something like 95% of listeners cannot tell the difference between 128kbps sample and the original
From what orifice did you grab that stat? I think that you are seeing more and more 192kbps and even a large minority of 256kbps mp3's on file sharing networks is at odds with the statement. In fact there isn't a codec out there that performs better than 'decent' at 128kbps. To get true transparency most lossy algorithms need somewhere north of 200kbps VBR. LAME and Vorbis both do extremely well at those rates, though depending on the sample the each have some shortcomings. My test setup is a SB Audigy going into a $300 bookshelf stereo acting as an amp for my $75 Sennheiser HD 495's. This is above average equipment but by no means expensive or elite. I can generally pick out Vorbis on about 1/20 tracks and LAME on about 1/30 vs each other and the source material in a double blind automated testing I conducted on myself. I have above average hearing (as a near perfect score on an isolation booth test for my recent pre-job physical showed) but not superhuman or anything. I listen to a wide swath of music (classical, jazz, rock, folk, metal, blues, etc) and so my potential to find the shortcomings in a codec are obviously greater than the person who only listens to overproduced pop which is compressed and squeezed to sound good on crappy equipment.
In reality the best price/pollution ratio today is a small turbodiesel. The best example is the Volkswagon Jetta TDI, the Jetta gets real world numbers within 20% of the hybrid's claims (probably higher than the real world performance of the Civic Hybrid for example). Modern turbo diesel engines have eliminated most of the historic problems of diesel engines (soot mostly) but they still have problems with NOX emissions.
Sure, I can do solar smelting on a limited scale. Place a large Fresnel lens over the material to be heated and you can easily achieve temperatures high enough to melt most metals (3000-3500 degrees Farenheit are often quoted). Just because you don't think solar is dense enough to power the worlds energy needs doesn't make it a fact. Besides we can easily make thin, light cells and place them in orbit, then beam the energy down as microwaves. The solar panels are more efficient because the solar energy they use isn't filtered by earths atmosphere and the power provided can be rectified at a fairly high efficiency. These solar farms could be tens of square miles or larger and could produce a significant percentage of the world's energy needs. The problem is launch costs, but if the space elevator becomes a reality then that problem goes away. This is the meeting of many areas of science coming together to eliminate an oncoming problem. Best of all most of the components will help reduce the problem on their own even if the grand plan doesn't come to fruition.
I think we have a winner, the ADC switcher seems to be designed for EXACTLY what the poster was asking. It's the only posted solution which supports 1920x1200 resolution.
In a ring environment there are NO collisions, there IS contention for the token but that doesn't reduce bandwidth at all and generally affects latency less than Ethernet's CSMA/CD. The problem was that it didn't scale well. Of course eventually we went back to ring topologies for FDDI which was very high speed but still not good for connecting large numbers of hosts (ask Case Western Reserve).
Bandwidth is expensive and recurring, server hardware is cheap and getting cheaper all the time (well ok bandwidth is too, just not as quickly). A Citrix farm for a midsized company (100-200 employees) is 3-4 Dual Xeon's with a couple gigs of ram per server, total cost, under $20K easily, whereas each T1 around here runs about $1K/month for managed bandwidth (not including installation). So you have paid for a server in a couple months if you can use one less T line.
Seven clients, whoopee. We run between 30 and 50 clients per server. These are typically Dual Xeon's with 2GB of ram, decent but not terribly expensive servers. The problem with X is that it sends the entire bitmap across every time there is an update, whereas RDP/ICA caches the bitmap and will reuse it, even across sessions. This makes for some heavy network traffic. Add to RDP the advantage of not losing the session because of network or client problems and it's a clear winner. I loved X when I first used it but since starting with my current employer where most of our clients run Citrix I have become a real convert, and not because of techno-religion but because the stuff is just downright better.
Yes, and I was talking about bandwidth, not the particular technology. An RDP session with bitmap caching running at 800*600*16bpp or 1024*768*8bpp is completely usable with the bandwidth of a 28.8 modem, whereas X is often unresponsive over a 604/128 DSL line.
