.
I have not yet used office 2K7 on a serious enough basis to comment on whether it is more or less shit than openoffice. It is certainly very different from both openoffice and older versions of MS office.
mmm, pretty high actually. Remember that while the resoloution was low by modern standards there was absoloutely no use made of data compression techniques. The composite video signal was more or less (there were some complications due to the recording gaps created by the helical scan system) recorded directly onto the tape.
most home recordable VHS tapes can store either three or four hours of PAL video at thier standard speed (tapes used for prerecorded releases are usually shorter and I have seen five hour tapes availible for sale before), according to wikipedia NTSC vhs tapes use a higher tape speed and therefore store less hours) most later VCRs also support "long play" and "extended play" modes which cram more onto the tape (at the cost of lower quality).
Wikipedia reckons VHS tape has about 3mhz of bandwith (they don't say if that is at PAL or NTSC tape speeds, i'm assuming NTSC since I suspect the wikipedia article was written from american sources) with a SNR of 43db.
If we assume the noise is white guasian noise then per the shannon-hartly theorem.
43db expressed as a power ratio is approximate 20000 the base 2 logarithm of this is approximately 14
per the shannon-hartly theorem (ignoring the +1 since it is negligable) with such a large SNR) this would give a maximum achivable reliable data rate of 14*3=42Mbps
multiply that by 10000 seconds (about 3 hours) and you get a theoretical capacity of about 52GB.
Of course considerations of maintaining reliability with poor quality equipment and the fact that perfect codecs don't exist mean the real capacity given by products that used vhs tape for data storage was much lower. Typically a few gigabyte per tape IIRC.
Though i'm sure the pirate networks will make sure it remains availible :/
thumb drives is the popular name for usb sticks
Maybe it is in your circle of aquaintences, arround here (EEE deparment, university of manchester) most people seem to call them pendrives with flash stick, memory stick and USB stick all being fairly common. I don't think i've ever met anyone in person who uses the name thumdrive.
Sony's memory sticks lost out when the thumbdrives hit 1GB and got real cheap and were further undermined by the widespread adoption of flash by cell and camera manufacturers -plus I'm not sure they were adopted by many 3rd parties
I was always under the impression that the only reason people bought sony memory sticks was because they had other sony hardware (camera, music player, aibo etc)that relied on them. IIRC thier price per megabyte was always higher than other flash formats (CF, SM, SD, XD, USB sticks and so on)
Psions in general were indeed nice machines
For the netbook though I would point out that there were both wince machines (e.g. the hp journada) and ultraportable laptops (e.g. the toshiba libretto) arround at the same time in similar form factors.
IMO the modern netbook fulfills two niches
1: those who want an ultraportable laptop to run thier standard software on the move (the psion netbook couldn't do this as it used an arm processor and epoc)
2: those who want a machine for mobile internet (which didn't really exist in any decent form when the psion netbook came out).
This leaves me wondering who exactly the psion netbook was aimed at.
The things you describe sound like military paranoia. I imagine things would be better on civilian repair ships.
The re-routing over the hours/days following the event is pretty much what would be expected given that theese are commercial links that companies have to pay to send thier data over.
Presumablly someone at each ISP has to do (perhaps only approximately and mentally) a cost-benifit analysis and decide that the cost of sending the data via an alternative route is worth the reputation saved by getting thier customers back online quickly.
Also afaict the middle east isn't the most politically stable of areas and therefore I wonder if a lack of land based cabling makes the problem worse.