Sorry but X SUCKS compared to RDP/ICA. I can be quite productive using RDP on a 28.8 dialup line if I turn on bitmap caching and turn down the resolution/bpp. X on the otherhand is almost unusable across a slow DSL line. Multiply this times hundreds or thousands of employees and the bandwidth savings are HUGE. X was great for when it was invented but it doesn't hold a candle to RDP/ICA.
Yep, it doesn't require any installation, it has subfolders under the directory where the application is located that it looks to for bookmarks and all the other settings are stored in it's.ini file (just like a good old win 3.1 program =). It doesn't touch the registry and no installation is needed. It will use whatever version of IE it finds on the PC (5.0 or greater is required I think, haven't tried it on an NT4 machine with 4.0 or less) as a rendering engine.
And it comes with all the free spyware you could ask for. IE sucks, it's a bug ridden virus vector. Wrapping it in a shell that adds some of the functionality of a real browser doesn't change that (though I do carry a copy of crazybrowser on my USB memory key for when I can't install a real browser on a clients system). When people see me browsing without any flashing ad's or popups the experience is so jolting that they invariably ask me what I use, I tell them that I use Mozilla with a couple of addons to make my browsing experience much better (I have a click for flash plugin, I've turned on popup blocking, turned animated gif's to once, etc). This almost always results in the person asking trying out Mozilla.
What standard doesn't Exchange follow? It's X.500, SMTP, POP3, IMAP, and LDAP compliant (maybe there are minor inconsistencies between the standards and Exchange's implementation but nothing world shattering, just the kinds of things almost all software that implements a large convoluted standard encounters) so there isn't much to complain about. Sure if you want full functionality you need a client that understands the advanced features but that's a problem with the open standards world not being up to snuff compared to Exchange, not Exchange failing to follow a standard. Just as Outlook added iCal support once the standard is finalized I'm sure that Exchange would support a usefull multiuser calandering standard if such a beast existed (it doesn't, dumping iCal's into a folder doesn't work, especially since large iCal's suck). Exchange is one of those MS products that is fairly difficult to bash for interoperability because it's so all inclusive that it works in just about any environment, and with Exchange 2003 it's hard to knock it for stability which was its historic weakness.
It exists, it's called Exchange server. OWA for webmail and Outlook 2003 for local client. Outlook 2003 defaults to blocking executables by default, but unfortunatly still uses IE for its preview engine (sigh). Luckily OWA works with Mozilla darn well so I don't have to risk infection, it only lacks meeting notification which is an ActiveX plugin for IE clients but has no equivilant for degraded clients.
Re:WEP (in)security assumptions
on
Wi-Fi in the Sky
·
· Score: 1
Simple, check the certificate. I'm assuming that your school is smart enough to have the proxy server doing authentication over SSL so if the page comes up unencrypted then you know you aren't talking to the schools proxy, if it prompts you to accept a new certificate I would likewise be inclined to investigate.
My guess it that they actually NEVER replace individual machines, they put new racks together as capacity demands and when a new rack is put together it is either assigned to an underserved DC or is sent out to replace the oldest/most degraded rack. This way they simply wheel in the new rack, hook it up, and wheel out the old rack, no downtime, no wasted personell time.
Who cares, MAE East and MAE West are MUCH more critical to the functioning of the internet and their position has been known for years. MAE East is on the other side of a wall from a parking deck, if you wanted to do a decent job of crippling the internet a truck bomb on the other side of that wall would do a pretty good job. Of course it would have been better before private peering facilities decentralized a lot of cross domain traffic but it would still cause enough traffic to be screwed up that it would have a very serious impact.
Yep, if MS released a more functional binary only driver for my MS Internet Keyboard Pro or MS Intellipoint Trackball I sure would use it, and I wouldn't care if it needed access to the kernel. MS hardware is a seperate division and I've found that their products are some of the best out there. As to the kernel maintainers, only if I could show that the bug was in fact in the kernel and not in the driver, but I expect that for any third party driver, not just closed source ones. It's not the job of the people on LKML to maintain all possible drivers under the sun, just the ones that are included in the blessed source releases.
Having just reinstalled XP this week I can say that my list is pretty similar: MS Powertoys (use a couple not just TweakUI) WinZip WinRar WinAmp Mozilla suite ACDC32 (classic, can't stand the new bloated version) Sys Internals PS Tools
Actually that's a red herring as no card out there can boost its output enough to go over the limits without an end user modified antennae anyways. The real reason for no open source drivers is that the chipset manufacturers sell the chipsets and driver development kits to the card manufacturers for loads of money, so having freely available drivers limits them to only selling the hardware. As an example Cisco has a fully open source driver for their 802.11b cards because they designed the chip themselves rather than relying on a third party provider so they were free to make open drivers.
The answer is obvious, make a GPL wrapper driver that does nothing but accesses the data structures and communicates via an interface to the closed driver. Playing stupid politics with system info is just a retarded dead end. The info is made available to some classes of drivers because it is usefull, in reality it is usefull to any driver that can benifit from the info, open or not. So witholding the data from closed drivers is just lessening the experience/reliability/etc of people who use Linux but who aren't Open Source zealots. That's the aim of the driver interface but it's a stupid one, and as I pointed out it's easy enough to circumvent.
Reducing net bottlenecks would require eliminating the profit motives of the Tier-2 and Tier-3 ISP's. It is impossible to run a profitable ISP that does not over-subscribe lines AND charges what most people seem willing to pay for broadband ( Europe connections and the single large North America Australia connection. These aren't things some academics studying net usage reports are going to be able to solve, they are purely based on economics.
Better to have planned downtime for system migration then unplanned downtime for system triage and intrusion cleanup. Running RH 7.2 on any kind of network with a chance of outside contact is just as bad if not worse then running Windows 2000 unpatched in a similar situation (at least with 2k Pro the only services running is file sharing). Personally I would be running RHEL 3.0 on most of those boxes and telling RH to get their freaking act together and fix the damn Oracle problems! That's why you are paying them good money and exactly why they have the partnership with Oracle, if your rep can't get you satisfaction ask to talk to his boss. 2-3 updates a year is probably bare minimum just for security updates, even OS/390 nee Z/OS has more than that!
Re:The 89 is banned as well dude...
on
TI-84 Plus Released
·
· Score: 5, Informative
Calculator Policy
You may use almost any four-function, scientific, or graphing calculator on the SAT I and Math Level IC, and Math Level IIC Subject Tests. You are not permitted to use:
* Hand-held minicomputers or laptop computers
* Electronic writing pads or pen-input devices
* Pocket organizers (PDAs) * Calculators with QWERTY (typewriter-like) keypads
* Calculators with paper tape
* Calculators that "talk" or make unusual noises
* Calculators that require an electrical outlet
The bolded entry is why the Ti92 is banned and the Ti89 is not.
Well it's actually a real world problem, a float might not work for several reasons (it's food grade material so cross contamination is a concern, there's no one there to move it because it's an automated process which runs even when the shop is closed, etc). Although it can be seen as a bit contrived because most containers are a known volume and most pumps are rated at flow rates as well as HP, but my dad who has his own chemical sales business does solve problems like this on an almost weekly basis.
Jesus, I took out my dad's calculus text from the 1960's and it was a joke compared to my text from seven years ago. The amount of work required to work complex integrations or multi-derivitives not to mention systems of equations was just insane. You could do maybe a handfull of problems and all you really learned was the steps involved, not the big picture. Going through and doing tons of examples including real world problems made me have not just a firmer grasp on the subject but also an apreciation of its power. A fine example of a real world problem we were given was "you have a cylindrical container of length x and diameter y, using a pump of horsepower z and assuming an efficiency of n how long would you set a timer t so that the tank was filled 90% by volume?" This is a problem that done by hand takes several pages and a lot of error prone work, yet for us it was just one of a series of about 50 homework problems.
Slashdot Mirror
something like 95% of listeners cannot tell the difference between 128kbps sample and the original
From what orifice did you grab that stat? I think that you are seeing more and more 192kbps and even a large minority of 256kbps mp3's on file sharing networks is at odds with the statement. In fact there isn't a codec out there that performs better than 'decent' at 128kbps. To get true transparency most lossy algorithms need somewhere north of 200kbps VBR. LAME and Vorbis both do extremely well at those rates, though depending on the sample the each have some shortcomings. My test setup is a SB Audigy going into a $300 bookshelf stereo acting as an amp for my $75 Sennheiser HD 495's. This is above average equipment but by no means expensive or elite. I can generally pick out Vorbis on about 1/20 tracks and LAME on about 1/30 vs each other and the source material in a double blind automated testing I conducted on myself. I have above average hearing (as a near perfect score on an isolation booth test for my recent pre-job physical showed) but not superhuman or anything. I listen to a wide swath of music (classical, jazz, rock, folk, metal, blues, etc) and so my potential to find the shortcomings in a codec are obviously greater than the person who only listens to overproduced pop which is compressed and squeezed to sound good on crappy equipment.
In reality the best price/pollution ratio today is a small turbodiesel. The best example is the Volkswagon Jetta TDI, the Jetta gets real world numbers within 20% of the hybrid's claims (probably higher than the real world performance of the Civic Hybrid for example). Modern turbo diesel engines have eliminated most of the historic problems of diesel engines (soot mostly) but they still have problems with NOX emissions.
Sure, I can do solar smelting on a limited scale. Place a large Fresnel lens over the material to be heated and you can easily achieve temperatures high enough to melt most metals (3000-3500 degrees Farenheit are often quoted). Just because you don't think solar is dense enough to power the worlds energy needs doesn't make it a fact. Besides we can easily make thin, light cells and place them in orbit, then beam the energy down as microwaves. The solar panels are more efficient because the solar energy they use isn't filtered by earths atmosphere and the power provided can be rectified at a fairly high efficiency. These solar farms could be tens of square miles or larger and could produce a significant percentage of the world's energy needs. The problem is launch costs, but if the space elevator becomes a reality then that problem goes away. This is the meeting of many areas of science coming together to eliminate an oncoming problem. Best of all most of the components will help reduce the problem on their own even if the grand plan doesn't come to fruition.
I think we have a winner, the ADC switcher seems to be designed for EXACTLY what the poster was asking. It's the only posted solution which supports 1920x1200 resolution.
In a ring environment there are NO collisions, there IS contention for the token but that doesn't reduce bandwidth at all and generally affects latency less than Ethernet's CSMA/CD. The problem was that it didn't scale well. Of course eventually we went back to ring topologies for FDDI which was very high speed but still not good for connecting large numbers of hosts (ask Case Western Reserve).
Bandwidth is expensive and recurring, server hardware is cheap and getting cheaper all the time (well ok bandwidth is too, just not as quickly). A Citrix farm for a midsized company (100-200 employees) is 3-4 Dual Xeon's with a couple gigs of ram per server, total cost, under $20K easily, whereas each T1 around here runs about $1K/month for managed bandwidth (not including installation). So you have paid for a server in a couple months if you can use one less T line.
Seven clients, whoopee. We run between 30 and 50 clients per server. These are typically Dual Xeon's with 2GB of ram, decent but not terribly expensive servers. The problem with X is that it sends the entire bitmap across every time there is an update, whereas RDP/ICA caches the bitmap and will reuse it, even across sessions. This makes for some heavy network traffic. Add to RDP the advantage of not losing the session because of network or client problems and it's a clear winner. I loved X when I first used it but since starting with my current employer where most of our clients run Citrix I have become a real convert, and not because of techno-religion but because the stuff is just downright better.
Yes, and I was talking about bandwidth, not the particular technology. An RDP session with bitmap caching running at 800*600*16bpp or 1024*768*8bpp is completely usable with the bandwidth of a 28.8 modem, whereas X is often unresponsive over a 604/128 DSL line.
Sorry but X SUCKS compared to RDP/ICA. I can be quite productive using RDP on a 28.8 dialup line if I turn on bitmap caching and turn down the resolution/bpp. X on the otherhand is almost unusable across a slow DSL line. Multiply this times hundreds or thousands of employees and the bandwidth savings are HUGE. X was great for when it was invented but it doesn't hold a candle to RDP/ICA.
Yep, it doesn't require any installation, it has subfolders under the directory where the application is located that it looks to for bookmarks and all the other settings are stored in it's .ini file (just like a good old win 3.1 program =). It doesn't touch the registry and no installation is needed. It will use whatever version of IE it finds on the PC (5.0 or greater is required I think, haven't tried it on an NT4 machine with 4.0 or less) as a rendering engine.
And it comes with all the free spyware you could ask for. IE sucks, it's a bug ridden virus vector. Wrapping it in a shell that adds some of the functionality of a real browser doesn't change that (though I do carry a copy of crazybrowser on my USB memory key for when I can't install a real browser on a clients system). When people see me browsing without any flashing ad's or popups the experience is so jolting that they invariably ask me what I use, I tell them that I use Mozilla with a couple of addons to make my browsing experience much better (I have a click for flash plugin, I've turned on popup blocking, turned animated gif's to once, etc). This almost always results in the person asking trying out Mozilla.
What standard doesn't Exchange follow? It's X.500, SMTP, POP3, IMAP, and LDAP compliant (maybe there are minor inconsistencies between the standards and Exchange's implementation but nothing world shattering, just the kinds of things almost all software that implements a large convoluted standard encounters) so there isn't much to complain about. Sure if you want full functionality you need a client that understands the advanced features but that's a problem with the open standards world not being up to snuff compared to Exchange, not Exchange failing to follow a standard. Just as Outlook added iCal support once the standard is finalized I'm sure that Exchange would support a usefull multiuser calandering standard if such a beast existed (it doesn't, dumping iCal's into a folder doesn't work, especially since large iCal's suck). Exchange is one of those MS products that is fairly difficult to bash for interoperability because it's so all inclusive that it works in just about any environment, and with Exchange 2003 it's hard to knock it for stability which was its historic weakness.
It exists, it's called Exchange server. OWA for webmail and Outlook 2003 for local client. Outlook 2003 defaults to blocking executables by default, but unfortunatly still uses IE for its preview engine (sigh). Luckily OWA works with Mozilla darn well so I don't have to risk infection, it only lacks meeting notification which is an ActiveX plugin for IE clients but has no equivilant for degraded clients.
Simple, check the certificate. I'm assuming that your school is smart enough to have the proxy server doing authentication over SSL so if the page comes up unencrypted then you know you aren't talking to the schools proxy, if it prompts you to accept a new certificate I would likewise be inclined to investigate.
My guess it that they actually NEVER replace individual machines, they put new racks together as capacity demands and when a new rack is put together it is either assigned to an underserved DC or is sent out to replace the oldest/most degraded rack. This way they simply wheel in the new rack, hook it up, and wheel out the old rack, no downtime, no wasted personell time.
Who cares, MAE East and MAE West are MUCH more critical to the functioning of the internet and their position has been known for years. MAE East is on the other side of a wall from a parking deck, if you wanted to do a decent job of crippling the internet a truck bomb on the other side of that wall would do a pretty good job. Of course it would have been better before private peering facilities decentralized a lot of cross domain traffic but it would still cause enough traffic to be screwed up that it would have a very serious impact.
Yep, if MS released a more functional binary only driver for my MS Internet Keyboard Pro or MS Intellipoint Trackball I sure would use it, and I wouldn't care if it needed access to the kernel. MS hardware is a seperate division and I've found that their products are some of the best out there. As to the kernel maintainers, only if I could show that the bug was in fact in the kernel and not in the driver, but I expect that for any third party driver, not just closed source ones. It's not the job of the people on LKML to maintain all possible drivers under the sun, just the ones that are included in the blessed source releases.
Having just reinstalled XP this week I can say that my list is pretty similar:
MS Powertoys (use a couple not just TweakUI)
WinZip
WinRar
WinAmp
Mozilla suite
ACDC32 (classic, can't stand the new bloated version)
Sys Internals PS Tools
Actually that's a red herring as no card out there can boost its output enough to go over the limits without an end user modified antennae anyways. The real reason for no open source drivers is that the chipset manufacturers sell the chipsets and driver development kits to the card manufacturers for loads of money, so having freely available drivers limits them to only selling the hardware. As an example Cisco has a fully open source driver for their 802.11b cards because they designed the chip themselves rather than relying on a third party provider so they were free to make open drivers.
The answer is obvious, make a GPL wrapper driver that does nothing but accesses the data structures and communicates via an interface to the closed driver. Playing stupid politics with system info is just a retarded dead end. The info is made available to some classes of drivers because it is usefull, in reality it is usefull to any driver that can benifit from the info, open or not. So witholding the data from closed drivers is just lessening the experience/reliability/etc of people who use Linux but who aren't Open Source zealots. That's the aim of the driver interface but it's a stupid one, and as I pointed out it's easy enough to circumvent.
Reducing net bottlenecks would require eliminating the profit motives of the Tier-2 and Tier-3 ISP's. It is impossible to run a profitable ISP that does not over-subscribe lines AND charges what most people seem willing to pay for broadband ( Europe connections and the single large North America Australia connection. These aren't things some academics studying net usage reports are going to be able to solve, they are purely based on economics.
Better to have planned downtime for system migration then unplanned downtime for system triage and intrusion cleanup. Running RH 7.2 on any kind of network with a chance of outside contact is just as bad if not worse then running Windows 2000 unpatched in a similar situation (at least with 2k Pro the only services running is file sharing). Personally I would be running RHEL 3.0 on most of those boxes and telling RH to get their freaking act together and fix the damn Oracle problems! That's why you are paying them good money and exactly why they have the partnership with Oracle, if your rep can't get you satisfaction ask to talk to his boss. 2-3 updates a year is probably bare minimum just for security updates, even OS/390 nee Z/OS has more than that!
Bullshit, from the collegeboard website:
Calculator Policy You may use almost any four-function, scientific, or graphing calculator on the SAT I and Math Level IC, and Math Level IIC Subject Tests. You are not permitted to use:
* Hand-held minicomputers or laptop computers
* Electronic writing pads or pen-input devices
* Pocket organizers (PDAs)
* Calculators with QWERTY (typewriter-like) keypads
* Calculators with paper tape
* Calculators that "talk" or make unusual noises
* Calculators that require an electrical outlet
The bolded entry is why the Ti92 is banned and the Ti89 is not.
Well it's actually a real world problem, a float might not work for several reasons (it's food grade material so cross contamination is a concern, there's no one there to move it because it's an automated process which runs even when the shop is closed, etc). Although it can be seen as a bit contrived because most containers are a known volume and most pumps are rated at flow rates as well as HP, but my dad who has his own chemical sales business does solve problems like this on an almost weekly basis.
Jesus, I took out my dad's calculus text from the 1960's and it was a joke compared to my text from seven years ago. The amount of work required to work complex integrations or multi-derivitives not to mention systems of equations was just insane. You could do maybe a handfull of problems and all you really learned was the steps involved, not the big picture. Going through and doing tons of examples including real world problems made me have not just a firmer grasp on the subject but also an apreciation of its power. A fine example of a real world problem we were given was "you have a cylindrical container of length x and diameter y, using a pump of horsepower z and assuming an efficiency of n how long would you set a timer t so that the tank was filled 90% by volume?" This is a problem that done by hand takes several pages and a lot of error prone work, yet for us it was just one of a series of about 50 homework problems.